Session PP1 - Poster Session VIII.
POSTER session, Thursday afternoon, November 18
Exhibit Hall A, SCC

[PP1.001] NSTX II: Startup, Fast Ions and Diagnostics, MAST, PEGASUS, CDX-U and Other ST

[PP1.002] Real-time equilibrium reconstruction and isoflux control of plasma shape and position in NSTX

The rtEFIT - isoflux algorithm for real-time plasma control, developed at General Atomics, has been modified for control of the plasma equilibrium in NSTX. Real-time equilibrium reconstruction using the rtEFIT code provides the shape of the plasma boundary based on 62 magnetic field and flux measurements, 11 poloidal field coil current measurements as well as 9 loop voltage measurements to estimate the axi-symmetric components of the eddy currents induced in the vacuum vessel. The calculated plasma boundary is used in isoflux control algorithms that generate voltage requests to the coil power supplies based on the difference from the requested boundary flux and positions of the X-points. These plasma boundaries compare well to those reconstructed using the off-line equilibrium reconstruction code EFIT. Recent efforts reduced the latency in the real-time data flow from the sensors to the power supplies from ~ 4 ms to less than 1ms and improved vertical position control which allows control of plasmas nearer stability boundaries.

[PP1.003] A solenoid-free current start-up scenario utilizing outer poloidal field coils*

Elimination of the in-board solenoid is not only required for the spherical torus reactors but would also be desirable for advanced tokamak reactors. The challenge for using only the outer PF coils for start-up is the difficulty of creating a sufficiently high quality field null region while retaining the poloidal flux needed for current ramp-up. It is shown that a few pairs of PF coils can provide a field null for a few ms with a large region of low transverse field in which an ionization avalanche can develop in the applied toroidal E-field with the aid of strong pre-ionization. Preliminary experimental and modeling work has been performed on NSTX aimed at quantifying the field null requirements in terms of the Lloyd parameter, the null size and its duration, while optimizing the loop voltage and the available flux. Different combinations of PF coils were used to investigate the relationship between the size of the region where E_TB_T/B_P = 0.1 kV/m and the breakdown. Fast camera and magnetic diagnostics clearly show plasma initiation for several ms. The vacuum field patterns and flux surfaces of the generated plasma and analysis of the plasma evolution with the DINA code will be presented.

*This work supported by KAIST and DoE Contract No. DE-AC02-76CH03073.

[PP1.004] Simulation of Non-Solenoidal Current Rampup in NSTX

The Spherical Tokamak (ST) concept relies on minimizing the inboard radial build of the device to remain compact. This requires that the solenoid, typically used for inductive current ramp, be eliminated. Present experimental ST�s still use a solenoid, but future devices will need to remove it, so that techniques for ramping up the plasma current without the solenoid need to be developed. Part of the NSTX program is to develop non-solenoidal breakdown and startup, and rampup techniques. The discharge breaks into three phases: the breakdown and startup achieved with Coaxial Helicity Injection (CHI), outboard Poloidal Field (PF) coils, or RF heating and PF coils; early current rampup with RF at low plasma current; and later current rampup obtained with Neutral Beam Injection (NBI) at higher plasma current. The time-scales for this type of current rampup are long compared to inductive ramps due to the use of non-inductive methods. Simulations with the Tokamak Simulation Code (TSC) indicate that is should be possible for NSTX to demonstrate this non-solenoidal current rampup, although the available pulse lengths do not allow the connection to a high performance plasma at the end of the current ramp. An experiment will be used to prescribe the density and temperature evolutions. High Harmonic Fast Wave (HHFW) is used for the early ramp phase, transitioning to NBI when Ip is sufficiently high to confine the beam particles. Various starting points for the HHFW will be examined. The poloidal flux, plasma current, and volt-second evolutions will be examined to understand the various contributions to the current buildup.

[PP1.005] Pitch angle resolved measurements of neutral beam ion loss from NSTX

A scintillator based fast ion loss diagnostic has been installed in NSTX to measure the energy and pitch angle distributions of neutral beam ion loss from NSTX plasmas. Loss is observed at the primary beam energy (80 keV ordinarily, but sometimes as low as 60 keV or as high as 100 keV in this year�s experimental campaign). Therefore the observed losses are prompt, occurring before the beam ions slow down significantly. The amplitude of the loss is larger at low plasma currents (<500 kA), as predicted. Various loss signatures have been seen, depending upon details of the discharge condition, including a continuous band of losses over a broad range in pitch angle, and loss localized to one or several discrete pitch angles. MHD activity is frequently correlated with enhanced loss at high pitch angle.

[PP1.006] Compressional Alfvén eigenmodes in NSTX help to explain anomalous ICRH driven fast ion energy diffusion

Observation and identification of Compressional Alfvén Eigenmodes (CAEs) in National Spherical Torus experiments (NSTX) offer new supporting evidencies to the idea of possible anomalous beam ion energy diffusion due to interaction with high frequency oscillations. Such anomaly observed in TFTR has not been explained and was earlier suggested to be driven by large amplitude cavity modes: such as CAEs. The polarization of the observed magnetic field oscillations along the equilibrium magnetic field and the instability frequency dispersion help to identify the eigenmodes responsible for the instability as CAEs. On the other hand theoretically it was pointed out that CAEs may be driven to a large amplitude by ICRH and result in strong energy diffusion of beam ions in TFTR plasma. In this work we bridge this theoretical hypothesis with observations.

[PP1.007] Effect of Modification of the Fast-ion Distribution Function on the Nonlinear Evolution of Alfvenic Instabilities

Injection of over 2 MW of deuterium neutral beams into a helium L-mode plasma produced instabilities with rapid frequency sweeps or ``chirping.'' Some instabilities with steady \sim100 kHz frequencies were also produced. In the Berk-Breizman model of frequency chirping, resonant fast ions form holes and clumps in phase space [1]. Increased collisionality of the fast ions can scatter resonant ions from the potential well, suppressing the chirping. To test this idea, 2-3 MW of high-harmonic fast wave radio-frequency (RF) heating was applied in 30-ms pulses during strong chirping. Although neutral-particle measurements indicate effective perpendicular heating of the fast ions, the chirping behavior was hardly affected. In contrast, RF heating altered the frequency and amplitude of the constant-frequency modes in ~10 ms, which is the timescale for modification of the entire fast-ion distribution function by the RF. The effect on the constant-frequency modes was most pronounced for more perpendicular angles of beam injection.

[1] H.L. Berk et al., Phys. Plasmas 6 (1999) 3102.

[PP1.008] Status of the High-k Scattering System on NSTX

A high-k scattering system is currently being installed on NSTX to directly observe density fluctuations on the scale of the electron gyro-radius as well as ion gyro-radius. A high power microwave source providing \sim200 mW at 280 GHz (\lambda=1.07 mm) is used as the probe beam source. The probe beam is launched in the ordinary mode at 5^\circ to the midplane with the 5 cm beam waist positioned in the scattering region. A five-channel heterodyne receiver system detects scattered signals from fluctuations with radial wavenumbers \left|k_r\right|\approx 0-20 cm^-1. The collection optics and detection system will be cross checked with a TPX acoustic cell which provides scattered signals with relatively calibrated fluctuation amplitudes at known wavenumbers. Diffraction, refraction, and polarization rotation effects of the probe and scattered beams as well as scattering volume constriction due to magnetic field shear and curvature are addressed.

This work was supported by the U.S. Department of Energy under contract numbers DE-AC02-76CH03073, DE-FG03-95ER54295, and DE-FG03-99ER54531.

[PP1.009] Solid State Neutral Particle Analyzer Array on NSTX

A Solid State Neutral Particle Analyzer (SSNPA) array has been installed on the National Spherical Torus Experiment (NSTX) to measure the energy distribution of charge exchange fast neutral particles. The array consists of four Si diode detectors on chords with fixed tangency radii (60, 90, 100, and 120 cm), which view across the three co-injection neutral beam (NB) lines. The calibrated energy range is 40~120KeV and its energy resolution is about 10KeV. Time resolved measurements have been obtained and compared with the E//B Neutral Particle Analyzer (NPA) results. It is observed that particle fluxes increase strongly and then decay rapidly to a steady level just after NB injection commences. Though this temporal behavior is also observed in the E//B NPA, it is not predicted in TRANSP simulations. In addition, the increase and decay rates in the two NPA systems are different. Example data from plasma discharges will be presented with explanations of these differences.

[PP1.010] Study of Uncertainties in NSTX Thomson Scattering Data

Having run reliably for several campaigns with signals in the range of 1e4 to 1e5 photoelectrons per spatial channel per laser pulse, the NSTX Thomson scattering diagnostic has provided a valuable database with which to study the significance of uncertainties leading to determinations of dTe/Te and dne/ne at the few percent level. As this system is about to be expanded from 20 to 30 spatial channels and will require a complete recalibration in several months, we intend to use this study to improve our calibration and operating procedures to optimize data quality. The database consists of \sim 2e5 laser pulses, each with 240 detected signals. In addition to our proposed changes, we present a summary of our analysis technique, a listing of the relative contributions from all known sources of error, and an assessment of our treatment of uncertainties based on chi-squared and residual distributions.

[PP1.011] High \beta electron micro-stability in Spherical Tokamaks

High \beta values achieved in spherical tokamaks, make these devices an attractive route to fusion energy. The attainable \beta depends on the anomalous transport, and so understanding the micro-instabilities responsible for transport is important.

Electron scale micro-instabilities are investigated using the gyrokinetic code GS2 [1]. Variations of \beta ^\prime=-\beta/L_p and magnetic shear \hats are treated by a local equilibrium expansion [2] around MAST flux surfaces [3]. In this way, stability is studied in regimes of steep pressure gradient and high local \beta.The effects of \beta^\prime and \beta on the ETG mode are presented.

At high \beta a micro-tearing mode is destabilised. GS2 simulations of a conceptual spherical tokamak power plant [4], show a regime in which this is the only unstable mode at \rho_e scales.

Nonlinear simulations of ETG turbulence show significant transport for MAST equilibrium parameters. The effect of \beta on ETG secondary stability is investigated.

High normalised transport is produced in nonlinear simulations of the micro-tearing mode, which show the evolution of the magnetic field leading to stochasticity.

[1]M.Kotschenreuther et al. Comp.Phys.Comm. \textbf88(1995)[2] J.M.Greene and M.S.Chance, Nucl.Fusion \textbf4(1981)[3] D.Applegate et al.\ submitted to Phys.Plas[4] H.R.Wilson et al.\ to be published in Nucl.Fusion

[PP1.012] Overview of Results from MAST

We present results from the diverse MAST experimental programme. Features include transport analysis of discharges with edge or internal transport barriers, and high confinement discharges, with and without ITBs, under counter-current NBI. H mode studies are extended to include confinement measurements for submission to the ITER database, as well as controlled experiments on the L/H transition to cast light on the transition process. An important aspect of exhaust studies on MAST is the measurement and control of transient and steady-state power loads, and ELMs are analysed in terms of both their dynamic structure and their impact on the first wall. During 2003-4 MAST has undergone significant enhancement including new divertors to handle increased NBI power and pulse length, error field correction coils to extend the operating space towards lower density and q, a new CXRS system and a new HFS gas injection system. Early results exploiting these new facilities are also presented.

[PP1.013] The role of magnetic equilibria in determining ECE in MAST

We consider two models for MAST magnetic equilibria, EFIT and SCENE, and the predictions that arise concerning ECE. Extensive ECE data from 16 to 60 GHz are available in MAST. The characteristic low magnetic field and high plasma density of a spherical tokamak do not permit the typical radiation of O and X modes from the first five electron cyclotron harmonics. Thus only electron Bernstein modes, (modes not subject to a density limit), which mode convert to electromagnetic waves in the upper hybrid resonance region, can be responsible for the measured radiation. A Gaussian beam formalism is used taking into account the MAST window/mirrors. Our model to determine the intensity of ECE to be detected by the antenna uses the magnetic equilibria as determined by EFIT or SCENE. The mode conversion (in particular EBW-X-O) efficiency is determined from a plane stratified plasma slab using adaptive finite element solution of the cold plasma equations around the UHR. 3D ray is solved simultaneously with the radiative transfer equation to describe the EBW propagation. Reabsorption of the radiation, which is important for non-local wave damping, is thus taken in account. We compare ECE signal and model computations in both L-modes and H-modes and the physics of the EFIT and SCENE codes.

[PP1.014] ECW/EBW heating and current drive experiments on the TST-2 spherical tokamak

Three types of ECW/EBW experiments (200 kW at 8.2 GHz) were performed on TST-2: (1) Electron heating of ohmically formed plasmas using the X-B mode-conversion scenario. A 10-15% increase of the stored energy, a doubling of soft X-ray emission, a centrally localized increase of photon emission, and a small increase of plasma current indicated power absorption in the plasma core. However, the absorption efficiency was only about 10% of the injected power. (2) Up to 4 kA of plasma current was formed and maintained for 0.28 s by RF power alone with constant toroidal and vertical fields. The time averaged electron temperature was 160 eV, and the plasma current centroid was located on the outboard side. (3) With 100 kW of EC preionization, up to 10 kA of plasma current was formed by induction from outer PF coils only. The center solenoid was not used.

[PP1.015] Stability Studies at High Field Utilization in the \sc Pegasus Toroidal Experiment

The \sc Pegasus Toroidal Experiment is exploring current and pressure limits in the high \beta_t, low-q operating space at near-unity aspect ratio. The first limit of interest is the external kink boundary that will define the accessible low-q, high I_N space. Initial operations were characterized by high \beta_t at very low toroidal field (B_t \leq 0.07 T) but were limited both by 2/1 tearing modes in the resistive, low-shear interior and by power supply waveform control capability. The experiment has been modified to avoid these limits with increased, time-variable TF, increased V-s, and improved position and shape control. The modifications allow for greater flexibility in q(r,t) tailoring and should provide access to the external kink boundary. Equilibrium and stability (DCON) modeling projects stable equilibria approaching I_p/I_tf \sim 3 (I_N \sim 20). The initial campaign with the upgraded facility is focused on first suppressing the internal MHD activity and then challenging kink limits by achieving the modeled parameters.

[PP1.016] Soft X-Ray and Radiated Power Measurements on \sc Pegasus

Understanding power losses and confinement is important to achieving high \beta_t plasmas at low aspect ratios in the \sc Pegasus ST and to contributing to the spherical torus database. To that end, new diagnostic systems are being deployed on the experiment. Two tangentially viewing, 16-channel XUV diode arrays will provide radiation power loss rate profiles in the plasma midplane. This profile is combined with a fitted plasma equilibrium to determine the total radiation losses. A novel silicon-drift-detector-based pulse height analyzer (PHA) will determine the electron temperature by continuum measurements from the 1-3 keV region. Direct digitization and fitting of individual photon events minimize pileup effects and dead-time losses. Count rates up to 500 kcounts/sec will allow T_e(t) measurements with a few-ms time resolution. A T_e profile can be obtained on a shot-to-shot basis. Finally, the conceptual design of a second generation very-high sensitivity SXR 2-D tangential imaging system is under development for determination of J(r).

[PP1.017] Impurity Spectroscopy and Wall Conditioning on \sc Pegasus

Impurity-induced radiation has significant influence on the V-s limited \sc Pegasus experiment, and efforts are underway to control and measure the impurities therein. Titanium gettering, helium glow discharge and cryogenic pumping are used for wall conditioning. Initial low power operations are characterized by high deuterium pump speeds up to 35000 l/s and low gas recycling rates. Impurity species mix is estimated from time-resolved VUV spectra using a SPRED multichannel spectrometer. A fast-scanning 1-D CCD detector array provides a full 10-110 nm spectrum every 0.2 ms. Quantitative estimates of Z_eff and the shape of the n_e(R,t) profile will be pursued with a tangentially viewing visible bremsstrahlung (VB) array. This system uses a slow-scan CCD camera to provide time-resolved VB spatial profiles using on-chip line transfer of the exposed spectrum to a masked region during the shot. Oxygen is expected to be the dominant impurity, and XUV diodes with multilayer Ross filters are used to isolate the He-like and H-like oxygen line emissions for central impurity monitors.

[PP1.018] Control, Data Acquisition and Analysis Systems on the \sc Pegasus ST

Recent facility upgrades on the \sc Pegasus ST required the development of flexible control and data systems. Diagnostic and event flow control is set by a master system, while most direct interfaces to hardware are enacted by an array of semi-autonomous slave systems. One subsystem locally controls the high-energy capacitor bank charging and safety functions, while another provides reference waveforms for control of the pulse-width modulators for the new multi-channel switching power supplies. Several satellite systems perform functions such as CAMAC data acquisition and local diagnostic control. Most communications use a simple TCP protocol and a few LabVIEW-based libraries. In addition, independent high-speed acquisition systems allow detection of very fast (\sim 10 ns) transients in the power supplies. Other satellite systems have been implemented for discrete data analysis tasks. For magnetic equilibrium calculations, a Grad-Shafranov solver with a Levenberg-Marquardt fitting algorithm offers ready expansion to include arbitrary diagnostic information.

[PP1.019] Facility and Programmable Power Supply Development for the \sc Pegasus ST

A rebuild of the \sc Pegasus ST facility with several major upgrades is in its final stages, and low-power ohmic operation has begun. A new low-inductance, fast-response toroidal field system allows variation of B_t within a few ms. Additional poloidal field coils, including divertor coils, provide shaping control. New modular IGBT and IGCT switching power supplies have been developed to provide full programmability control of all the magnet coils. The toroidal and poloidal systems are powered by several 900 V/16 KA IGBT switch assemblies, while the OH system requires a 2700 V/48 KA IGCT system. An array of IGBT switch modules has been successfully deployed for low-power OH operation --- albeit with full power for poloidal and toroidal fields --- while initial deployment of the IGCT system is expected this Fall. Together, these systems provide increased V-sec, programmability for V_loop and flexible position and shape control, and increased toroidal field strength.

[PP1.020] Fueling and Plasma Initiation Tests with a Plasma Gun on the \sc Pegasus Toroidal Experiment

Developing non-inductive startup and efficient fueling techniques is important for the ultralow-A \sc Pegasus ST experiment, and the ST concept in general. A single low impurity, high current (\sim 1 KA) plasma gun( G. Fiksel et al.), Plasma Sources Sci. Technol. 5, 78 (1996). has been installed in the divertor region of \sc Pegasus to test auxiliary plasma injection and toroidal current drive during plasma startup. This is effectively a form of DC helicity injection. Direct plasma fueling reduces ionization losses during startup, and for a \sc Pegasus windup factor of ~15-25 the gun can provide a significant target I_p (\sim 10 KA) at the beginning of an ohmic discharge. This significantly eases plasma startup, allows access to a much wider range of plasma currents and shapes than present volt-second limitations allow, and provides critical tests for a future multiple-gun array. Experimental results from the single gun tests are reported.

[PP1.021] Potential Electron Bernstein Wave Heating Experiments on \sc Pegasus

The electron Bernstein wave (EBW) is being studied for use in a wide variety of high-beta confinement devices where the plasma is overdense. The EBW is of particular interest in spherical torus (ST) experiments, where it could be used both as an electron temperature diagnostic and as a technique for heating and current drive. The \sc Pegasus Toroidal Experiment provides an attractive opportunity for investigating the physics and implementation of EBW heating and for developing scenarios for larger experiments such as NSTX. It operates at low toroidal field (B_t < 0.15 T), allowing the utilization of abundant, low-cost 2.45 GHz hardware and sources. Recent upgrades to the experiment provide for programmable position control which is essential for good coupling of the fast X-mode to the EBW. Planning has begun for a 1 MW heating and current drive experiment on \sc Pegasus. Raytracing calculations, antenna designs, diagnostic requirements and experimental possibilities are presented.

[PP1.022] The HIT-II Spherical Torus: Physics and Key Experimental Results

Discharges in the HIT-II spherical torus device [Redd et al., Phys. Plasmas 9, 2006 (2002)] can be driven by either Ohmic or Coaxial Helicity Injection (CHI) current drive. A new CHI operating regime has been explored, with toroidal plasma currents of up to 350 kA, I_p/I_TF ratios of up to 1.2, and internal probing data which may demonstrate the formation of a closed-flux core. The key to acheiving these results is the magnetic field shear in the CHI injector region, with a minimum shear necessary for current build-up. Ohmic plasma performance has also improved, with peak currents up to 300 kA, with and without transient CHI startup. The CHI startup technique [Raman et al., Phys. Plasmas 11, 2565 (2004)] provides more robust discharges, with a wider operating space and more efficient use of the transformer Volt-seconds, than unassisted Ohmic. Finally, CHI can be used to enhance an Ohmic plasma current without significantly degrading the quality of the discharge. Results will be presented for each HIT--II operating regime, including empirical performance scalings and applicable parametric operating spaces.

[PP1.023] Internal Magnetic Probing of HIT-II CHI Plasmas

The HIT--II device, a low aspect ratio (R_o = 0.3m, a = 0.2m) torus with B_T = 0.5T on axis, operates with both inductive drive and Coaxial Helicity Injection (CHI) to initiate and sustain the plasma equilibrium. Recently, a CHI operational regime has been found that shows a significant departure from previous plasma behaviour and may be attributed to Taylor relaxation to a more Spheromak-like equilibrium. These plasmas have produced record plasma currents of 350kA. An internal magnetic probe consisting of an array of eight 3d coils spanning 8.8cm(3.5") has successfully probed these plasmas to a depth of 15cm(6"). Internal field measurements have shown the relaxed plasmas have significantly higher poloidal flux than the bias injector flux and that the plasmas are highly paramagnetic. Only the highest current discharges and 6" probing depths have shown significant probe perturbations (20% degradation in I_p) and even these shots demonstrate the typical dynamics of the relaxed plasmas. The internal field structure and dynamics of the two CHI operational modes are described and compared. Also, the transient bubble burst initiation of the CHI plasmas and the ubiquitous n=1 rotating mode is well characterized by the probe and will be described.

[PP1.024] Far Infrared Interferometry and Ion Doppler Spectroscopy on the HIT Program

Ion Doppler Spectroscopy (IDS) is used to measure impurity ion velocity and temperature on HIT--II and HIT--SI. The spectrometer is scannable through 10 impact parameters on a shot to shot basis. Oxygen V temperatures for CHI discharges in HIT--II routinely exceed 250 eV during current drive, exceeding measured electron temperatures by a factor of >2. Ohmic discharges exhibit lower ion temperatures which are in better agreement with measured electron temperatures. This suggests the presence of relaxation activity during CHI current drive which is preferentially heating the ions. A tangentially viewing Far Infrared (FIR) interferometer is used to measure density on the HIT experiments. The FIR source is an optically pumped difluoromethane gas laser which operates at 185 um. A heterodyne signal is generated by doppler shifting part of the beam off of a rotating diffraction grating. The interferometer is in a Martin Puplett configuration, and is scanable through 6 impact parameters on a shot to shot basis. The instrument is sensitive to chord integrated density fluctuations less than 10^18 ~m^2, and has time resolution of up to 200 kHz. Density fluctuations associated with the CHI n=1 mode, and ohmic drive Internal Reconnection Events (IREs) will be presented.

[PP1.025] CHI during an ohmic discharge in HIT-II

Coaxial Helicity Injection (CHI) has been used on the National Spherical Torus Experiment (NSTX), the Helicity Injected Torus (HIT) and HIT-II to initiate plasma and to drive up to 400 kA of toroidal current. The primary goal of the CHI systems is to provide a start-up plasma with substantial toroidal current that can be heated and sustained with other methods. We have investigated the use of CHI systems to add current to an established, inductively driven plasma. This may be an attractive method to add edge current that may modify the stability characteristics of the discharge or modify the particle and energy transport in a spherical torus. For example, divertor biasing experiments have been successful in modifying particle and energy transport in the scrape-off layer of tokamaks. Use of IGBT power supplies to modulate the injector current makes analysis of current penetration feasible by comparisons of before and after CHI using EFIT analysis of the data.

[PP1.026] Pressure-driven instabilities of high-beta ST in TS-3 reconnection heating experiment

The high-beta STs (toroidal beta=0.2-0.7) have been generated by two merging STs in TS-3 reconnection heating experiment. The rapid heating of reconnection made some of the high-beta STs unstable to ballooning mode. To study the properties of this ballooning mode, we measured magnetic fluctuation of high-beta STs near the separatrix using a set of pick-up coils with spatial resolution as fine as 15mm. The fluctuation amplitude decreased with the external toroidal magnetic field B_ex, as often observed in conventional tokamaks. However, when the toroidal magnetic field exceeded a certain critical value, the fluctuation amplitude increased significantly. The plasma pressure gradient was found to increase with B_ex due to its improved confinement effect. The study of s(magnetic shear)-\alpha(pressure gradient) indicates that the onset of the fluctuation is caused by the ballooning related instability in the increased pressure gradient regime.

[PP1.027] Investigations of Low and Moderate Harmonic Fast Wave Physics on CDX-U

Third harmonic hydrogen cyclotron fast wave heating studies have begun on CDX-U to investigate the potential for bulk ion heating. In preparation for these studies, the available RF power in CDX-U has been increased to 0.5 MW. The operating frequency of the CDX-U RF transmitter was lowered to operate in the range of 8-10 MHz, providing access to the ion harmonic range 2Ømega\sim4Ømega in hydrogen. A similar regime is accessible for the 30 MHz RF system on NSTX, at 0.6 Tesla in hydrogen. Preliminary computational studies over the plasma regimes of interest for NSTX and CDX-U indicate the possibility of strong localized absorption on bulk ion species. Numerical simulation studies for relavent ST geometries will be presented, along with result for the CDX-U antenna experiments.

[PP1.028] Design and Testing of a Supersonic Gas Injector on CDX-U.

Recent experiments on the CDX-U spherical torus have successfully achieved a significant reduction in recycling with large-area liquid lithium plasma-facing surfaces. Such wall conditions have also demonstrated the need to improve plasma fueling. To address this challenge, a supersonic gas injector, based on a Mach 8 Laval nozzle design,[1] has been constructed, characterized, and tested for installation on CDX-U and on the NSTX spherical torus. Supersonic gas injectors show promise of being a much more efficient fueling method than gas puffing, while remaining much easier to implement than pellet injection. Fueling efficiency of the nozzle will be examined, as well as its affect on particle confinement.

[1] M. Baumgartner, Ph. D. thesis, Princeton University (1997)

[PP1.029] Recent Liquid Lithium Limiter Experiments in CDX-U

Studies with a large area liquid lithium plasma-facing surface in the CDX-U spherical torus have demonstrated its ability to lower recycling dramatically and reduce impurities. Observed improvements to plasma performance include a lowering of the loop voltage consumption by at least a factor of four, and a more than doubling of the central ion temperature. These results are consistent with simulations that suggest a decrease in internal inductance and effective charge. Recent experiments have also shown the need for careful temperature control to keep the lithium from migrating over the edge of the limiter tray as it �wets� its stainless steel surface. The tray heaters have been modified to minimize this effect, and measurements with improved plasma diagnostics in liquid lithium limiter experiments under these new conditions will be reported.

[PP1.030] Effects of a Liquid Lithium Limiter on Edge Plasmas in CDX-U

The fully-toroidial liquid lithium limiter tray in CDX-U is intended to reduce recycling and impurities in the discharge. We have observed a 3-4x reduction in loop voltage during liquid lithium operation, which suggests broadened current and temperature profiles. Measurements of the edge temperature have now been performed with a triple Langmuir probe to determine if there is an increase in the edge electron temperature. The primary measurement of recycling is spectroscopic H-alpha light from the edge plasma. These spectroscopic measurements use two beams with a common light dump; one is focused directly on the beam dump to measure background light, while the other is first reflected off of the lithium surface of the limiter tray. Experimental results will be presented.

[PP1.031] The Lithium Tokamak eXperiment

A fully nonrecycling first wall has been predicted to fundamentally alter the nature of a tokamak plasma. Recent experimental data indicate that a surface of liquid lithium can provide greatly reduced recycling. The Lithium Tokamak eXperiment (LTX) is designed to eliminate recycling with a full molten lithium wall. Electron temperature and current profiles, transport and stability properties which are qualitatively different from a conventional high recycling tokamak are expected to result. If successful, LTX will be a critical step in the development of the lithium wall tokamak, which may well provide the shortest, lowest cost, and most environmentally attractive path to the implementation of fusion energy. The design and progress in the construction of LTX will be summarized, and the results of preliminary tests of the lithium coated wall concept developed for LTX will be presented.

[PP1.032] Stellarators

[PP1.033] Progress in NCSX Engineering

The engineering development of the National Compact Stellarator Experiment is proceeding on schedule toward First Plasma in May, 2008. The design has matured and the fabrication phase is beginning. Recent changes in design details and project plans will benefit the physics program. The vacuum vessel has been designed to be heated to 350C to facilitate the implementation of carbon first wall components. The ports were optimized in orientation, shape, size, and number (a \sim50% increase) for diagnostic access. Insulating breaks in the coil structure were added to inhibit eddy currents that could interfere with plasma control. Testing activities were added to the plans to ensure the reliability and accuracy of the coil system prior to operation.

[PP1.034] Finite Beta Boundary Magnetic Fields of NCSX

The magnetic field between the plasma surface and wall of the National Compact Stellarator (NCSX), which uses quasi-symmetry to combine the best features of the tokamak and stellarator in a configuration of low aspect ratio is mapped via field line tracing in a range of finite beta in which part of the rotational transform is generated by the bootstrap current. We adopt the methodology developed for W7-X, in which an equilibrium solution is computed by an inverse equilibrium solver based on an energy minimizing variational moments code, VMEC2000[1], which solves directly for the shape of the flux surfaces given the external coils and their currents as well as a bootstrap current provided by a separate transport calculation. The VMEC solution and the Biot-Savart vacuum fields are coupled to the magnetic field solver for finite-beta equilibrium (MFBE2001)[2] code to determine the magnetic field on a 3D grid over a computational domain. It is found that the edge plasma is more stellarator-like, with a complex 3D structure, and less like the ordered 2D symmetric structure of a tokamak. The field lines make a transition from ergodically covering a surface to ergodically covering a volume, as the distance from the last closed magnetic surface is increased. The results are compared with the PIES[3] calculations.

[1] S.P. Hirshman et al. Comput. Phys. Commun. 43 (1986) 143.

[2] E. Strumberger, et al. Nucl. Fusion 42 (2002) 827.

[3] A.H. Reiman and H.S. Greenside, Comput. Phys. Commun. 43, 157 (1986).

[PP1.035] Overview of the QPS Experiment

The Quasi-Poloidal Stellarator (QPS) is a very-low-aspect-ratio (R/a = 2.7) compact stellarator in which the dominant magnetic field components are poloidally symmetric in flux coordinates, which leads to large reductions in: neoclassical transport at low collisionality; bootstrap current; and poloidal viscosity, which allows large E x B poloidal flows for suppression of anomalous transport. The magnetic configuration is relatively insensitive to increasing beta. The experiment under design has R = 0.95 m, a = 0.35 m, B = 1 T for a 1.5-s pulse, and P(heating) = 2�4 MW. Nine independent coil currents allow varying: neoclassical transport by a factor of 12-36, degree of poloidal symmetry by a factor of 9, and poloidal viscosity by a factor of 6�30. QPS can study regimes in which either the anomalous transport or neoclassical transport is dominant, stability limits at beta up to 5and equilibrium robustness at finite beta. Recent progress on physics issues, the relationship to other stellarator concepts, the QPS project status, and the proposed experimental program are presented.

[PP1.036] Design of the QPS Experiment

The Quasi-Poloidal Stellarator, QPS, is a low-aspect-ratio (R/a = 2.7), compact stellarator under design at ORNL. The device parameters are R = 0.95 m, a = 0.35 m, and B = 1 T for 1.5 s with 2 MW of ECH and 3.5 MW of ICRF for plasma heating. A nonplanar modular coil set provides the primary magnetic field configuration. It has two field periods with ten modular coils per period. Due to stellarator symmetry, there are only five different coil types. Flexible copper cable conductor will be wound on a stainless steel form, vacuum impregnated with epoxy, and canned for vacuum compatibility. The coil form allows the coils to be connected into an integral structural shell. Unlike most stellarators, a vacuum vessel surrounds the coils rather than fitting inside the coils, allowing excellent access for plasma diagnostics and heating. In addition there are external vertical field and toroidal field coils and an ohmic current solenoid for configuration flexibility. First plasma operation is planned for early 2009. Details of the engineering design and analysis will be presented.

[PP1.037] Control of Magnetic Islands and Deviation from Quasi-Symmetry in Compact Stellarators

Techniques have been developed to minimize magnetic islands due to field perturbations and to determine deviations from ideal symmetry in compact stellarators. The measured field on a dense 3-D grid around a non-planar modular coil is compared with the design field from an ideal coil. The 3-D path representing the current center of the coil pack is found by minimizing the difference between the measured and design fields. The spatial orientation of each coil is determined to give the best fit to the design field. Residual magnetic islands are minimized by calculating the current for each modular coil by following magnetic field lines and locating periodic trajectories in a symmetry plane. The residues, a measure of island size, are minimized in a nonlinear optimization. The departure from ideal poloidal symmetry for QPS is found by following electron beam orbits. The deviation from the flux surface is a direct measure of the non-poloidally symmetric components of the magnetic field.

[PP1.038] Initial Operation of the Compact Toroidal Hybrid (CTH)

CTH is a stellarator/tokamak hybrid device that will use an ohmic current of I_p\leq50 kA to investigate both ideal and resistive current-driven instabilities in low aspect ratio stellarators and to test 3-D equilibrium reconstruction using the V3FIT code under development.\footnote[2] S. Hirshman, E. Lazarus, J. Hanson, S. Knowlton, and L. Lao, Phys. Plasmas v.11, p.595 (2004) CTH will have an edge vacuum rotational transform variable from t_V(a) = 0.2 to 0.5 with a current-generated transform of t_J(a)\leq0.5 at the planned magnetic field B_o\leq0.5T. The average major radius is R_o = 0.75 m, the vessel minor radius is a_v = 0.29 m and the maximum average plasma minor radius is a_p = 0.20 m. Following completion of the assembly of CTH, vacuum surface mapping will be performed at low magnetic field to determine the accuracy of the coil winding and placement, and the extent to which the flexibility of the magnetic design was realized. The effect of error correction coils on suppression of static magnetic islands will be assessed. Initial plasmas will be generated by 2^nd harmonic ECH at 18GHz.

[PP1.039] Equilibrium Reconstruction in Stellarators: V3FIT

Equilibrium reconstruction is a crucial capability in the interpretation of tokamak experiments. As stellarator plasma beta or bootstrap currents increase, the configuration of flux surfaces deviates further from that of the vacuum, and reconstruction becomes necessary to determine the equilibrium state. V3FIT is the stellarator equilibrium reconstruction code that we are writing. It is designed to be 1) fast, 2) flexible, and 3) easy to modify. To make it fast, we are closely coupling the equilibrium reconstruction iterations with the equilibrium convergence iterations. To make it flexible, the code is designed to be able to work with different equilibrium codes, and with multiple types of diagnostics. Initially, we will use VMEC as the equilibrium code, and magnetic diagnostics as the primary type of experimental information about the equilibrium. To make V3FIT easy to modify, we are writing the code in Fortran 95, and making extensive use of modules and derived data types.

[PP1.040] Overview of HSX Experimental Program

The quasisymmetric stellarator approach to confinement continues to be explored in the Helically Symmetric Experiment. We have demonstrated improved single particle confinement through experiments showing faster plasma breakdown, increased microwave absorption, higher x-ray flux and decreased trapped particle losses. Recently we showed that parallel viscous damping is reduced with quasisymmetry. Presently our program is concentrating on exploring whether the huge reduction of neoclassical transport with quasisymmetry leads to an observable difference in the plasma parameters during 2nd harmonic X-mode ECH at 0.5T. We have begun to investigate what role turbulence plays in limiting the confinement improvement with quasisymmetry. One unforeseen feature of quasisymmetry is perhaps related to the improved confinement of trapped particles; we have seen evidence of a possible Global Alfven Eigenmode that disappears when the symmetry is broken. Future plans and upgrades will be discussed.

[PP1.041] ECRH and ECE on HSX stellarator

Plasma heating on HSX stellarator is made by the extraordinary wave at the second harmonic of the electron cyclotron frequency. Absorption of launched power is measured with 6 microwave antennas. Three receiving antennas are mounted in vicinity of ECRH launcher and three others are distributed along the machine. The results of these measurements show a high multi-pass absorption (>0.8) in both quasi-symmetric and 10% mirror configurations. The measurements are supported by intensive ray tracing calculations in Maxwellian and bi-Maxwellian plasmas. The electron cyclotron emission spectrum is measured by 8-channel radiometer. At high plasma density (>1.7x10^18 m^-3) the emission is thermal while at low plasma density the emission is mostly due to supra-thermal electron. The interpretation of ECE spectra is done with modified 3-D ray tracing code. Autocorrelation function technique is used to find out the level of electron temperature fluctuations in HSX plasma. Future upgrades on ECRH and ECE systems will be discussed. Work Supported by US DOE under grant DE-FG02-93ER54222

[PP1.042] Profile Measurement of HSX Plasma Using Thomson Scattering

At the HSX plasma laboratory, a 10 channel Thomson scattering system is now operational. The system has been absolutely calibrated for density measurement using Raman scattering with nitrogen gas. It is found that for the QHS configuration the electron temperature gradually decreases when we increase the density at a fixed ECRH power and that the whole temperature profile increases with the heating power at a fixed plasma density. The central temperature increases from about 500 eV to 950 eV while the launched heating power increases from 37 kW to 150 kW for the QHS configuration plasma with a density of 1.5\times 10^12cm^-3. At the present density and heating power, the difference between the QHS and Mirror mode is not pronounced. Detailed results will be presented at the conference.

[PP1.043] Neoclassical Transport in HSX

The magnetic field of HSX can be varied from a quasihelically symmetric (QHS) configuration with very low neoclassical transport, to a field with no symmetry, called the Mirror configuration, which has neoclassical transport similar to a conventional stellarator. Particle transport is measured with a suite of absolutely calibrated H� detectors, coupled with 3D calculations of the particle source. This yields the experimental particle flux, which is compared to neoclassical expectations. It is found that in the core (r/a < 0.3) of Mirror plasmas, the neoclassical contribution to the particle flux dominates, and that neoclassical transport is negligible in QHS configuration. The experimental heat diffusivity calculated from the plasma profiles shows similar trends; it is close to neoclassical in the core of Mirror plasmas, and anomalous otherwise.

[PP1.044] Evidence for Alfv�nic Fluctuations in Quasi-Helically Symmetric HSX Plasmas

Interferometer, Langmuir probe and magnetic probe time-series traces reveal the existence of high-frequency fluctuations in the range of 20-120 kHz. These fluctuations are observed for ECRH produced plasmas in the quasi-helically symmetric HSX device and have an m=1 structure. The fluctuations are coherent and global, peaking in the plasma core. For low-density collisionless plasmas, satellite modes are also observed. The Alfv�n continua for the n = 1 mode family was calculated and shows there exists an n=1, m=1 global Alfv�n eigenmode (GAE) for typical plasma parameters. Fast electrons associated with 2nd harmonic X-mode ECRH are thought to drive the instability. The measured frequency range as well as scaling with electron density and ion mass are consistent with GAE predictions. When quasi-helical symmetry is broken, fast electron confinement deteriorates and the mode is no longer observed.

[PP1.045] Magnetic fluctuations measurements in the low beta HSX Stellarator

Even though the Helically Symmetric eXperiment (HSX) presently operates at low beta of less than 0.1 fluctuations of about 0.5 less than 1.0x10^12 cm^-3. There are two types of magnetic fluctuations observed. The first type is a coherent mode with a large amplitude and growth rate of 10�s of mseconds. This coherent mode shows some characteristics of a GAE mode. The amplitude of this mode depends on plasma density and the heating location and is well correlated with density fluctuations. This mode has no apparent effect on any plasma parameters. A second mode, bursty in nature, has a large amplitude and has a large impact on the plasma. We observe large decreases in the stored energy and ECE signal at the onset of this mode. This mode is observed only when heating is near the magnetic axis. We are planning to measure the m and n spectrum of these modes. The initial results of these magnetic studies will be presented.

[PP1.046] Edge Turbulence in Various Magnetic Configurations of HSX

Multi-pin Langmuir probes have been used to measure the edge and SOL characteristics of HSX plasmas in multiple locations and under various magnetic configurations. Auxiliary coils provide the flexibility to change the vacuum magnetic spectrum, well depth, rotational transform, and effective minor radius. The ion saturation current and floating potential probes measure broadband, large level fluctuations (10 - 40%) in the edge and SOL. Poloidal wavenumbers, measured via two displaced probes, are in the range of 0.5 - 1.5 cm^-1, with \rho _s k_\theta = 0.1 - 0.2. From the estimated density gradient scale lengths in the edge (L_n = 2 - 5 cm), these fluctuation levels are consistent with mixing length type arguments (n`/n \sim 1/k_\theta L_n). The phase velocities of the fluctuations follow the E x B velocities inferred from floating potential profiles. In the quasi-helically symmetric configuration, the direction of measured electrostatic turbulent flux is inward at line-averaged densities below 1.7 x 10^12 cm^-3, and outward above. Probe measurements inside the separatrix indicate decreased fluctuation levels with increased well depth.

[PP1.047] Plasma Turbulence in the HSX Stellarator

Multi-point measurements of plasma density and potential and correlation analysis are used to find the structure of particle transport events (``blobs'') in the plasma edge.The quasi-symmetry of the confining magnetic field can be gradually broken by auxiliary fields, changing both the neoclassical and the turbulent transport properties. The impact on the turbulent part of the transport is investigated and compared to global plasma confinement properties (energy confinement time.) The wave number spectra and density-potential cross phases are used to discover the underlying instability (i.e. drift wave vs. interchange.) In addition, the parallel correlation along magnetic field lines is investigated both with passive and active techniques.

[PP1.048] Calculations of Neoclassical Viscous Damping Due to Magnetic Islands in the HSX Stellarator

Neoclassical viscous damping rates have been calculated for the HSX stellarator, concentrating on the effects of magnetic surface shape changes due to magnetic islands. Results indicate that small islands present in the quasi-helically symmetric configuration can give rise to an observable but not problematic deviation from quasi-symmetry. These islands can be eliminated via a small reduction in the rotational transform. Large islands can be produced by the introduction of the iota-bar equals 4/4 resonance inside the last closed magnetic surface. These islands result in not only a significant reduction in the volume of closed nested magnetic surfaces, but also an elevated level of viscous damping on magnetic surfaces adjacent to the islands. This research is funded by the United States Department of Energy.

[PP1.049] Profile Characteristics of Edge Transport Barrier on CHS

The edge transport barrier (ETB) has been observed for the two co-injected neutral beam heated helical plasma in CHS. When the heating power exceeds the power threshold of P \sim 800kW, H_\alpha signal showed a clear spontaneous drop followed by the increase of line-averaged density after the second neutral beam injection. The stored plasma energy with the diamagnetic measurement also increases about 20 ms after the transition. The optimization for the magnetic configuration using quadrupole coils is effective for the ETB formation. The increase of the stored energy by the optimization achieves to \sim 40%,and H-factor (ISS04v03) of \sim 1.3.

The profile characteristics of the electron density and temperature of the ETB formation has been investigated with a multi channel YAG Thomson scattering system. A considerable increase of the edge density by \sim 50-100% in \rho>0.5 has been observed, while the electron temperature in \rho>0.7 increases by 30-50%. These results show the increase of the stored energy is mainly caused by the improvement of a particle transport in the edge region.

[PP1.050] Collisional ripple transport of neutral beam-injected energetic ions in low aspect ratio helical system CHS

Losses of partially thermalized, pitch angle-scattered NB ions have been experimentally observed by use of a lost fast ion probe (LIP) at the small R side of CHS. The LIP indicates that the energy of escaping fast ions at the probe position ranges from 10 to 20 keV and their pitch angle is around 75~80 degrees although the NB(Eb=38keV) is tangentially coinjected in CHS. Judging from this, collisional ripple transport looks important to understand beam ion behavior and confinement in CHS. Particle simulation in the presence of collisions is performed by the DELTA5D code to verify whether the experimental observation can be explained by classical drift orbit phenomena. The DELTA5D suggests that escaping fast ions can reach the probe position via collisional ripple transport and the their energy is consistent with that observed by the LIP. Comparison will be made between the results from experiment and simulation for ripple transport and resulting losses of beam ions in CHS.

[PP1.051] A new technique to optimize coil winding path for the arbitrarily distributed magnetic field and application to CHS-qa modular coils

A new technique to calculate an arbitrarily shaped coil winding path for a target magnetic field distribution has been developed. The technique is called DUCAS (Design tool Using Current potentials And SVD, SVD= Singular Value Decomposition). The coil winding surface (CWS) is modeled by triangular finite elements (FEs). The SVD is applied on the response matrix from the current potentials (CPs) of the FE nodes to the magnetic field, to get eigen distribution functions of CPs and singular values (SVs). Using the eigen functions with large SVs, the CP distribution is determined on the CWS so as to reproduce a given magnetic field distribution. Discrete coil shapes are determined along the contour (flow) lines of CPs. The arbitrarily formed CWS is acceptable in DUCAS. We applied the DUCAS on CHS-qa (quasi-axisymmetric Compact Helical System) modular coils and confirmed that the technique is applicable on designs of helical system modular coils.

[PP1.052] Impact of field line label and ballooning parameter on infinite-n ballooning stability in compact stellarators

The infinite-n ideal ballooning mode stability of a non-axisymmetric equilibrium is a function of the ballooning parameter, \zeta_k = k_\iota / k_\alpha and the field line label, \alpha = \theta - \iota\zeta. Here, k_\iota and k_\alpha are the components of the wavenumber perpendicular to the magnetic field and \iota is the rotational transform. In this work, the impact of field line label and ballooning parameter on the infinite-n ballooning stability of compact, quasi-poloidal symmetric stellarators is investigated. Previously, the ballooning stability of quasi-poloidal stellarators has been examined for fixed-boundary, very-high b (b >10tokamak-stellarator hybrid configurations [1] and free-boundary, moderate b (b >4Quasi-Poloidal Stellarator (QPS) [2]. These previous calculations were performed with \zeta_k, \alpha = 0. Here, these results are extended to include other possible values of \zeta_k and \alpha. The first ballooning instability \beta-limits for these devices are well described by the \zeta_k, \alpha = 0 results. Changing either \zeta_k or \alpha increases the \beta required for first instability. The \beta values required to enter second ballooning stability are higher when \zeta_k, a \ne 0. The plasma is still first-unstable to modes with \zeta_k, \alpha \ne 0 even after modes with \zeta_k, \alpha = 0 (and regions nearby in parameter space) become second stable. These results are compared with calculations of the stability of finite-n ballooning modes in both the hybrid configuration and the QPS configuration.

[1] A. S. Ware, S. P. Hirshman, D. A. Spong, et al., Phys. Rev. Lett. 89, 125003 (2002).

[2] A. S. Ware, D. Westerly, E. Barcikowski, et al., Phys. Plasmas 11, 2453 (2004).

[PP1.053] Bootstrap Current compensation with Electron Bernstein Waves in Reactor-Size Stellarators

The Bootstrap Current (BC) arises from the interaction between trapped and passing particles. In high axisymmetry devices the BC contributes to increase the rotational transform (\iota), while in configurations with dominant helical symmetry it counteracts it. It has been shown that this alteration of the \iota profile may lead to instabilities. The aim of this work is to explore the use of Electron Bernstein waves (with no density cut-off) for current drive to locally compensate the BC, to stabilize the configuration and to assess the amount of power required to achieve the goal. Specifically, the BC-consistent equilibrium is calculated iteratively with the VMEC and TERPSICHORE codes. The ART code is used for ray-tracing, mode conversion and power deposition. A module is implemented to calculate the current drive. The method is applied to quasi-axisymmetric and quasi-helically symmetric reactor-size configurations.

[PP1.054] Controlling Chaos in 3D Magnetic FIelds

The ability to find integrable magnetic fields whose field lines lie on tori is absolutely essential for designing fully three-dimensional plasma confinement systems. The theory of controlling chaos in 3D magnetic fields by deforming a given chaotic field into a nearby integrable one is presented and demonstrated computationally. The numerical work converges as the number of modes increases and enjoys extremely fast convergence to a solution when run in a Kolmogorov series-type approach. The 1.5 degree of freedom Hamiltonian case is presented as well, but the full magnetic field problem is much more interesting in that one must treat all components of the vector potential on equal footing. By combining geometry with numerical analysis, this research utilizes computational differential geometry to solve the important problem of designing integrable Stellarator equilibria.

[PP1.055] Edge Plasma Theory and Computation

[PP1.056] On dust particle dynamics in tokamak edge plasmas

The presence of substantial amounts of dust has been observed on the first walls of fusion devices. Although, the impact of dust on plasma parameters in current fusion devices is not clear, dust in burning plasma experiment may cause a significant safety threat. In Ref. 1 was shown that once dust particle comes from the wall into tokamak edge plasma it accelerates by plasma flows and can quickly traverse distances comparable to tokamak radii before being disintegrated due to erosion caused by the interactions with plasma. As a result, the dust deposition areas on the wall structures can be spread far away from the origin of the dust. In this paper we review the results of [1] and consider some new aspects of dust dynamics in tokamak plasmas, which we found from numerical modeling of dust particle motion. We also discuss an impact of dust on core plasma contamination.

[1] S. I. Krasheninnikov, Y. Tomita, R. D. Smirnov, and R. K. Janev, Phys. Plasmas, 11 (2004) 3141

Research was supported in part by the U. S. Department of Energy under Grant No. DE-FG02-04ER54739 at the UCSD

[PP1.057] Blob modeling in the scrape of layer of tokamak with the NIMROD code

Experimental observations and turbulence simulations of the far scrape off layer (SOL) suggest that the transport has non-diffusive nature and the density and particle fluxes are intermitted in this region of tokamak plasmas. Such observations can be explained by the strongly localized filaments of plasma pressure propagating in the radial direction. These filaments were observed in the experiments and are often called ``blobs''. It has been also shown that the blobs are correlated with the edge localized modes (ELMs). Inclusion of non-diffusive blob transport, which can be responsible for large fraction of SOL particle transport, is important for understanding of the edge physics in tokamaks. This report presents the first results of blob modeling with the 3D nonlinear non-ideal MHD NIMROD code [1]. The stability of blobs is discussed. The numerical results are compared with analytical results and with results of 2D MHD simulations (see poster by G. Yu et al. at this meeting). [1] C. R. Sovinec et al. Jour. of Comput. Phys. 195 (2004) 355.

[PP1.058] Two Dimensional Modeling of Blob/ELM Dynamics in Edge Tokamak Plasmas

In many cases strongly intermittent convective rather than diffusive, anomalous cross-field plasma transport plays a key role in both SOL plasma dynamics and plasma-wall interactions in tokamaks, stellarators, and linear devices. Such features are rather typical for ELMs in H-mode while in L-mode and in between of ELM bursts are associated with localized coherent plasma structures (blobs). Simple 2D model based on gravitational drive, which resembles the Boussinesq approximation for thermal convection, allows to explain many essentials of blob dynamics observed experimentally. Here we report on the modeling results with improved 2D model, which goes beyond the Boussinesq approximation. We also adapt our model to be able to simulate the dynamics of coherent structures with high beta, which can strongly bend the magnetic field lines and, thus, avoid their intersections with divertor targets.

Work supported by DOE

[PP1.059] Coherent Structures in the SOL

Long-time simulations of density ``blobs'' in the tokamak scrape-off layer show that most cannot be categorized as ``coherent structures.'' Born near the separatrix, blobs are expected to propagate a distance 5-10 times their own linear dimensions before reaching the chamber walls. Using a simple two-field model commonly used in the literature, we find that small, fast-moving blobs go unstable to Kelvin-Helmholtz modes, as also observed by previous workers. Nonlinearly, these modes lead to repeated vortex-shedding, with its accompanying mass loss, leaving behind only a small fraction of the original blob mass. Large, slow-moving blobs, on the other hand, tend to be unstable to Rayleigh-Taylor modes. The ``RT Fingers'' quickly and violently break up the blob into smaller filaments that continue to propagate radially as they spread the mass poloidally. In this case, only a diffuse cloud of particles makes it to the wall. Blobs of the ``correct size'', however, do behave coherently, retaining most of their original mass. These coherent structures exhibit a steepened leading edge that remains KH-stable, and a long tail. Together, these characteristics agree with the experimental observations for ``intermittent-events'' in the SOL with steep rise and slow decay times. Extension of this work to 3D using a more sophisticated physics-model is being contemplated.

[PP1.060] Convective structures in edge plasmas

Effective plasma gravity caused by magnetic curvature and neutral wind were suggested earlier as mechanisms responsible for nonlinear evolution and radial advection of meso-scale structures as blobs and ELMs. Here we examine the role of driving forces associated with the \nabla T_e instability in the scrape-off-layer and the instability caused by the parallel shear of the \bfE\times B drift velocity. We discuss the linear stage of these instabilities and suggest the nonlinear model which describes two-dimensional convective structures driven by the parallel shear of the \bfE\times B drift velocity and \nabla T_e in the SOL. Estimates of the characteristic size and velocity for these structures are consistent with experimental observations.

[PP1.061] UEDGE simulations of helium plasma discharges on PISCES and NAGDIS-II

In recent years, it was found that divertor plasma simulators like NAGDIS-II and PISCES exhibit many features similar to that found in edge plasmas of large fusion devices like tokamaks. Therefore, taking into account the rather simple geometry, stationary operational conditions, and relatively small scale of experiments, divertor simulators are considered to be a useful test bed for verification of different plasma physics models, in parallel with or even before applying these models to tokamaks. The edge plasma physics code UEDGE has been modified to simulate the hydrogen-helium-impurity mixture plasma in cylindrical geometry. The fast intermittent blobby non-diffusive cross-field transport, which was observed in PISCES and NAGDIS, is modeled in UEDGE as cross-field velocity Vconv directed to the wall. Along with plasma diffusivities, the 2D profile of Vconv is adjusted to match experimental probe measurements. We present simulation results for attached and detached helium plasmas. The synergistic effects caused by different phenomena in linear-machine plasmas such as intermittent non-diffusive transport, fast electrons, ion-electron recombination, MAR, and radiation opacity are discussed. Work supported by DoE grant DE-FG02-04ER54739.

[PP1.062] Density Limit due to SOL Convection

Recent measurements on C-Mod(M. Greenwald, Plasma Phys. Contr. Fusion \bf44), R27 (2002). suggest there is a density limit due to rapid convection in the SOL: this region starts in the far SOL but expands inward to the separatrix as the density approaches the Greenwald limit. This idea is supported by a recent analysis(D. A. Russell et al., Lodestar Report LRC-04-99 (2004).) of a 3D BOUT code turbulence simulation(X. Q. Xu et al., Bull. APS \bf48), 184 (2003), paper KP1-20. with neutral fueling of the X-point region. Our work suggests that rapid outwards convection of plasma by turbulent coherent structures (``blobs'') occurs when the X-point collisionality is sufficiently large. Here, we calculate a density limit due to loss of thermal equilibrium in the edge plasma due to rapid radial convective heat transport. We expect a synergistic effect between blob convection and X-point cooling. The cooling increases the parallel resistivity at the X-point, ``disconnects'' the blobs electrically from the sheaths, and increases their radial velocity,(D.A. D'Ippolito et al., 2004 Sherwood Meeting, paper 1C 43.) which in turn further cools the X-points. Progress on a theoretical model will be reported.

[PP1.063] Implications of blobs for reactors

Blobs rapidly convect plasma outward, increasing sputtering of wall material. Also, return convective flows rapidly carry impurities across the SOL and into the core. Previous calculations(M. Kotschenreuther, T. Rognlien, P. Valanju, FED (accepted).) indicate this might lead to radiative collapse for a reactor. We have developed a model for blobs arising from resistive ballooning turbulence, extending previous work(S. I. Krasheninnikov, Phys. Lett. 283 368-370(2001), and D. A. D'Ippolito, et al. Phys. Plasmas 9 22-233 (2002).) to include temperature dependent coefficients, kinetic neutral transport, and impurities (Be or W). Our model describes resistive ballooning fluctuations, plasma blob transport, plasma background, neutral transport, and impurity transport back to the core due to inward convection of low-density blobs. Impurity density at the last closed flux surface due to wall sputtering is computed. Potentials applied in the divertor might shear stabilize the blobs, reducing impurity influx and improving the density limit. Such biasing is enabled in practice by novel divertors (see adjacent related posters).

[PP1.064] Inducing Alpha Particle Losses for Stable High-\beta Advanced Tokamak Reactors

We describe a scheme where deliberate application of small non-axisymmetric magnetic perturbations induce controlled alpha particle losses to enable stable high \beta plasma operation of ignited advanced tokamaks. The loss of a moderate fraction of alphas (\sim 30-60%) due to \delta B/ B \sim 1-2% can induce plasma rotations with Mach \sim 0.5. This allows rotation and heating profiles to be controlled to enable the stabilization of resistive wall modes, produce and control internal transport barriers, and enhance alpha exhaust. This control knob is far less expensive and more robust than other available external reactor controls (external current drive, external rotation drive, radiating core impurities, etc.) For AT equilibria, the ripple lost trapped alphas become passing particles after diffusing past the separatrix and exit through the divertor before hitting the main chamber. With thin (< 1mm) static liquid metal films on divertor surfaces (e.g. tin, gallium etc.), acceptable material erosion and heat exhaust appear possible. This concept is being developed using quantitative numerical calculations of alpha orbits in realistic magnetic equilibria with novel divertors (See adjacent related posters).

[PP1.065] Coil Designs for Novel Magnetic Geometries to Cure the Divertor Heat Flux Problem for Reactors

Coil designs are developed for novel magnetic divertor geometries with a second axi-symmetric x-point and flux expansion region along the separatrix. Adjacent posters describe how these lead to spreading of heat flux and the possibility of stable, complete detachment to overcome serious physics and engineering problems in reactors. The principal feasibility issue is creating, with simple coils, additional X-points on the separatrix without extensively deforming the magnetic field in the main plasma. For the spherical tokamak NSTX, we show that adding one or two poloidal coils suffices to create a divergent flux at the divertor, i.e., a new x-point. The currents and forces for the extra coils are small. We also modify ARIES ST design to show reactor feasibility. Optimized coil designs for PEGASUS, ARIES RS/AT, and a modular ITER retrofit are also being developed. For our calculations we used self consistent code FBEQ, which was used to design NSTX. We also use NCSX tools for optimization of designs with competing physics and engineering constraints.

[PP1.066] Novel Magnetic Geometries to Cure the Divertor Heat Flux Problem for Reactors

A novel magnetic divertor geometry with a second axi-symmetric x-point and flux expansion region along the separatrix is analysed. It can provide a stable, completely detached plasma state compatible with reactor operation; avoiding serious physics and engineering problems: 1) extreme divertor heat fluxes, 2) poor global confinement and high disruptivity due to low edge temperatures, 3) lack of access to lower edge densities with acceptable power exhaust, 4) high radiation fractions in the main chamber, and 5) first wall heat fluxes in the high neutron fluence region. In traditional divertors, detachment results in the propagation of the ionization-recombination front� towards the main plasma energy source, cooling the bulk plasma boundary. Simple robust physical arguments imply that the extra x-point will act as a local attractor for the front. Thus, a high bulk edge temperature would be maintained at the bulk boundary, and the completely detached/highly radiating region can be programmed to occur at whatever location is convenient.� Complete detachment would also be enabled for a much lower bulk boundary density and/or higher SOL exhaust power. See adjacent related posters.

[PP1.067] Divertor-leg instability for finite beta and radially-tilted divertor plate

Plasma in the divertor leg may experience a fast instability caused by sheath boundary conditions (BC). Perturbations cannot penetrate beyond the X point because of very strong shearing in its vicinity. Accordingly, this instability could increase cross-field transport in the divertor leg, and thereby reduce the heat load on the divertor plate, without having any appreciable negative effect on core plasma confinement. A way of describing the role of shearing in terms of the surface resistivity attributed to a ``control plane'' below the X point has recently been suggested (Contr. Plasma Phys., v. 44, p. 168, 2004). We use this BC, plus sheath BC at the divertor plate. We include effects of finite beta and of the radial tilt of the divertor plate. We optimize the radial tilt in order to maximize radial transport in divertor legs. We discuss experimental signatures of the instability: i) phase velocity and wave-numbers of the most unstable modes; ii) correlations between fluctuations of various parameters; and iii) the differences between fluctuations in the common and private flux regions.

[PP1.068] Kinetic effects on parallel heat flow and ionization rate in divertor plasmas

1-D simulations of parallel heat flow in divertor plasmas, with and without recycling are made with the UEDGE fluid code, comparing runs using classical flux limited heat flow to nonlocal heat transport [1], now implemented in UEDGE. Comparative simulations are made with the electron kinetic code FPI. For the latter, we prescribe the power input source, which emulates cross field transport, to be identical to that of our UEDGE runs. But, the temperature profile computed by FPI is found to depend very strongly on the assumed velocity dependence of this source, even if the integrated power is the same. The atomic hydrogen ionization module in FPI uses cross sections such that, for Maxwellian plasmas, the rates are the same as those used by UEDGE and DEGAS; this is necessary because step-wise ionization is dominant. There is strong enhancement of the total ionization rate (including stepwise ionization) in cold, detached plasmas, due to nonlocal transport effects.

[1] F. Alouani Bibi and J.P. Matte, Phys. Rev. E \textbf66, 066414 (2002)

[PP1.069] Non-Maxwellian ions and radial force balance equation in a diverted tokamak edge

Experimental interpretation of the electric field profile around the edge pedestal region is usually based upon the fluid radial force balance equation, using the measured ion flow velocity and pressure profiles under the assumption that the ion distribution function is Maxwellian. However, we find from a numerical XGC analysis that the kinetic orbit effect across strongly sheared radial electric field and steep pressure pedestal drives the edge ion distribution to non-Maxwellian. In this work, the non-Maxwellian property of the ion distribution function in the plasma edge will be numerically analyzed and the non-Maxwellian correction to the radial force balance equation will be presented. Comparison with an experimental observation will also be presented.

[PP1.070] Neoclassical polarization drift in a sheared radial electric field in tokamaks

Neoclassical polarization drift of plasma ions is of critical importance in the dynamics of a sheared radial electric field. Neoclassical polarization drift speed V_np of collisionless single ions is studied numerically in a time-varying, spatially sheared radial electric field in a realistic tokamak geometry, using a Hamiltonian guiding center code, together with a qualitative analytic study. A simple dependence of V_np on the conventional time varying radial electric field dEr/dt and its radial shear dEr^\prime/dt is identified. The latter depedence is due to the finite banana width effect. A simple approximate analaytic formula has been developed.

[PP1.071] Comparative Study of Neutral Transport Models for Edge Plasmas

Neutral particles in the edge region of fusion plasmas are strongly coupled to the plasma through the atomic physics processes. While the most accurate treatment of neutral transport is provided by Monte Carlo codes, a much simpler fluid model such as the one used in the fluid edge code UEDGE(F. Wising et al., Contrib. Plasma Phys. 36) (1996) 136. is generally believed to provide reasonably accurate results for typical parameters in the edge plasmas of fusion devices at a low computation cost and absence of random noise. However, the fluid model cannot handle long neutral mean free path regimes. A relatively new model of neutral transport based on the Transmission amp; Escape Probabilities (TEP) method and implemented into the 2D code GTNEUT(J. Mandrekas, Comput. Phys. Commun. 161) (2004) 36. stands between the Monte Carlo and the fluid approaches in terms of complexity and computational cost. It is fast, has no statistical noise and can handle both long and short mean free paths. In this study a series of benchmark tests for all three models is conducted in 1D and 2D model geometries and real 2D geometry of the DIII-D tokamak and the relative merits of all three models are discussed.

[PP1.072] New developments in the TEP neutral transport methodology

The Transmission and Escape Probabilities (TEP) method (W.M. Stacey, J. Mandrekas, Nucl. Fusion 34) (1994) 1385. is a computationally efficient and accurate technique for the calculation of neutral transport in edge plasmas. The method has been implemented into the GTNEUT code(J. Mandrekas, Comput. Phys. Commun. 161) (2004) 36. which has been benchmarked extensively against Monte Carlo and experiment. Recently, the TEP methodology and the GTNEUT code have been extended to relax certain restrictive assumptions in the original formulation, namely the requirement of an isotropic distribution function at the interfaces and the assumption of a spatially uniform first collision source. A double P1 (DP1) expansion allows distributions with linear anisotropies at the interfaces, extending the accuracy of the TEP method to cases where anisotropic effects are important. Three different approaches are compared to deal with the non-uniformity of the first collision source (subdivision into smaller computational regions, spatially-dependent expansion functions and diffusion theory calculation of the directional escape probabilities). Benchmarks with Monte Carlo simulations are presented.

[PP1.073] Effect of Divertors in NCSX

We have used magnetic field data generated by the PIES 3D MHD equilibrium code (M50 coil set) and a new vacuum field code [1] together with the latest numerical model of the first wall [2] to compute wall heat-loading in the National Compact Stellarator Experiment (NCSX). Heat flow is traced by following field lines, with field-line diffusion used to mimic the effect of particle scattering, and the local heat flux estimated from the strike-point density of escaping field lines. This extends our earlier work [3] by including the effect of divertors, whose size, location and configuration are varied to minimize estimated wall damage. Error scaling of the field-line integrator is also presented.

1. Michael Drevlak, MPIPP, Greifswald, Germany, private communication 2. Art Brooks, PPPL, private communication. 3. T. B. Kaiser, et al, BAPPS 48, 304 (2003).

[PP1.074] Effect of Transport Changes on the MARFE Density Limit in TEXTOR

A transport model for the edge plasma of TEXTOR in normal limiter and Dynamic Ergodic Divertor (DED) operation is described. The model includes the poloidal variation of transport characteristics due to the Shafranov shift and the magnetic field perturbations of the DED. The Shafranov shift depends on the poloidal beta. The DED affects external perturbations of the magnetic field which modulate the particle and energy fluxes, and induces changes in the plasma rotation. In addition, variation of the vertical magnetic field shifts the plasma column horizontally, changing the deposition of particles and energy. The effects of these changes in transport and the plasma-wall interaction are modelled for TEXTOR. Based on this model, we try to explain the experimentally observed changes in the MARFE density limit.

[PP1.075] Revisiting the BOUT simulation of Quasi-Coherent mode in Alcator C-Mod

Simulation of the Quasi-Coherent (QC) mode in Alcator C-Mod (A. Mazurenko et al., Phys. Rev. Lett. 89 (22) (2002).) has been one of most successful applications of the tokamak edge turbulence code BOUT. The code was able to reproduce the frequency (f \sim 100 kHz) and the wavenumber, and other features of the QC-mode. In that simulation the QC mode was interpreted to result from a three-wave coupling process involving two high-frequency branches, a shear Alfven wave and a geodesic acoustic mode, both with frequency over 1 MHz. However, these high frequency modes were not observed in recent dedicated experiments. Recent revisions of BOUT uncovered several problems which have been eliminated in the code, and the QC mode simulations were revisited with the upgraded BOUT. A mode similar to the QC mode is still reproduced by the code, while the two high frequency branches are not present, indicating that the three-wave coupling process is not necessary for the QC-mode. We discuss the current status of the QC-mode simulation.

[PP1.076] Testing an H-mode Pedestal Model Using DIII-D Data

Tests against experimental data are carried out for a model of the pedestal at the edge of H-mode plasmas based on double-Beltrami solutions of the two-fluid Hall-MHD equations for the interaction of the magnetic and velocity fields.(S.M. Mahajan and Z. Yoshida, PRL 81 (1998) 4863, Phys. Plasmas 7 (2000) 635.) The width and height of the pedestal predicted by the model are tested against experimental data from the DIII-D tokamak. The model for the pedestal width, which has a particularly simple form, namely, inversely proportional to the square root of the density, does not appear to capture the parameter dependence of the experimental data. When the model for the pedestal temperature is rescaled to optimize agreement with data, the RMS error is found to be comparable with the RMS error found using other pedestal models.(T.~Onjun, G.~Bateman, A.H.~Kritz, G.~Hammett, Phys.~Plasmas 9 (2002) 5018.)

[PP1.077] Calibration of ASTRA Model for H-mode Simulations

Experimental data from the International Profile Database is used to calibrate the model that is implemented to predict H-mode temperature profiles in the ASTRA integrated modeling code.(A.Y. Pankin et al, submitted to Plasma Phys. Cont. Fusion (2004).) The transport component of the model consists of four contributions: \hbox(1) Electron thermal transport from the Electron Temperature Gradient mode controls the electron temperature at the top of the pedestal at the edge of the plasma; (2) Neoclassical transport plays a significant role in the ion thermal transport through the pedestal; (3) Resistive ballooning modes can dominate near the edge of the plasma core; and (4) Ion drift modes (Ion Temperature Gradient and Trapped Electron Modes) often dominate in the deep core of the plasma. First, the transport model is calibrated with a subset of the discharges. Then, the ASTRA code will be used to predict the core temperature profiles as well as the height of the pedestal and the frequency of the Edge Localized Modes in a larger set of discharges.

[PP1.078] XGC Study of the Neoclassical Pedestal Scaling Law

After the turbulence suppression and L-H transition, the pedestal build-up in the H-mode layer may be neoclassical. However, the neoclassical physics for pedestal plasmas cannot be analyzed by a conventional theory due to ion orbital spread in steep gradient, ion loss cone near the X-point, and the neutral effects. Massively parallel Monte Carlo guiding center ion code XGC (X-point included guiding center code) is used to study the maximal edge pedestal buildup and to find pedestal scaling law in the absence of ELM. A pedestal is formed by balance between the radial orbit mixing, particle source from neutral ionization, and the strong convective particle loss around the X-point. A strong electric field well develops in the pedestal region and plasma flows develop. Agreement of the pedestal density profile with the Tanh-fit is remarkably good. The pedestal width is in rough agreement with the experiments. The density pedestal width shows an offset linear behavior in the square root of the pedestal ion temperature. Scaling studies with magnetic field show rather surprising results. The pedestal width does not show a 1/B_\theta dependence. But, it rather shows a 1/B_T dependence. Thus, the density pedestal width does not scale as the poloidal ion gyroradius!

[PP1.079] Multi-dimensional structure of the electric field in tokamak H-mode

Formation of the poloidal shock structure has been predicted theoretically under the existence of a large poloidal flow as in tokamak H-mode. The poloidal electric field induces convective transport in the radial direction, so general understanding of the steep structural formation mechanism is needed to include the poloidal structure. We propose a two-dimensional model with a shear viscosity term, which couples radial and poloidal structures. This model gives a two-dimensional electric field profile involving steep structures both in the radial and poloidal directions. The poloidal shock structure is influenced by the smoothing effect of shear viscosity, but there remains poloidal asymmetry when the poloidal flow has large shear in the radial direction. In the case of the electrode biasing H-mode, the poloidal electric field gives a E�~B flow oriented to the radial direction, whose maximum velocity exceeds 10[m/s].

[PP1.080] Current Drive and RF

[PP1.081] Electron cyclotron current drive near low order rational magnetic surfaces in tokamaks

The reduction of power absorption and generated current density in the vicinity of a low order rational magnetic surface in tokamak is demonstrated numerically for the 2-nd harmonic electron cyclotron current drive using the extraordinary wave propagating in the mid-plane. Power and current density profiles have been obtained from the 4D Monte Carlo modeling of the electron distribution function taking into account the variation of this function on the magnetic surface and the finite wave amplitude. The reduction of absorption is caused by a local plateau formation on the electron distribution function on the field lines which re-enter the radiation beam after a small number of toroidal turns. The influence of the above effect on the tearing mode stability index, \Delta^\prime, is discussed.

[PP1.082] ITER-ECRF TOP LAUNCHER OPTIMISATION STUDIES

Control of neoclassical Tearing Modes (NTMs) by localized electron cyclotron current drive (ECCD) is seen to be one of the key functions of the ITER-ECRF Top Launcher. The RF beams, launched at an �optimum� toroidal angle, should be poloidally steered in order to provide far off-axis current drive capability for (3,2) and (2,1) NTMs stabilization. Extensive calculations have been carried out by the beam tracing ECWGB code [1] to assess how the top launcher can be optimized to best fulfill its task, taking into account that the figures of merit for the NTMs control are a good localization of the driven current at the relevant surfaces, maximum CD efficiency and minimum width of the current profile. Detailed calculations by ECWGB code are presented here for various ITER-FEAT relevant plasma scenarios. The required steering range is evaluated and the results showing the effects of different beams on the peak value and on the width of the driven current density profile are discussed.

[1] D. Farina, S. Nowak, G. Ramponi, ECWGB: a beam tracing code for EC heating and current drive, IFP Report FP 03/6 (October 2003), http://www.ifp.cnr.it/publications/2003/FP03-06.pdf

[PP1.083] Kinetic Modeling of Electron Berstein Waves

The goal of this work is to be able to nonlinearly model edge effects of Electron Bernstein Wave (EBW) propagation into a plasma. Nonlinearity could be especially important at the edge, where the plasma pressure is low. EBW's must be modeled by a full kinetic approach as they depend on the details of the velocity distribution. Thus, we propose to use particle-in-cell methods. Initial simulations show that standard PIC methods are too noisy for modeling, with the noise generated fields being large compared with the launched fields. To improve this situation, we have implemented variable-weight particles in VORPAL, a massively parallel, hybrid plasma modeling code. We are further implementing quiet loading mechanisms and delta-f particles to further reduce the plasma noise. Results will be presented.

[PP1.084] Current Drive by Electron Bernstein Waves

Electron Bernstein waves (EBW) could be suitable for driving current in overdense plasmas in spherical tori (ST) like NSTX and MAST. From numerical studies we have found that EBW current drive (CD) by the Fisch-Boozer method is effective in the plasma core, while the Ohkawa method is more effective off-axis. These studies are being extended to self-consistently and kinetically include the synergistic effect of the bootstrap current. We are using the fully relativistic Fokker-Planck code DKE to solve the drift kinetic equation for electrons. The EBW-electron interaction is included in DKE through a quasilinear diffusion operator with the diffusion coefficient obtained from the relativistic dispersion code R2D2. Our objective is to gain insight into the physics of EBW-CD, with bootstrap current, for applications to STs. The dependence of EBW-CD efficiency on the properties of EBWs, the plasma equilibrium, and the location of wave-particle resonance will be discussed.

[PP1.085] Poloidal field effects in multidimensional mode conversion in tokamaks

We consider a D-T plasma in 2-d geometry, representing the poloidal plane of a tokamak, and study wave propagation in the ion-gyrofrequency range. We use the cold-plasma dispersion tensor, a function on 4-d phase space, with realistic geometry and parameters. Ray dynamics is generated by a Hamiltonian H, equal to the determinant of the dispersion tensor; the rays lie in the 3-d dispersion surface H=0. Plots of this surface give significant insight into the ray dynamics. By displaying various 2-d and 3-d slices, we identify regions of conversion (between magnetosonic and ion-hybrid waves), where tunneling occurs between the two sheets of the surface [1]. The poloidal magnetic field induces a smooth deformation of the dispersion surface, making it asymmetric with respect to the midplane. Projecting onto the 2-d physical plane, the ion-hybrid resonance changes smoothly into propagation, whose direction depends on the poloidal-field structure, in agreement with the results of Jaeger et al. [2] and of Nelson-Melby et al. [3]. In conversion regions where the gradients of density and magnetic-field strength are appreciably nonparallel, the rays are helical [4], greatly complicating the conversion process. A simple wedge model is analyzed, capturing the essential behavior of ray dynamics in such regions.

1. E R Tracy, A N Kaufman, A Jaun, PhysLettA91(2001)309 2. E F Jaeger et al, PhysRevLet90(2003)195001 3. E Nelson-Melby et al, PhysRevLet90(2003)155004 4. E R Tracy, A N Kaufman, PhysRevLet91(2003)130402

[PP1.086] Ion cyclotron heating of non-Maxwellian components in fusion plasmas

An important problem in radio frequency (rf) heating of fusion plasmas is the absorption of rf power by non-Maxwellian components such as minority ion species, fusion-born alpha particles, and fast ions associated with neutral beam injection. Heating of these components often occurs at high harmonics of the ion cyclotron frequency where conventional 2-D full-wave models for rf heating are not valid. In this work, the 2-D all-orders full-wave model AORSA is extended to include non-Maxwellian velocity distributions. Results show that the non-Maxwellian nature of the velocity distribution function can affect wave propagation as well as power absorption. For example, the power absorbed by neutral beam ions in NSTX appears to be more localized near the high harmonic resonances than previous calculations with equivalent Maxwellians suggest. Also, fusion-born alpha particles in ITER absorb more power than predicted with analytic approximations.

[PP1.087] Field Induced Chemistry to Prevent RF Breakdown in Fusion Applications

It is known that microprotrusions form in metallic electrodes in times less than 15\mus, due to a combination of high temperature and electric fields. Microprotrusions are field-concentration points that may induce breakdown in antennas, which can degrade performance. We are studying the possibility of using the same concentrated electric fields that cause breakdown to induce chemical reactions that can either cover the protrusions with high work function materials or etch the protrusions away as they form. RF breakdown is being studied using both a low-power DC and a RF parallel plate experiment at the University of Illinois and a high-power antenna simulator at Oak Ridge National Lab. The effects of surface coatings, gas pressure, and magnetic field strength/orientation on breakdown is being determined. Initial data will be presented.

[PP1.088] Nonlinear wave-particle interaction in helically symmetric stellarators with bumpy fields

In a stellarator with helical ripple, the spatial extent of the local minima in the magnetic field can be of order of the resonant interaction region of applied ECRH waves due to short wavelength |B| variation. In this event, particles trapped in the helical ripples in stellarators bounce back and forth along the magnetic field line, and trapped particles get energy from the wave repeatedly. Quasi-linear theory is not valid in describing the wave-particle interaction when the turning points are very close to the resonance region. In this case, the particles stay in the resonance region long enough to make several energy excursions caused by relativistic nonlinear wave-particle interaction. When this is the case, the wave and particles are correlated and the stochastic process is not viable. Nonlinear wave-particle interaction for deeply trapped particles has been described in previous work. In this work, energy absorption of particles from wave-particle interaction is quantified by a combination of analytical and numerical methods accounting for finite parallel motion and magnetic drifts. For energetic electrons, the relativistic mass shift changes the resonance region. This effect is included in the calculation of the energy excursion. For the case relevant to helically trapped electrons, the equivalent diffusion operator of a phase space is derived accounting the dominant nonlinear wave-particle interaction. * Research supported by U.S. DoE under grant no. DE-F02-99ER54546

[PP1.089] Velocity-Space Difffusion Coefficients Due to Full-Wave ICRF Fields in Toroidal Geometry

Bounce-averaged ion velocity-space diffusion coefficients resulting from full-wave code electromagnetic fields in tokamak geometry are calculated by two methods: (1) appropriate averaging over an ensemble of initial conditions of the RF induced velocity ``kicks'' during one transit of the torus cross-section, calculated by direct numerical integration of the Lorentz equation of motion in tokamak and full-wave EM fields; and (2) local Fourier analysis of full-wave fields to obtain wavenumbers and polarizations, followed by analysis with a previously implemented ray-tracing/quasilinear Fokker-Planck code. Diffusion coeffient results from the two approaches are compared. The diffusion coefficients are used in the FP code for calculation of the RF-driven nonthermal ion distributions for C-Mod and ITER test cases.

[PP1.090] Accurate calculation of shielding factor for high energy beam currents

The theoretical basis for neutral beam current drive was well developed nearly two decades ago. A key element of the net driven current is the shielding factor, which includes both classical (friction) and neoclassical (viscous) responses of the thermal plasma to the fast ion current. We review the strengths and limitations of existing formulations of the shielding factor in terms of their completeness in covering: plasma composition (multiple thermal ion species), finite fast ion velocity relative to the thermal electrons (e.g., 80 keV ions in NSTX and 1 MeV ions in ITER), strong shaping of the plasma geometry (low A effects on trapped fraction and viscosity for neoclassical corrections), full beam geometry, etc. We also report progress in the development of an extension to the velocity moments approach to the force balance equations that can incorporate all the above in a consistent manner. A recursive formulation for computing the collisional matrix elements to arbitrary order is developed. Using this procedure, the required number of terms in the Laguerre expansion of the classical electron (shielding current) parallel response to arbitrary beam momentum input (via electorn/beam collisions Ceb) can be accurately estimated as a function of the electron/beam velocity ratio. Extensions to include higher-order electron viscosity responses will be described.

[PP1.091] Increasing Ion Temperature Using RF Power in Hanbit Mirror Device

In Hanbit mirror device, an RF antenna has been used to produce plasma and to heat the ion simultaneously. The plasma density has been sufficiently high but the ion temperature was not high for the RF power until last year. Since neutral-pressure must be sustained highly for easy discharge, the charge exchange collision between ion and the neutral particle is regarded as main reason for the low temperature. In this year, pre-ionization using electron heating by Klystrone is introduced and plasma discharge scenarios with help of pre-ionization are investigated. As primary results of this series of discharges, the ion temperature and anisotropic factor in velocity space are increased. In this paper, a mechanism for the increasing of temperature will be explained using heating code and conditions for MHD stable discharges be disscussed using MHD stability calculation.

[PP1.092] Dusty, Non-Neutral, and Strongly Coupled Plasmas

[PP1.093] Langmuir Probe Interpretation for Plasmas with Secondary Electrons from the Wall

A method is presented for analyzing the electron current to a cylindrical Langmuir probe in a low pressure, hot-filament discharge plasma containing secondary electrons from the wall in addition to colder bulk plasma electrons. Orbit-motion-limited probe theory is applied to each of the electron components, taking into consideration that the secondary electron current is in the saturation region for probe potentials more positive than the wall potential. The method resolves the probe current into ion, secondary electron, and bulk electron components and finds parameters for each. The fitted model curve follows the probe data with less than 5% relative error from below the floating potential to the saturation region. The analysis shows that the probe current of the bulk electrons alone is indistinguishable from zero for probe potentials more negative than the wall potential, indicating that there are indeed no bulk electrons with energies exceeding the ambipolar potential.

[PP1.094] Particle and energy balance in low-density plasma discharges

For nearly-collisionless, unmagnetized plasma discharges, we show that electron particle and energy balance can be found from a model that includes 1) the energy distribution of the newborn electrons from ionization, 2) the rate of heating of confined electrons by collisions with more energetic electrons, and 3) the rate at which electrons are lost over the confining potential barrier. The model is applied to a simple low-density, hot-filament discharge (double plasma device without a grid or surface magnetic fields). The plasma density, electron temperature, and plasma potential are shown to have approximately the values given by the model.

[PP1.095] Field Penetration and Ion Separation in Multispecies Plasmas

Recent plasma opening switch (POS) experiments have characterized the interaction of a pulsed magnetic field with a weakly collisional multispecies plasma[1-3]. Among the observed effects were simultaneous field penetration and ion pushing. Spatially and temporally resolved measurements of the magnetic field, electron density, and heavy-ion velocities provide rich ground for theoretical investigation. Two- and three-dimensional simulations were conducted with a hybrid electromagnetic particle-in-cell code to help understand these phenomena. A coaxial POS was modeled with initial densities and temperatures measured in experiment. Several ion species combinations were initialized to demonstrate ion species separation and field penetration as a function of plasma composition. Field penetration, plasma density evolution, ion velocities, and electron temperature are discussed. [1] R. Shpitalnik et al., Phys. Plasmas 5, 792 (1998). [2] R. Arad et al., Phys. Rev. Lett. 87, 115004-1 (2001). [3] R. Arad et al., Phys. Plasmas 10, 112 (2003).

[PP1.096] The Construction of the Columbia Non-neutral Torus

The Columbia Non-neutral Torus (CNT) is a small (R=0.3 m, a=0.2 m, B=0.3 T) and relatively simple stellarator being built at Columbia University. CNT will be used to study the equilibrium, stability, and transport of non-neutral plasmas confined on closed magnetic surfaces. CNT has four circular coils: two interlocking coils with a variable tilt angle, plus two poloidal field coils. By varying the angle between the interlocking coils, the rotational transform can be varied from 0.2 to 0.6 and the magnetic shear from essentially zero to 20%. The design of CNT is now completed and the construction phase is nearing completion. Pressures in the 10^-10 Torr range have been achieved in our vacuum chamber. To create an electron plasma, the electrons will be injected from multiple tungsten meshes placed directly on the magnetic surfaces. Each mesh will be controlled separately to explore a variety of plasma profiles. A combination of Langmuir, emissive, and sector probes will be used to diagnose the plasma. PC-based PXI system will be used for experimental control and data acquisition. Details on the design, construction and testing of each of these systems in CNT will be presented. Progress on mapping the magnetic surfaces and producing the first plasma will also be reported.

This work is supported by U.S. DOE Grant # DE-FG02-02ER54690

[PP1.097] Studies of non-neutral plasmas in the CNT stellarator

The Columbia Non-neutral Torus (CNT) is a small stellarator (B=0.3 T, R=0.3 m, =0.2 m) designed to study the confinement of non-neutral plasmas on magnetic surfaces and the effects of extreme electric fields in stellarators. The equilibrium, stability, and transport of non-neutral plasmas confined on magnetic surfaces is different from previously studied configurations. The large space charge in a non-neutral plasma allows a study of stellarator confinement in the limit of extreme electric fields, ie. when the electrostatic potential difference \Delta \phi between the plasma edge and the plasma core is much larger than kT/e. The design of CNT is unique and simple. Four circular coils create an ultralow aspect ratio stellarator with good magnetic surfaces within a cylindrical ultrahigh vacuum chamber. An overview of the theory and physics goals of CNT, in particular the plans for the first year of physics research, will be presented. The current status of the construction, which is nearing completion, will also be discussed.

[PP1.098] Numerical Simulation of Three-Dimensional Single-Species Plasma Equilibria on Magnetic Surfaces

The Columbia Non-Neutral Torus (CNT) will be the first stellarator designed specifically to confine non-neutral plasmas on magnetic surfaces. This simple (four circular coils), ultra-low aspect ratio (1/\epsilon=1.8) stellarator provides an experimental means to examine the equilibrium, stability, and transport of such plasmas. In terms of fusion research, this will allow us to study the confinement properties of stellarators in the presence of large (e\phi>kT) electric fields. The device can also potentially confine electron-positron plasmas (the simplest plasma due to mass and charge symmetry), and antiproton-positron plasmas (for production of neutral anti-hydrogen). Presented for the first time are numerical studies of the three-dimensional equilibria of single-species plasma on magnetic surfaces. Applying a finite temperature to the plasma implies that neither the density nor the potential are in general constant on magnetic surfaces. However, a conducting boundary on the outer surface of the plasma forces the convergence of contours on the edge. Equilibrium profiles will be presented for a single-species plasma in [1] a pure toroidal magnetic configuration (tokamak), [2] a twisting ellipse configuration (simple stellarator), and [3] an asymmetric configuration (outer magnetic surface doesn't match conducting boundary). Equilibria for the CNT magnetic configuration will be calculated in the near future. This material is based upon work supported by the National Science Foundation under Grant No. 0317359.

[PP1.099] Plans to observe magnetic pumping transport in a toroidal electron plasma

Electron plasmas with densities of 5\times 10^6 cm^-3 are trapped for 18 ms in a partially toroidal trap with a purely toroidal magnetic field (B_o=200 G, R_o=43 cm, a=4.5 cm) (M.R. Stoneking \textitet al.,) Phys. Rev. Lett. \textbf92, 095003 (2004). The measured confinement time is more than two orders of magnitude longer than all characteristic single-particle drift timescales and unambiguously confirms theoretical expectations that toroidal equilibria exist for non-neutral plasmas. Confinement is believed to be limited by either collisions with neutrals (P \approx 10^-7 Torr) or by transport due to field asymmetries. A new experiment (R_o=15 cm, a=2 cm) is under construction with improved vacuum conditions (P \approx 10^-9 Torr), enhanced magnetic field strength (B \approx 1 kG), and improved field symmetry. The scientific goals of the new project are 1) to observe the magnetic pumping transport mechanism of Crooks and O�Neil (S.M. Crooks and T.M. O�Neil, Phys. Plasmas \textbf3), 2533 (1996)., and 2) to develop a strategy for controlled charge injection into a completely toroidal trap. This work is supported by the National Science Foundation.

[PP1.100] Shear-Limited Test Particle Transport in 2D Plasmas.

Measurements of test-particle transport in pure ion plasmas show 2D enhancement over the 3D diffusion rates, limited by the shear in the overall E \times B drift rotation ømega_E (r).(C.F. Driscoll et al., Phys. Plasmas 9), 1905 (2002). For finite plasma length L_p, axially bouncing particles may undergo many (N_b) correlated collisions before rotational shear separates them in \theta. In the 3D regime of N_b < 1, the measured diffusion agrees quantitatively with theories of long-range E \times B drift collisions, and is substantially larger than predicted for classical velocity-scattering collisions. For shorter plasmas with 1 < N_b < 100, the measured diffusion is enhanced by a factor roughly proportional to N_b. For short plasmas with exceedingly small shear ( N_b > 1000 ), we observe transport rates consistent with estimates for shear-free 2D plasmas dominated by thermally-excited ``Dawson-Okuda'' vortices. Most interestingly, recent measurements in the low shear regime show non-diffusive transport, as would be expected from convection in large slow rotating thermal eddies. Presumably these eddies are excited by thermal fluctuations similar to the one exciting plasma modes.(F. Anderegg et al.,) Phys. Plasmas 10, 1556 (2003).

[PP1.101] Thermal Fluctuations as a Non-destructive Temperature Diagnostic.

We have detected the thermally excited charge fluctuations in pure electron plasmas over a temperature range of 0.05 < k_BT < 10~eV. At low temperatures, the m_\theta = 0, k_z = 1,2,3, ... Trivelpiece-Gould modes are weakly damped and dominate the spectrum. As the temperature increases, the broad random particle component increases between the modes. We have developed 3 different non-perturbative temperature diagnostics. First, the near-Lorentzian emission spectrum of weakly damped modes, calibrated by the plasma-antenna impedance, integrates to energy 1/2 \, k_B T. This method agrees with ``dump'' temperature measurements when \lambda_D / R_p < 0.3. Second, the broad emission spectrum encompassing several modes is compared directly to a kinetic theory calculation. This method works well if L_p / R_p > 20 so that theory properly describes Landau damping. Third, the total frequency-integrated fluctuating charge \delta N^2 on the antenna is compared with a thermodynamic calculation. This method does not depend on the spectral shape of the fluctuation; but one needs to separately determine the plasma density n, L_p, and R_p to calculate the expected \delta N^2 (T).

[PP1.102] Overview of Trapped-Particle-Mediated Transport and Damping Effects in NNPs.

Recent experiments and theory(C.F.~Driscoll et al.), in AIP Conf. Proc. 692, 3 (2003). show that exceedingly small populations (\sim3%) of axially trapped particles, together with confinement field \theta-asymmetries, cause strong bulk plasma expansion and mode damping in pure electron plasmas. The particles are trapped locally by weak magnetic field ripples ( \delta B (z) / B \sim 10^-3 ), or by small variations in wall potentials (\delta V (z) \sim 0.1~ Volt). These trapped particles produce a drag on the DC currents arising from the plasma rotation through confinement asymmetries, producing bulk transport; and a drag on the AC currents associated with excited modes, causing mode damping. This trapped-particle-mediated drag dominates in plasmas with low collisionality \nu, apparently because TPM dissipation scales as (\nu / ømega_R)^1/2. Thus, TPM effects may explain some of the ``anomalous'' transport and damping effects observed on a variety of apparatuses. An overview of TPM transport and damping scalings will be given in relation to prior experimental observations, so that further experiments or analysis can identify (or rule out) TPM effects.

[PP1.103] Trapped-Particle Mediated Asym\-etry-Induced Damp\-ing of Diocotron Modes.

The nominally stable electron plasma diocotron modes (\mboxk_z = 0, m_\theta = 1,2,..., frequency f_m) exhibit strong exponential damping (rate \gamma_m) when the confinement fields have weak \theta- and z-asymmetries. This represents ``trapped-particle-mediated, asymmetry-induced'' damping, due to collisional dissipation at the velocity-space separatrix between axially trapped and sloshing particles. The damping is intimately connected to the concurrent bulk plasma expansion (rate \nu_P \propto \alpha^2 B^-2, when the \theta-asymmetry is a tilt angle \alpha); and to the previously studied damping of the ``trapped particle mode.'' We find that the diocotron damping rate \gamma_m exhibits a robust scaling of \gamma_m = - \nu_P (f_m / f_* ) \alpha^2, where f_* is a dimensional factor depending only on the type of trapping separatrix. For a magnetic mirror separatrix within weak trapping, we find f_* \approx 1 Hz. This scaling is observed for a wide range of plasma parameters, for various asymmetry types and strengths, and for both collisional and stimulated separatrix crossings. Its strong overall dependence on the asymmetry and strength of the magnetic field (\gamma_m \propto \alpha^4 B^-3) explains the ``anomalous'' diocotron mode decay observed in various prior experiments at low magnetic field.

[PP1.104] Motion of Guiding Center Drift Atoms in the Electric and Magnetic Field of a Penning Trap.

The ATHENA and ATRAP collaborations have produced antihydrogen atoms by recombination in a cryogenic antiproton-positron plasma. This paper discusses the motion of the weakly bound atoms in the electric and magnetic field of the plasma and trap. The electric field in the moving frame of the atom polarizes the atom, and then gradients in the field exert a force on the atom. An approximate equation of motion for the atom center of mass is obtained by averaging over the rapid internal dynamics of the atom. The only remnant of the internal dynamics that enters this equation is the polarizability for the atom, proportional to the binding radius cubed. This polarizability is large for the weakly bound and strongly magnetized (guiding center drift) atoms produced in the antihydrogen experiments. Application of the approximate equation of motion shows that the atoms can be trapped radially in the large space charge field near the edge of the positron column. Also discussed are the curved trajectories followed by the atoms in moving from the plasma to a field ionization diagnostic. Finally, the critical field for ionization is determined as an upper bound on the range of applicability of the theory.

[PP1.105] De-Excitation of Guiding-Center Atoms.

The rate \nu at which guiding-center antihydrogen atoms relax to the ground state is determined through theory and simulation. The rate is found to be slow compared to the rate atoms leave the trap in current antimatter recombination experiments.(G. Gabrielse et al.), Phys. Rev. Lett. 89, 213401 (2002); M.~Amoretti et al., Nature (London) 419, 456 (2002). These experiments operate in the strongly magnetized regime where guiding-center atoms(M.E.~Glinsky and T.M.~O'Neil, Phys. Fluids B 3), 1279 (1991). are expected, defined by \chi = r_c / b \ll 1, with r_c the positron cyclotron radius and b = e^2 / kT the classical distance of closest approach. The atoms evolve to deeper binding through two distinct collisional processes: drag on the positron orbit from large impact parameter collisions, and positron replacement from small impact parameter collisions. The rate of energy loss from drag, previously predicted to increase monotonically with binding energy,(L.I.~Men'shikov and P.O.~Fedichev, JETP 81), 78 (1995). is actually marked by an adiabatic cutoff,(E.M.~Bass and D.H.E.~Dubin, Phys. Plasmas 11), 1240 (2004). making rare close collisions the dominant relaxation process at deep binding. A Monte-Carlo simulation confirms this result.

[PP1.106] Electronic and Positronic Guiding Center Ions.

A novel type of guiding center drift ion is described. These ions only occur in strong magnetic fields. They consist of a neutral atom to which either an electron or a positron is weakly bound, at sufficiently large radius that it may be described by \mathbf E \times \mathbf B drift dynamics. The attractive electric field arises from the weak induced dipole moment of the neutral atom in the field of the outer charge. Such ions may occur naturally in astrophysical plasmas and may also have been formed in recent antihydrogen experiments, where their presence would provide proof that deeply bound øverlineH atoms are being created. Binding energies and orbital dynamics are described in two limits: (i) a ground state H atom along with an outer charge in a zero-angular-momentum orbital, and (ii) a classical guiding center drift H atom (a proton about which an electron \mathbf E \times \mathbf B drifts in the Coulomb field) with an electron or positron bound at larger radius. For case (i) the affinity of a positronic H^+ ion is shown via a full quantum calculation to be 2.23 (B/B_0 )^2 e^2 /a, where B_0 = 2.35 \times 10^5 Tesla and a is the Bohr radius. This scaling with B agrees with a recent estimate.(V.G.~Bezchastnov et al., Phys. Rev. A 61), 052152 (2000). For case (ii) much larger binding energies (of order meV) are found because the induced dipole moment of a guiding center atom is much larger than that of ground state hydrogen.

[PP1.107] Screening Enhancement of Energy Equipartition in a Strongly Magnetized Nonneutral Plasma.

An analogy is uncovered between the nuclear reaction rate in a dense plasma and the energy equipartition rate in a strongly-correlated (\Gamma = e^2 / aT \gg 1) strongly-magnetized (\kappa = e^2 Ømega_c / øverlinev T \gg 1) nonneutral plasma. [Here øverlinev = \sqrtT/m.] When \kappa \gg 1, cyclotron energy is an adiabatic invariant. This energy is shared with other degrees of freedom only through rare close collisions that break the invariant. If \Gamma > 1, the probability of such close collisions is greatly enhanced because surrounding charges screen the colliding pair. In the regime \Gamma < \kappa^(2/5), we find that the equipartition rate \nu defined by d T_c /dt = \nu (T - T_c) (where T_c is the cyclotron temperature) is the rate without screening(M.E.~Glinsky et al.), Phys. Fluids B 4, 1156 (1992). multiplied by an enhancement factor f (\Gamma). Interestingly, f(\Gamma ) is identical to the enhancement factor appearing in the theory of nuclear reaction rates in dense plasmas.(E.E. Salpeter and H. Van Horn, Ap. J. 155), 183 (1969). We present molecular dynamics simulations of equipartition. Rate enhancements of up to 10^10 are measured. The greatly enhanced rate may help to explain recent experiments that observed rapid equipartition in a Be^+ plasma.(Jensen et al., submitted to PRL. See also the adjacent poster.)

[PP1.108] Rapid heating of a strongly coupled plasma at the solid-liquid phase transition

Between 10^4 and 10^6 ^9Be^+ ions are trapped in a 4.5 Tesla Penning trap and laser-cooled to \sim1 mK, where the ions form a crystalline plasma with an interparticle spacing of \sim20 \mum. This system is a realization of a strongly coupled one-component plasma. Using Doppler laser spectroscopy on a single-photon transition, we measured the temperature and heating rate of this plasma when not being laser-cooled. We measured a slow heating rate of \leq 100 mK/s due to residual gas collisions for the first 100-200 ms after turning off the cooling laser. This slow heating is followed by a rapid heating to 1-2 K in 100 ms as the plasma undergoes the solid-liquid phase transition at T=10 mK (\Gamma \sim 170). We will present evidence that this rapid heating is due to a sudden release of energy from weakly cooled degrees of freedom involving the cyclotron motion of trapped impurity ions. We will also discuss the prospects for observing the latent heat associated with the phase transition.

[PP1.109] Stability of a Penning trap with a quadrupole rotating electric field

We will present theoretical and experimental studies of the center-of-mass (COM) stability of ions in a Penning trap with a quadrupole rotating electric field. The rotation frequency of an ion plasma in a Penning trap determines the plasma density and shape, and it can be precisely controlled by a rotating electric field. A quadrupole rotating-field scheme can control pure single species plasmas in contrast to a dipole field, which is effective only for plasmas composed of two or more species of ions(X.-P. Huang, et al.), Phys. Plasmas 5, 1656 (1998).. However, the quadrupole field can modify the trap stability because of the spatial dependence of the electric field. In this study we theoretically and experimentally determine the COM stability condition for ions in a Penning trap with a rotating quadrupole field. The experimental results agree with the theoretical prediction.

[PP1.110] Torque-balanced steady states of pure electron plasmas.

The ``rotating-wall'' (RW) technique has been used extensively to compress and increase confinement in single component plasmas. We describe the results of experiments demonstrating a new, robust operating regime in which compressed, torque-balanced steady states of electron plasmas can be achieved over a broad range of RW frequencies, specifically without tuning to plasma modes. The experiments are done in a regime in which the transport is roughly independent of density, and RW heating is balanced by cyclotron cooling in a 5T magnetic field. The plasma density increases until the central E \times B rotation frequency matches the frequency of the applied RW field. The implications of these results for the ultimate limits on single-component plasma confinement and cold, bright beam formation will be discussed.

[PP1.111] A multicell trap for long-term confinement of large numbers of positrons.

There are a number of potential applications of high-capacity and portable antimatter traps. Previously, we proposed a design for a high-capacity, multicell Penning-Malmberg trap for positrons(C.M. Surko and R.G. Greaves, Rad. Phys. and Chem. 68), 419 (2003).. Here, we discuss an improved design based on the results of recent experiments to confine and tailor electron plasmas using the ``rotating wall'' (RW) technique. We are now able to access a regime in which careful tuning of the RW frequency is unnecessary, and transport is insensitive to plasma density and length(J.R. Danielson and C.M. Surko, adjacent poster.). Operating a high-capacity positron trap in this regime offers a number of advantages. The design of a 95 cell trap for N \ge 10^12 positrons and directions for future work will be discussed.

[PP1.112] Temperature profile measurements and equilibrium calculations in a non-neutral plasma.

In 1992 Eggleston, et.al.^1 developed a technique for measuring the radial temperature profile in a pure electron plasma by partially dumping the plasma onto a charge collector. They used a model which described the plasma as a flat-ended cylinder to determine the midplane potential of the plasma and ignored the plasma's contribution to the confining potential hill. These assumptions are fine if the plasma is long and the ring to which the confining potential is applied is also fairly long. For short plasmas and short confining rings, a more general calculation is needed. In this paper we present a variation on the standard equilibrium calculation that allows us to use dumped charge vs. confining potential data to calculate the temperature profile of a pure electron plasma. The fact that part of the Maxwellian velocity distribution has escaped forces a generalization of the form of the equilibrium condition from the usual e^-q \phi/kT. Experimental data and results will be discussed. A simplified measurement similar to the simple r=0 velocity-tail measurement of Eggleston will also be discussed.

^1D.L.Eggleston, C.F. Driscoll, B.R. Beck, A.W. Hyatt and J.H. Malmberg, Phys. Fluids B 4, 3432 (1992).

[PP1.113] Comparison of Calculated Plasma Mode Frequencies with Experiment

We have measured the diocotron and Trivelpiece-Gould mode frequencies, radial density profile and central temperature in a long (0.6 m), cylindrical Malmberg-Penning electron trap at four different magnetic field strengths. The total particle count varied by a factor of 10 and the magnetic field varied by a factor of 3.5. The temperatures were fairly constant. Using an equilibrium code (EQUILSOR), a 2-D particle-in-cell code (RATTLE), and a 3-D particle-in-cell code (INFERNO) we have calculated the frequencies corresponding to the experimental conditions. We will discuss the limitations of the codes and the conditions in which they agree with experimental results.

[PP1.114] Simulations of Trapped-Particle Diocotron Modes with Electrostatic and Magnetic Mirror Barriers

Trapped-particle diocotron modes are formed in Malmberg-Penning traps when an electrostatic barrier is created at the midpoint (z=0) of a non-neutral electron plasma by a ``squeeze voltage'' applied to a narrow ring at z=0. The asymmetric mode (m_\theta=1, k_z\neq 0) decays exponentially with a behavior that scales at low rigidity as B^-2 in simulations, but as B^-1 in experiments.(C. F. Driscoll, et al.) in Non-Neutral Plasma Physics V (AIP, New York, 2003). Magnetic mirror analogs of the electrostatic modes can be created in simulations by introducing a current ring at z=0. For relatively large axial field perturbations needed to create an effective barrier (\Delta B/B \sim 0.25), the seeded asymmetric modes exhibit frequencies and decay constants comparable to electrostatic barrier modes.

[PP1.115] Do Magnetically Trapped Particles Play a Role in Asymmetry-Induced Transport?

It has been suggested(C.~Fred Driscoll et al., in Non-Neutral Plasma Physics V), Martin Schauer et al., eds., p.3 (2003). that magnetically trapped particles play a role in the asymmetry-induced transport observed in our experiment(D.L. Eggleston and B. Carrillo, Phys. Plasmas 10\rm, 1308 (2003).). This magnetic trapping would occur due to the small increase (\beta \equiv \delta B/B \approx 0.4%) in magnetic field at the center of our solenoid and would keep low velocity particles confined to the ends of the trap. To test this suggestion, we have added three coils of additional windings to our solenoid that allow us to adjust the axial field variation \delta B, and have examined the effect of these adjustments on the radial flux resonances we typically observe. Making B as uniform as possible reduces \beta by a factor of five, but this produces little change in the transport. Varying \beta over the broader range -8.5% to 9.5% gives variations of 20-50% in the magnitude, peak frequency, and width of the flux resonances. The flux magnitude decreases with increasing \beta while the resonance width increases. The resonance peak frequency increases with \left|\beta\right|. We have not yet found a model that can explain these results.

[PP1.116] Ion resonance instability of the diocotron mode in the ELTRAP device

The ion resonance instability of the l=1 diocotron mode, due to the presence of a small fraction of positively charged ions, produced by ionization of the background gas, has been investigated in the Malmberg-Penning trap ELTRAP, using electrostatic and CCD diagnostics. The spectral analysis of the induced charge on a sectored electrode exhibits a high peak, determined by the diocotron mode, whose amplitude is related to the radial plasma displacement. It is found that at low electron energy (determined by the bias of the source with respect to the grounded extraction grid) the instability can be triggered by a fast ramp up of the confining voltage. The mode is stabilized when suitable cleaning potentials are applied to the electrodes, to remove the ions in a short time. The dependence of the instability growth rate on the ion lifetime has been investigated using time dependent double-well configurations.

[PP1.117] Non-neutral Plasmas in Strong Quadrupole Fields

Experimentalists at CERN recently created slow anti-hydrogen in double well Malmberg-Penning traps. The most interesting anti-hydrogen physics experiments require that the anti-hydrogen be trapped. The CERN researchers plan to use a multipole, minimum-B configuration, superimposed on the Malmberg-Penning trap, to trap the anti-hydrogen. Experiments at Berkeley and at LANL have established that quadrupole fields are strongly detrimental to the confinement of the non-neutral species from which the anti-hydrogen is formed, but these experiments were performed in traps with relatively low solenoidal fields. It is controversial whether quadrupole fields are detrimental at the higher solenoidal fields used in the CERN experiments. We hope to report on lifetime measurements with full strength solenoidal and quadrupolar fields.

[PP1.118] Numerical simulations on phase space holes and synchronized BGK modes: comparisons with analytical and experimental results

We present the results of numerical simulations of the autoresonant excitation of high-amplitude BGK modes in a pure electron plasma confined in a Malmberg-Penning trap. The numerical results are compared with the results obtained in recent experiments [1], in which the generation of these structures was achieved by driving the plasma with an oscillating external potential, whose frequency was adiabatically decreased in time. As explained theoretically in [2], this downward chirp of the drive frequency traps resonant electrons and, owing to the velocity chirp, decelerates them. This results in significant changes to the velocity distribution function (initially Maxwellian), and leads to the formation of a well-localized ``hole'' in phase-space. As the driving process continues, this soliton-like structure is associated with density perturbations of increasing amplitude, which then persist for many bounce periods after the drive is turned off. Due to the presence of two distinct time scales, a two-dimensional numerical study of the system is rather time-consuming using conventional PIC codes. For this reason, a one-dimensional, radially-averaged PIC code, was developed. Numerical studies of different excitation processes (corresponding to different methods of manipulation of the distribution function) are presented, and are seen to be in good agreement with comparisons with analytical and experimental results.

[PP1.119] Pattern Formation by Passage Through Resonances in Pure Electron Plasmas

Transverse dynamics of trapped pure electron plasmas can be modeled by Euler�s equations of ideal fluids. A class of uniform m-fold symmetric plasma/vortex equilibria in this approximation were discovered by Deem and Zabusky [1], but creation of these states required some nontrivial initial conditions. I will describe a more realizable approach to formation of non-axisymmetric, shape preserving patterns in non-uniform plasmas. We start from an axisymmetric plasma equilibrium, but add an oscillatory driving potential of an appropriate spatial symmetry. We chirp the driving frequency and pass through resonances with either a discrete Kelvin mode [2] or a continuum of �kinetic� modes of the initial plasma equilibrium. Under certain conditions, the driven system enters a persistent, nonlinear resonance regime, yielding a nontrivial, shape-preserving plasma state in the process of evolution. The �kinetic� application is related to excitation of synchronized BGK modes in pure electron plasmas by passage through axial bounce resonances [3,4]. Work supported by the Israel Science Foundation and INTAS. [1] G. Deem and N. Zabusky, Phys. Rev. Lett. 40, 859 (1978). [2] L. Friedland and A. Shagalov, Phys. Fluids 14, 3074 (2002). [3] W. Bertsche, J. Fajans, and L. Friedland, Phys. Rev. Lett. 91, 265003 (2003). [4] L. Friedland et al, Phys. Plasmas (September, 2004).

[PP1.120] NON-NEUTRAL PLASMA RESONANCES IN CROSSED FIELD DEVICES

Currently there is a good understanding of the physical processes involved in the initiation of crossed-field devices [1], however an understanding of the underlying physics of the steady-state operation of these devices is still lacking. As detailed in Ref. [1], a crossed field device is initiated by a linear wave-particle, Rayleigh-like instability in a shear flow. This linear instability then triggers a nonlinear instability, which is a quasilinear diffusion process, which reshapes the density profile into one which will be in equilibrium with the ponderomotive forces of the growing linear rf wave. Here the distribution of the rf wave can be described by a second-order ordinary differential equation (ODE).

As the rf wave grows and saturates, the device enters into a new regime: one where the nonlinear diffusion process creases, the growth rate vanishes, and one wherein there is a steady flow of electrons from the cathode to the anode [2]. Here there is a dramatic shift in the structure of the equations that govern the distribution of the rf fields, with the order of the ODE�s shifting from two to five. This shift in the order becomes apparent when one includes the dc flow of electrons from the cathode to the anode, and shows up as an activation of three new fast modes. These fast modes have three narrow resonances: one at the wave-particle resonance and the other two at electron cyclotron resonances (upper and lower), which correspond to Slater orbits. We will use a multi-scale analysis to study the structure of the solutions of these fast modes, and to describe the physical processes that occur. References: 1. D.J. Kaup, Phys. of Plasmas, 2001, 8, 2473-80. 2. D.J. Kaup, Proc. of SPIE Aerosense meeting, 2002, 4720, 67-74.

[PP1.121] Nonlinear structures and turbulence in the inhomogeneous dusty magnetoplasma

We discuss the magnetodynamics of dusty plasma with application to astrophysics. Magnetic drift waves have recently been discussed [1] in the context of dusty plasmas although it was first addressed long ago with regard to electron MHD. We find that the cut-off for the electromagnetic waves, between the dust and ion cyclotron frequencies, is due to a first order light ion fluid rotation. This leads to the nonlinear Schrodinger equation for the system, which suggests the possibility of strong structural turbulence [2]. For frequencies between the dust cyclotron and the rotation frequencies, the magnetic drift waves can enable magnetic field penetration in the form of magnetic shock wave or electrical current vortex. The scale sizes of these phenomena are found to be below the dust inertial scale, which for dense molecular clouds is on the order of 100 AU. Creation and collapse of such structures may suggest an alternate paradigm for evolution of turbulence in the interstellar plasma than the widely used cascade transport approach. This work is supported by ONR and NASA.

1. Rudakov, L.I., Physica Scripta, T89, 158, 2001. 2. Ganguli, G. and L. Rudakov, Phys. Rev. Lett., submitted, 2003.

[PP1.122] Rocket-born instrument to detect charged smoke and cloud particles in the mesosphere

An instrument has been developed to detect charged, sub-visible aerosol particles in the upper atmosphere. The instrument is designed to fly on a sounding rocket. The air sample flows between four pairs of graphite electrodes biased symmetrically with increasing bias potentials. Electrons, light ions, and heavier charged particles of both polarities are collected mass-selectively on the electrodes that are connected to sensitive electrometers. The design of the detector helps to reduce the effect of the shock on the motion of the aerosols. The gas flow dynamics and collection efficiency are calculated using numeric codes. A laboratory prototype of the instrument has been fabricated and is currently under testing using ion beams.

[PP1.123] High-frequency normal modes of a complex plasma disk

A complex plasma disk is a two-dimensional circular arrangement of particles confined by a parabolic potential well and interacting via a shielded Coulomb force. This system has finite degrees of freedom and therefore a finite number of discrete normal modes. We excite high-frequency compressive modes of a complex plasma disk using the Dusty O. N. U. experimenT (D.ONU.T). We find that some of these modes are localized oscillations involving only three or four of the particles in the disk. Properties of these oscillations will be discussed.

[PP1.124] Direct observation of microparticle gyromotion in a magnetized direct current glow discharge plasma

Laboratory observations of oscillatory motion of charged microparticles have been made in an argon dc glow discharge plasma created within a strong dc magnetic field.^1 Measurements of the oscillation frequency and amplitude are consistent with the expected gyromotion of magnetized dust grains under the ambient plasma conditions. The measurements provide an effective method for the noninvasive determination of the charge on the observed microparticles. The observations also seem to indicate that the neutral drag force on the dust grains may be smaller than anticipated from the classical estimation. This work is supported by the Office of Naval Research. ^1W. E. Amatucci, D. N. Walker, G. Gatling, and E. E. Scime, Phys. Plasmas, 11, 2097 (2004).

[PP1.125] A model for the condensation of a dusty plasma

A model for the condensation of a dusty plasma is constructed [1] by considering the spherical shielding layers surrounding a dust grain test particle. The collisionless region less than a collision mean free path from the test particle is shown to separate into three concentric layers, each having distinct physics. The method of matched asymptotic expansions is invoked at the interfaces between these layers and provides equations which determine the radii of the interfaces. Despite being much smaller than the Wigner-Seitz radius, the dust Debye length is found to be physically significant because it gives the scale length of a precipitous cut-off of the shielded electrostatic potential at the interface between the second and third layers. Condensation is predicted to occur when the ratio of this cut-off radius to the Wigner-Seitz radius exceeds unity and this prediction is shown to be in good agreement with experiments. [1]P. M. Bellan, Phys. Plasmas 11, 3368 (2004).

[PP1.126] Theory and Simulation of a Nonlinear Fluid Model for Void Formation in Dusty Plasmas

A void in a dusty plasma is typically a small and stable centimeter-size region within the plasma that is completely free of dust particles and characterized by sharp boundaries. We present new developments in the theory and numerical simulation of a recently proposed [Phys. Rev. Lett. 90, 075001 (2003)] nonlinear time-dependent fluid model for void formation. This model consists of an initial instability caused by the ion drag and a nonlinear saturation mechanism to a state containing a void. General features of this model have been confirmed in both 1D and 2D numerical simulations. We report further developments of this model, including the effect of convective dust nonlinearity, a more complete momentum equation for ions, and more realistic boundary conditions, in both 1D and 2D. Analytical as well as numerical results will be presented.

[PP1.127] Initial application of stereoscopic particle image velocimetry to transport and boundary phenomena in dusty plasmas

Over the past five years, the Auburn Plasma Sciences Laboratory (PSL) has applied two-dimensional particle image velocimetry (2D-PIV) techniques [E. Thomas, Phys. Plasmas, 6, 2672 (1999)] to make measurements of particle transport in dusty plasmas. Although important information was obtained from these earlier studies, the complex behavior of the charged microparticles clearly indicated that three-dimensional velocity information is needed. The PSL has recently acquired and installed a stereoscopic PIV (stereo-PIV) diagnostic tool for dusty plasma investigations [E. Thomas. et al, Phys. Plasmas, L37 (2004)]. It employs a synchronized dual-laser, dual-camera system for measuring particle transport in three dimensions. Results will be presented on the initial application of stereo-PIV to dusty plasma studies. Additional results will be presented on the use of stereo-PIV for measuring the controlled interaction of two dust clouds.

[PP1.128] Preliminary measurements of kinetic dust temperature using stereoscopic particle image velocimetry

A dusty (or complex) plasma is a four-component system composed of ions, electrons, neutral particles and charged microparticles. The presence of the microparticle (i.e., dust) component alters the plasma environment, giving rise to a wide variety of new plasma phenomena. Recently, the Auburn Plasma Sciences Laboratory (PSL) has acquired and installed a stereoscopic PIV (stereo-PIV) diagnostic tool for dusty plasma investigations [Thomas, et. al., Phys. Plasmas, 11, L37 (2004)]. This presentation discusses the use of the stereo-PIV technique for determining the velocity space distribution function of the microparticle component of a dc glow discharge dusty plasma. These distribution functions are then used to make preliminary estimates of the kinetic temperature of the dust component. The data is compared to a simple energy balance model that relates the dust temperature to the electric field and neutral pressure.

[PP1.129] Effect of boundary condition on chaotic behavior of fluctuation excited in electron cyclotron resonance plasma

The effect of the boundary condition on chaotic behavior of the fluctuation observed in an electron cyclotron resonance plasma is investigated by chaos analysis using measured time series data. It is found that when a permanent magnet cage of 290 mm in diameter and 500 mm in length is introduced in the vacuum chamber, the fluctuation changes from turbulent state to chaotic state in high microwave power region while it changes from chaotic state to periodic state in low microwave power region. The similar result is obtained when a stainless steal cage is used. These results suggest that the geometry of the cage affects the state of the instability. Moreover, it is found that the superposition of multicusped fields created by small permanent magnet cage reduces the chaotic dimension of the instability.

[PP1.130] Spatial Distribution of Cold Antihydrogen Formation

Part P of program listing