Session LP1 - Poster Session VI.
POSTER session, Wednesday afternoon, October 31
Exhibit Hall B,

[LP1.001] DIII-D Transport, Boundary Plasmas

[LP1.002] DIII-D Confinement and Boundary Physics Programs

The DIII-D Research Program has strong efforts in confinement and boundary physics studies. In the area of transport, significant results are reported in the areas of internal transport barriers, the quiescent Double barrier (QDB) regime, fundamental turbulence studies, tests of theory based transport models, dimensionless parameter scaling studies, H-mode access conditions, and the H-mode pedestal structure. In the boundary physics area, studies have focussed on pellet fueling, density control for advanced tokamak scenarios by divertor pumping, impurity transport and sources, scrape-off layer (SOL) characteristics of QH and QDB regimes, effects of divertor geometry, mechanisms of cross-field transport in the SOL, the role of drifts and electric fields, surface erosion mechanisms, and density limits. Recent results will be summarized and the overall program described.

[LP1.003] Edge Similarity Experiments on C-Mod and DIII-D

An experiment to match the dimensionless pedestal parameters in C-Mod H-modes was carried out in DIII-D to determine whether plasma physics alone determines the width of the H-mode pedestal or atomic physics plays a significant role. The \mboxC-Mod shape was scaled up in DIII-D by a factor of about 2.5. Dimensionally scaled C-Mod pedestal heights of electron temperature and density were matched to within 10% at a dimensionally scaled level of power through the pedestal region (a^3/4P \sim\,constant). The phenomenology obtained in DIII-D was similar to C-Mod H-modes, including ELM-free periods (stable to Type\,I and III ELMs) with steady-state pedestal heights and a quasi-coherent edge mode when the heating power was just above the H-mode threshold. This ELM-free phase evolved into a grassy ELMing H-mode comparable to a high density H-modes in C-Mod.

[LP1.004] Development of Methods to Contral Internal Transport Barriers in DIII-D Plasmas

A very important aspect of improving the performance and duration of plasmas with internal transport barriers (ITBs) is to develop the means of controlling the ITBs. Experiments have been performed in DIII-D to investigate the use of ECH, pellet injection and impurity injection in controlling ITBs in plasmas with L-mode edges or QH-mode edges (QDB plasmas). On-axis and off-axis localized ECH was applied (with up to 2\,MW of total ECH power) inside or outside the ITB location to change the ITB position. ECH was applied to \rho=0.1, 0.36, 0.57, and 0.7. This resulted in a substantial reduction in toroidal rotation and increased density. Pellet injection into the plasma edge and at tangency to \rho\approx 0.7 was investigated. Impurity injection (neon, argon, and krypton) into QDB plasmas was investigated as a method of improving the performance of these plasmas with neon being the best of the 3 gases.

[LP1.005] Density Fluctuations During Modulated Off-Axis ECRH in DIII-D

Time-resolved measurements of density fluctuations using beam-emission spectroscopy (BES) have been made on DIII-D discharges with modulated off-axis electron-cyclotron-resonance heating (ECRH). In addition to the local electron temperature, the amplitude and poloidal phase velocity of the fluctuations about the resonance location are found to be modulated as well. The amplitude and phase of the fluctuation modulations are compared to predictions of the GRYFFIN (or ITG) nonlinear gyrofluid code(M.A.\ Beer and G.W.\ Hammett, Phys.\ Plasmas 3), 4046 (1996). supplemented by the GS2 nonlinear gyrokinetic code.(W.~Dorland, et al., in Fusion Energy 2000 (International Atomic Energy Agency, Vienna, 2000).)

[LP1.006] Comparison of DIII-D Experimental Ion Temperature Gradients with the Critical Gradient as Calculated by the GKS Code

The ion thermal diffusivities (\chi_i) in DIII-D discharges exhibit a strong nonlinear dependence on the measured temperature gradients. This non linear dependence has the appearance of a critical gradient in the sense that when the temperature gradient is less than a certain value \chi_i is small and when it reaches or surpasses this value then \chi_i increases rapidly. Here we present a comparison between the measured ion temperature gradients and the ``critical" gradient as calculated by the GKS code. The existence of a ``critical" gradient can depend on whether the electrons are treated adiabatically or kinetically. It also depends on the relative size of the density gradient. For large density gradients the transport due to trapped electron modes can produce transport even when the ion temperature gradient mode is stable. This could eliminate the effect of a critical gradient. We will compare the incremental ion thermal diffusivity deduced from the experimental data with the predictions of gyrofluid and gyrokinetic ITG turbulence simulations

[LP1.007] Rotation Speed Differences of Impurity Species in the DIII-D Tokamak and Comparison with Neoclassical Theory

Toroidal rotation velocity profiles of carbon, helium, and neon have been measured on the DIII-D tokamak with the charge exchange recombination (CER) spectroscopy diagnostic. Neoclassical theory predicts a relation between the toroidal rotation speeds of the different impurities. In order to make an accurate comparison with theory, the CER data analysis requires taking into account the energy dependent charge-exchange cross sections for the different species. Upgrades in the CER analysis code to take this cross-section effect into account are made and checked with transitions that have different energy dependence. The toroidal rotation speed of these impurities was measured in quiescent double-barrier (QDB), RI-mode, and PEP-mode discharges. Differences in the toroidal rotation speeds of the impurities in the same discharge are compared with neoclassical theory using the FORCEBAL/NCLASS and TRV codes.

[LP1.008] Analysis of Angular Momentum Transport in DIII-D Using CALTRANS

We've added simulation and analysis of axial angular momentum transport to our CALTRANS modeling code, and have begun using the new capability to study toroidal rotation in DIII-D. The analysis-mode functionality was benchmarked by comparing derived transport coefficients for diffusion and pinch velocity with model coefficients used in simulations and found to give good agreement. Results are presented for DIII-D discharge conditions representative of high performance negative central shear, quiescent double barrier and electron cyclotron heating targets.

[LP1.009] Overview of H-Mode Pedestal Studies on the \mboxDIII-D Tokamak

H-mode pedestal studies on DIII-D are motivated by the impact of this region on the global energy confinement and stability, and, through ELMs, on the divertor. Studies were divided into work on edge stability, the width of the H-mode transport barrier, and the ELM energy loss mechanism. A lower n edge localized ideal kink-ballooning mode model for edge stability is consistent with the variation in edge pressure gradient with shape, and with the fast growing lower n modes seen as ELM precursors. The transport barrier width is found to be proportional to the edge poloidal \beta with no explicit temperature dependence. A dimensionally similar comparison with Alcator C-Mod suggests that the barrier width is set by plasma physics rather than the edge neutral source, other work however indicates that the neutral source may set the shape of the density profile. The radial extent of the lower n kink ballooning mode, which is a function of the overall q and pressure profiles, may determine the ELM size.

[LP1.010] Dependence of the H-mode Power Threshold on Plasma Shape in DIII-D

The H-mode power threshold has been measured in lower single null (LSN), double-null (DN), and upper single-null (USN) discharges where the only operational difference is the shape of the plasma. The direction of the ion grad B drift was toward the LSN in all cases. In contrast to our previous results at low triangularity (0.3) where the LSN shape had the lowest power threshold, at high triangularity (0.8) the DN had the lowest power threshold of the three configurations (DN: 1.8\,MW, LSN: 2.7\,MW, USN: 6.8\,MW). This result suggests improved access to H-mode for future advanced tokamaks that are based on a high triangularity DN design. However, the high triangularity LSN case had a higher power threshold (2.7\,MW) than recent low triangularity LSN discharges (1.0\,MW). This could indicate that the power threshold increases with triangularity. At this time we can not rule out other effects that may be important to the L-H transition that occurred as a result of changing the triangularity, such as different neutral recycling behavior, x-point to target distances, and wall conditions. An analysis of these experiments will be presented.

[LP1.011] Comparison of Fluctuations in Lower and Upper Single Null Plasmas in DIII-D

Measurements were performed on DIII-D to compare the turbulence in upper single-null (USN) and lower single-null (LSN) plasmas with otherwise identical parameters, focusing on the effect of the geometry on the H-mode transition. Data was acquired in L-mode plasmas slightly below the H-mode threshold and ELMy H-mode plasmas with phase contrast imaging (PCI), beam emission spectroscopy, Langmuir probes, and reflectometry. In the L-mode discharges the poloidal velocity of the fluctuations shows a large shear in LSN (\nablaB drift toward the X-point) a few cm inside the left closed flux surface, while the USN at the same power does not. PCI data show that the k_\theta=0 modes are dominated by radially outward k_r modes in USN and inward modes in LSN. The ELMs are similar in both geometries; the fluctuations above 20\,kHz increase sharply 0.3\,ms before the rise in D_\alpha emission and start to decay immediately, while lower frequencies peak with the emission.

[LP1.012] Recent Results from the Quiescent Double Barrier Regime on DIII-D

The Quiescent Double Barrier (QDB) regime combines internal transport barriers with a quiescent, ELM-free H-mode edge (QH-mode), yielding sustained, high performance plasmas. In recent experiments, physics understanding of the mechanisms leading to the formation of the QDB plasmas has been improved, and both absolute (\beta \leq 3.8%, S_n \leq 5.5 \times 10^15\,s^-1) and relative (\beta_N H_89=7 for 10\, \tau_E) performance increased A signature of operation with a QH-mode edge appears to be very large radial electric fields in the edge and SOL. In the plasma core, simulations and modeling replicate many of the features of the observed transport and fluctuation behavior, including the ion temperature profile and turbulence correlation lengths.

[LP1.013] Evidence for Reynolds-Stress Driven Shear Flows Using Bispectral Analysis

Spontaneous shear flow generation in magnetized fusion plasmas is thought to occur by an interaction of the turbulent Reynolds stress with the shear flow. In this paper we discuss the theory of mean flow generation when described via the bispectrum. Predictions from this theory are then compared with results from simple numerical simulations of plasma turbulence. Finally, we present results from analysis of edge fluctuations from DIII-D during an L-H transition. These results indicate a transient rise and fall of three-wave coupling localized to inside the separatrix. Future work is also discussed.

[LP1.014] Modeling of Steady-State Non-Inductive ITB Discharges with Application to DIII-D

Establishment of near steady-state high-performance discharges with internal transport barriers in the electron and ion heat and the toroidal momentum channels is investigated using the GLF23 and Weiland confinement models. A combination of neutral beam and electron cyclotron heating and current drive is used to optimally shape the current profile for near non-inductive steady-state operation. The GLF23 and Weiland confinement models have had some success in modeling DIII-D discharges and consequently represents our best choice for DIII-D AT scenario development at this time. By starting the modeling with actual high-performance DIII-D discharges, we expect to obtain experimentally realized results. The stability of our simulations is monitored with the BALOO and GATO codes and rf heating and current drive is modelled with TORAY-GA. This computationally instensive modeling approach requires concurrent computing methods in order to be used routinely. We discuss our efforts to date in producing a parallel computational transport environment.

[LP1.015] Analysis Techniques for Observing Zonal Flow Characteristics in the Turbulence Flow Field

Zonal flows, predicted to be nonlinearly generated by turbulence and in turn act to saturate the turbulence through time-varying E\timesB shear, are toroidally and poloidally uniform (n=0, m=0) electrostatic potential structures that are predicted to exist at low frequencies relative to the ambient turbulence. Such structures should be manifest as a time-varying feature in the density fluctuation poloidal flow field through a v_\theta = E_r \times B_T oscillation, and thus might be experimentally identifiable through time-dependent analysis of the poloidal flow. A wavelet-based and time-shifting least-squares minimization technique are used to estimate the time-dependent poloidal flow velocity in two dimensional density fluctuation data and thus search for such zonal flow characteristics. Initial results from application of these techniques to experimental data will be presented.

[LP1.016] Fluctuation Characteristics of the QDB Regime in DIII-D

A new sustained high-performance operating mode, the quiescent double barrier (QDB) regime has been identified in DIII-D. The QDB regime contains compatible core and edge transport barriers. FIR scattering and reflectometer data show that core turbulence is not eliminated. High frequency quasi-coherent modes are often visible in scattering data, and reflectometer data indicates these are localized to \rho \sim\,0-0.4. In QDB plasmas, the core correlation lengths are significantly lower than observed in L-mode. A continuous edge harmonic oscillation (EHO) is normally associated with QDB operation. High resolution reflectometer measurements show that this scrape-off layer density profile is modulated at the fundamental EHO frequency, and indicate that the peak of the EHO is located at the base of the edge profile pedestals, at or slightly outside the separatrix, in good agreement with beam emission spectroscopy measurements.

[LP1.017] Heating Induced Toroidal Rotation and Other Consequences of Anomalous Momentum Transport in Tokamaks

Viscous stress due to driftwave turbulence plays an important role in determining the electric field in a tokamak plasma. Interaction of the neoclassical viscous stress, which damps poloidal rotation, and the anomalous contributions due to driftwaves leads to some surprising phenomena. A derivation and discussion of the momentum transport equations will be given. The GLF23 transport model is used to compute the fluxes and stresses due to driftwaves. Three examples of numerical solutions to the full set of transport equations will be given. The first example illustrates the toroidal rotation generated in heated plasmas even without external torques. The numerical solutions are compared to the qualitative features observed in the Alcator C-Mod tokamak. The second example is the viscous shear layer at the last closed flux surface (separatrix or limiter). The sensitivity of the L/H power threshold to the open field line poloidal flow is demonstrated. The third example is the poloidal spin-up precursor to an internal transport barrier. This was observed on TFTR with balanced neutral beam heating.

[LP1.018] Treatment of Kinetic Electrons in Eulerian Gyro- kinetics

An accurate description of electromagnetic turbulence in fusion plasmas requires an equally accurate treatment of the essential features of the electron dynamics. It is well-known, however, that a numerical treatment of kinetic electrons is challenging. Fast parallel dynamics places a tight Courant limit on explicit solvers. The difficulties are further exacerbated by the appearance of new unstable numerical ``box modes". In this poster we discuss the most robust numerical methods that can be used to overcome the new challenges posed by a kinetic treatment of electrons. We note that these box modes are generally reduced in severity by the addition of electromagnetic fluctuations and the corresponding Alfven timescale of oscillations. Finally, we report on first results for global simulation of electromagnetic turbulence.

[LP1.019] DIII-D Transport Simulations with CalTrans

The LLNL/GA CalTrans project(http://wormhole.ucllnl.org/caltrans/) incorporates the transport codes ONETWO and Corsica under a common framework and user interface to better predict and analyze the DIII-D tokamak experiment. Recent work has focused on ECCD(St.~John et al., 28th EPS, Funchal, 2001; Casper et al., this meeting.) and on angular-momentum transport, the focus of this paper.(See also T.~B.~Kaiser et al., this meeting.) The GLF23 transport model has been added to the code; the model includes velocity-shear-dependent transport coefficients and has been shown to well describe various aspects of turbulence-driven transport in DIII-D. Here we present predictive simulations of DIII-D QDB discharges including angular-momentum transport self-consistently evolved with the GLF23 model. We also present simulations studying rotation in the DIII-D edge, including a simple model for the effects of an externally applied breaking field.

[LP1.020] Nonlinear Electron Response to Electromagnetic Fluctuations in the Zero Electron Mass Limit

In gyrokinetic simulations of electromagnetic turbulence, the electron parallel motion sets a very small limit on the time step for stability - the electron Courant limit - when explicit finite difference schemes are used. Particularly troublesome is the nonlinearity arising from electron motion along perturbed magnetic field lines, when realistic ratios of electron and ion masses are used. We derive reduced equations for the limit of zero electron mass. These do not require solving the kinetic equation for passing electrons, whose contribution to the parallel current is given explicitly in terms of macroscopic quantitites which are the solutions of nonlinear differential equations. Since the electron mass does not appear, the electron Courant limit is removed.

[LP1.021] Solution of Linearized Drift Kinetic Equations in Neoclassical Transport Theory by the Method of Matched Asymptotic Expansions

The classic solution of the linearized drift kinetic equations in neoclassical transport theory for large-aspect-ratio tokamak flux-surfaces relies on the variational principle and the choice of ``localized" distribution functions as trialfunctions.(M.N.\ Rosenbluth, et al., Phys.\ Fluids 15) (1972) 116. Somewhat unclear in this approach are the nature and the origin of the ``localization" and whether the results obtained represent the exact leading terms in an asymptotic expansion int he inverse aspect ratio. Using the method of matched asymptotic expansions, we were able to derive the leading approximations to the distribution functions and demonstrated the asymptotic exactness of the existing results. The method is also applied to the calculation of angular momentum transport(M.N.\ Rosenbluth, et al., Plasma Phys.\ and Contr.\ Nucl.\ Fusion Research, 1970, Vol.\,1 (IAEA, Vienna, 1971) p.\,495.) and the current driven by electron cyclotron waves.

[LP1.022] Finite Orbit Effects in a Rotating Tokamak Plasma

Close to the tokamak axis, the particle orbit widths are not small compared with the minor radius, and the usual assumption of thin orbit widths breaks down. A previous analysis of finite orbits in a stationary plasma(S.C.\ Chiu, V.S.\ Chan, Bull.\ Am.\ Phys.\ Soc.\ 45), 51 (2000) paper BP1-132. has indicated significant deviations of phase space topology from thin orbit theory. In the present work, it is shown that the phase space topology of a rotating tokamak plasma is significantly different from a stationary plasma. Specifically, there exists a region at low energies where particles are trapped at all pitch angles. More significantly, the trapped passing boundary can disappear for co-moving particles within a certain radius when the rotation speed or rotation shear exceeds some thresholds. The implications of these effects on transport are discussed. The orbit widths are calculated as functions of phase space parameters. The transport of particles at low collisionality is estimated and the scaling is discussed. The effect of adding rf heating shall also be discussed.

[LP1.023] A Model for Energy Confinement Scaling of Tokamak Plasmas with Double Transport Barriers

The formation of double transport barriers, one in the plasma interior and the other at the plasma edge, has been observed in many Tokamak experiments for enhancing the plasma energy confinement to a new level. Naturally, the next question is how the confinement varies with changes in global plasma parameters. With the assumption that the discharge is a combination of H-mode and neoclassical transport, a model is being developed to study possible confinement scaling relations. The model takes the plasma as a single fluid (barrier affects both electrons and ions) and divides it into three regions: the central, the barrier and the outer regions, with neoclassical diffusivity for the barrier region and the H-mode diffusivity for the others. The H-mode diffusivity is derived entirely from experimental observations and it is capable of reproducing the electron temperature profile and the confinement scaling relation of H-mode plasmas. The model leaves the location and the width of the barrier region as variables since it is not clear at present what is their dependence on the plasma parameters.

[LP1.024] Co-Toroidal Rotation With Electron Cyclotron Heating in DIII-D

RF electron cyclotron heating (ECH) and current drive in DIII-D are observed to typically reduce the core toroidal rotation velocity and ion temperature when added to target discharges with rotation established by neutral beam injection (NBI). These discharges have T_i > T_e. The explanation most consistent with the data is that higher T_e results in an enhancement of turbulent ion transport due to an increase in T_e/T_i, consistent with a measured increase in turbulence with increased T_e/T_i.(G.R. McKee, this conference.) In contrast, a series of discharges with ECH alone indicate a significant level of co core toroidal rotation, approaching 40\,km/s. Short NBI pulses must be used for ion velocity and temperature measurements, but comparison of times with and without ECH indicate the development of rotation in the direction of the plasma current with the addition of ECH. Profile data and transport analyses will be presented.

[LP1.025] Comparison of Pellet Injection Measurements with a Pellet Cloud Drift Model on the DIII-D Tokamak

Deuterium pellet injection has been used on the DIII-D tokamak from different injection locations to study pellet-fueling efficiency. When the injection point is inside the magnetic axis of the plasma, the fueling efficiency is significantly higher (approaching 100% in some circumstances) than when the injection point is outside of the magnetic axis. Drifting of the pellet cloud from regions of high to low magnetic field has been hypothesized to explain the experimental results. A pellet cloud drift model(P.B.\ Parks, Phys.\ Plasmas 7), 1968 (2000). has been extended and implemented in a code to compare with the experimentally measured pellet deposition profiles. Measurements of the H_\alpha spectra emitted from the pellet cloud have been made and are used to compare with assumed cloud parameters in the drift model. Comparisons of the resulting fuel deposition profile from different injection locations and the model calculations are presented.

[LP1.026] Effects of Known Non-axisymmetric Radial Magnetic Perturbations on the DIII-D Boundary Plasma

Experimental data shows that radial magnetic perturbations from the DIII-D C-coil can change the plasma properties near the separatrix. In some cases, when the C-coils are energized, we observe a flattening of the pressure profile and a repositioning of the edge pressure gradient relative to that predicted by the axisymmetric EFIT model. These changes could result from the formation of a narrow stochastic boundary near the separatrix due to interactions of the C-coil harmonics on resonant surfaces in the high magnetic shear region. We investigate the implications of such a model with a field line tracing code (TRIP3D), that includes both the non-axisymmetric C-coil perturbations and the axisymmetric EFIT equilibrium.

[LP1.027] Turbulent Heat Flux in the Edge of DIII-D

Measurements of the cross-field turbulent heat flux have been performed in the edge of DIII-D using a fast reciprocating Langmuir probe array fitted with a fast (100\,kHz bandwidth) electron temperature diagnostic. Both conductive (due to correlation between electron temperature and poloidal electric field fluctuations) and convective (associated with the turbulent particle flux) terms of turbulent heat flux were measured in L- and H-mode discharges. In H-mode the two terms are comparable near the separatrix and fall off rapidly with radius (e-folding length \sim1.5\,mm). In L-mode the convective term is usually larger than the conductive one and radial decay is much slower (e-folding length up to 10\,mm). Both conductive and convective heat fluxes have ``bursty" character and essentially non-Gaussian statistics. About 50% of the total heat flux is carried by intermittent large amplitude (>2.5 times the root-mean-square fluctuation level) events. Statistics of such events and characteristics of an average event are discussed.

[LP1.028] The effect of drifts opn the edge plasma in DIII-D

Plasma drifts arising from electric fields and gradients of the magnetic field play a significant role in establishing the characteristics of the edge and scrape off layer regions of a tokamak plasma. The effect of these drifts in DIII-D is examined by simulating the edge/SOL plasma using the 2D fluid plasma code UEDGE. We find the drifts create significant poloidal non-uniformities in the plasma potential and pressure. The pressure non-uniformity is supported by plasma viscosity. In addition to changing the distribution of plasma between the inner and outer divertor regions, and thus the divertor power flow, the drifts are instrumental in creating sheared radial electric fields on the closed field lines. This sheared electric field is an important ingredient in the physics of H-mode confinement, creating regions of suppressed turbulence, and large radial gradients in plasma temperature and density at the edge. Simulation of plasmas with the ion grad B drift away from the X-point results in lower shear in the radial electric field. Hence such operation requires higher power to achieve H-mode confinement, as seen experimentally. We describe the simulation results in detail and compare the results with detailed measurements from the DIII-D tokamak in this paper.

[LP1.029] Modeling of neon and argon dynamics during ELMs in DIII-D

Type I ELMs can provide beneficial core impurity regulation during injection of noble gas impurities to reduce the peak divertor heat flux. Since the quantitative relation between ELM heat flux spikes and impurity expulsion efficiency is poorly known, we analyze transient recycling during Type I ELMs in upper single null configurations, for which fast spectroscopic measurements (f~1 kHZ) of NeI and D-alpha emission have been made. Double null configurations with D pellet and Ne gas injection are also analyzed, to explore the recycling impurity response. Observations of stepwise core neon accumulation resulting from discrete ELM events, even in the absence of neon injection, are used to calibrate a detailed neon recycling model. Finally, the effect on neon and argon enrichment values of neutral charge exchange processes during the ELM deuterium flux pulses is compared with dependence on local deuterium puffing ('puff and pump' effect) in lower single null configurations. A detailed MIST ELM-event impurity recycling model, the "b2.5" code, and the b2-Eirene code (IPP-Garching) are used.

[LP1.030] Edge Plasma Simulations for DIII-D Double-Null Configurations

The up/down asymmetry in the edge plasma of a double-null configuration is studied with the 2-d plasma fluid code UEDGE which includes the effect of magnetic imbalance and cross-field drifts. The magnetic balance is characterized by dRsep, the radial distance between the upper divertor separatrix and the lower divertor separatrix, as measured at the outboard midplane. Recent experiments on DIII-D explored the variation in divertor heat load and particle flux when the magnetic configuration was varied from a balanced double null (DN) divertor to slightly unbalanced configurations. For attached plasmas, the variation of the peak heat flux on upper and lower divertors is very sensitive to the magnetic balance, which is consistent with a narrow scrape-off width for the parallel heat flux. For exact magnetic balance, the divertor particle and heat fluxes are different on upper and lower plates due to cross-field drifts. The direction of the ion \nabla B drift determines the sense of the asymmetry, as observed in the DIII-D experiments and in our UEDGE simulations.

[LP1.031] Calculated X-Point Neutral Density and Plasma Distributions for Lower Single Null Discharges in DIII-D

Analyses of plasma diagnostic data from lower single null discharges in DIII-D yield 2-D neutral density distributions that are compared to measurements along a chord through the X-point. The discharges are simulated with the B2.5 plasma and DEGAS neutrals transport codes. The study is focused upon comparison of results for L-mode plasmas having the same X-point location and ion grad-B drift directions toward the X-point (normal Bt) and away from the X-point (reversed Bt). A density scan at fixed input power for normal Bt, and a power scan at fixed density for reversed Bt are reported. For both directions of Bt the core plasma particle confinement times are less than, but within 50corresponding energy confinement times. For approximately the same plasma conditions and input power, the calculated particle confinement, and neutral density and plasma/neutral interactions in the X-point region of the core, show only a weak dependence on Bt.

[LP1.032] UEDGE Analysis of Core Fueling and Carbon Density with Open and Closed Divertor Geometries in DIII-D

The 2-D fluid code UEDGE was used to study the effect of divertor closure on core fueling and carbon density in DIII-D. We analyzed experiments with the open and the closed divertor of two similar, unpumped discharges characterized by n_e/n_e,Greenwald \approx 0.8, I_P\,= 1.37\,MA, q_95\,=\,4.1, \delta_TRI\,=\,0.8, and P_INJ\,= 4.5\,MW. For these higher density discharges, we estimate that the deuterium ionization in the core (the fueling rate) was \approx20% lower with the closed configuration. UEDGE simulations show a doubling of the recycling current at the divertor targets in the closed divertor case, which is offset by lower efficiency of the recycled particles returning to fuel the core from the closed divertor, yielding numerical agreement with the experiment. UEDGE calculated both the absolute carbon density (Z_eff\, \leq 1.6) and its relative change due to divertor closure in good agreement with experiment (15%-30% less). These results tend to confirm the effectiveness of the closed divertor in restricting carbon flow out of the divertor, a factor in lowering carbon ion flux into the core.

[LP1.033] Edge Similarity Experiments on C-Mod and \mboxDIII-D

The detachment of the scrape-off layer plasma from the divertor target plates in tokamaks is considered as the main provision to reduce particle and heat fluxes onto the material surfaces. Plasma detachment arises when the electron temperature at the target falls below 5\,eV, causing an ionisation front to be formed that is well separated from the solid surfaces. Numerical simulations of scrape-off layer plasmas in a simplified 2D geometry using UEDGE show that increasing the core plasma density and decreasing the power into the scrape-off layer plasma reduces the electron temperature at the separatrix, broadens the neutral density profile and, consequently, moves the ionisation front farther away from the target. The position of the ionisation front is controlled by transport processes that take place both at the high temperature and the low temperature side of the front. This contribution examines the significance of these processes, and attempts to describe the location of the ionization front in terms of simple scaling laws.

[LP1.034] Changes in Edge and Scrape-off Layer Plasma Behavior Due to Variation in Magnetic Balance

We report on recent experiments in which the magnetic balance of highly-triangular H-mode plasmas was systematically varied. ``Magnetic balance" is quantified in terms of dRsep, which is the radial distance between the upper divertor separatrix and the lower divertor separatrix, as determined at the outboard midplane. The \nablaB drift was toward the X-point for which dRsep was negative. Sizable changes in edge and core plasma properties, e.g., pedestal density (n_e,ped) and energy confinement (\tau_E), were observed when dRsep was varied between 0 and +2\,cm. However, \tau_E and n_e,ped were insensitive to changes in magnetic balance for dRsep\, <\,0. ELMing was absent on the high field side of the core and scrape-off layer plasmas for double-null divertors (DND), even as ELMing was clearly present on the low field side. Thus, DND operation may largely relax the requirement for ``ELM protection" on the inboard divertors and centerpost in future tokamaks.

[LP1.035] Divertor Plasma Conditions and Pumping Efficiency in DIII-D

This paper describes detailed studies of the upper divertor plasma and pump performance using the upper divertor pumping system for particle control of high triangularity discharges in DIII-D. Inner and outer cryopumps were used independently to control the removal of neutral particles from each divertor leg. Langmuir probes installed across the upper divertor surfaces were used to measure the plasma conditions during pumping. Significant changes in target plate profile shapes were observed as the strike point position was varied. Plasma interaction with divertor structures, such as the inner dome, can significantly affect the core plasma density. The direction of the magnetic field (and ExB drift) affects the total particle removal efficiency. H-mode densities less than <n>/n_G <\,0.3 have been achieved.

[LP1.036] Scrape-off Layer Characteristics of QH and QDB Plasma Compared with ELMing H-mode and Advanced Tokamak Plasma

A constraint for design of future tokamaks is the transients in divertor heat load and other scrape-off layer (SOL) parameters during ELMs. Quiescent \mboxH-mode (QH) and quiescent double barrier (QDB) discharges offer steady-state confinement comparable to H-mode without ELMs. We examined the SOL characteristics of these discharges compared with ELMing \mboxH-mode and Advanced Tokamak discharges, including heat and particle flux profiles. The most obvious difference is increased steady-state peak heat flux in QH and QDB due to operation at lower density. We examined power balance for these discharges by measuring radiated power and first-wall heat flux compared with input power and stored energy. More power is missing than in ELMing H-mode. We find the ELM peak heat flux transient is eliminated in QH and QDB. The large steady-state heat flux stretches the limit of existing first wall technology and suggests a need for enhanced radiative dissipation.

[LP1.037] ELM Energy Transport in DIII-D

The scaling of ELM energy lost from DIII-D plasmas is explored as a function of edge pedestal characteristics. The spatial profile of energy lost from the pedestal is measured for regular, repeating ELMs with the DIII-D Thomson scattering diagnostic by ordering the data in time with respect to the nearest ELM. Fitting the evolution of the edge electron temperature and density produces a pedestal profile just before and just after an ELM. The energy lost from the pedestal can be represented by convected \Delta n, and conducted, \Delta T, energy. The conducted energy is a significant fraction of the ELM total energy at low density, but this fraction becomes smaller at high density. A model is explored where the ELM instability effectively opens field lines allowing parallel transport into the SOL and divertor. Processes that control the parallel energy loss include ELM duration, effective parallel length, flux limits to conduction, parallel convective transport time and sheath limits to the target heat flux. Implications of the DIII-D data with respect to these processes will be explored.

[LP1.038] Evolution of the 2D Spatial Profile of Visible Emission During an ELM in the DIII-D Divertor

The transient particle and energy loads due to Edge Localized Modes (ELMs) are a significant problem for the design of divertors in future tokamak reactors. Detailed understanding of the effect of the ELM perturbation on the 2D distribution of radiation in the divertor is needed to validate computer simulations and investigate mitigation schemes. Gated, intensified, tangentially viewing cameras with wavelength filters were used in combination with tomographic reconstruction techniques to provide 2D profiles of carbon and deuterium emission during ELM evolution in the DIII-D divertor. Preliminary 2D reconstructions of D_\alpha and CIII visible emission during large Type-I ELMs will be shown. The dramatic broadening of the D_\alpha emission profile near the target will be compared with the broadening of the heat flux profiles during ELMs from IRTV. Plans for obtaining the detailed temporal evolution of the 2D spatial profile of the divertor emission throughout an ELM will be described.

[LP1.039] Transport Simulations of DIII-D Discharges with Impurity Injection

Several recent DIII-D discharges with external impurity injection into L-mode plasmas are analyzed with a coupled main plasma and multi-charge state 1\frac 12--D impurity transport code. These discharges exhibit various degrees of confinement improvement, which has been attributed to the synergistic effects of impurity induced enhancement of the E\timesB shearing rate and reduction of the drift wave turbulence growth rate (M. Murakami, et. al., Nucl. Fusion 41) (2001) 317.. Impurity transport is described by empirical and neoclassical transport models. Both the standard neoclassical theory as well as an enhanced theory which takes into account the effects of external momentum input and radial momentum transport (W.M. Stacey, Phys. Plasmas 8) (2001) 158. have been considered.

[LP1.040] Carbon Release Mechanisms in the DIII-D Divertors

Carbon release mechanisms are examined through analysis of fluxes and spectral profiles of C I, CD and C_2 emissions. Physical and chemical sputtering are the major production processes. The C I influx,\Gamma_CD^total, reflects the total carbon fueling rate; contributions from chemical sputtering are estimated from the measured molecular fluxes using \Gamma_CD^total = 52\times \Gamma_C_2 + (\Gamma_CD - 8\times \Gamma_C_2). The first term accounts for dissociation of C_2D_y and C_3D_y and the second for dissociation of CD_4. When flux measurements indicate chemical sputtering dominates, the effective C I temperatures tend to lie in the range 1.0\pm0.2\,eV, as expected from molecular breakup. When P_inj \geq 9\,MW, molecular emissions are not usually evident, and effective C I temperatures reach values consistent with high-energy physical sputtering, 4-5\,eV. These results suggest the fractions of C I generated by each mechanism may be evaluated from its effective temperature. The apparent suppression of chemical sputtering from the plasma-facing a-C:H/DLC layer has been studied with the BBQ and CASTEM codes.

[LP1.041] Thermal Instability Analysis of Gas Fueled DIII-D \mboxH-mode Shots that Achieved Densities in Excess of the Greenwald Density

H-mode discharges with densities well in excess of the empirical Greenwald density have recently been achieved by continuous gas fueling in DIII-D. These results are in marked contrast to early attempts to achieve high densities in DIII-D H-mode discharges by continuous gas fueling, where degradation of confinement and/or the formation of a MARFE coincident with a H-L transition limited the achievable density to the Greenwald density or less, which can be attributable to thermal instabilities in the edge plasma transport barrier.(W.M.\ Stacey and T.W.\ Petrie, Phys.\ Plasmas 7), 4931 (2000). Examination of shots that achieve densities above and below the Greenwald density identify factors that are important to achieving high densities without MARFEs: steep edge temperature gradients, low edge neutral concentrations and an inward pinch.

[LP1.042] Determination of Ar Concentration Evolution Within DIII-D Core Plasma by X-ray Ross Filter Method

Injection of the non-recycling noble gas Ar into the DIII-D divertor is a promising technique for reducing the heat load on the plates; it also seems to improve thermal transport in an advanced operating mode. During such experiments core plasma contamination by migrating Ar can be investigated by measuring the evolution of the Ar concentration profile using the Ross filter method implemented on the fan shaped X-ray poloidal diagnostics on DIII-D. A Ross filter with energy pass band centered on the ArXVII K_\alpha line at 3.14\,keV, discriminating Ar K_\alpha line against background radiation, was used on DIII-D. A high sensitivity to the injected quantity of Ar and good discrimination against Ne was observed. We present reconstruction of Ar concentration profiles from the chord-integrated measurements using the measured T_e and n_e profiles. Using the transport code MIST, the impurity diffusion coefficients can be determined by matching the time evolution of the Ar concentration evolution.

[LP1.043] Accurate Measurements of the Pitch-Angle Scattering of Beam Ions

The pitch-angle scattering rate of a dilute population of 75-keV deuterium ions is measured in a well-diagnosed, relatively quiet, magnetically-confined deuterium plasma. Neutral particle diagnostics detect the fast-ion density in velocity space following a short 10-ms pulse of injected beam ions. The data are compared to the classical theory of diffusion in velocity space caused by many, small-angle, Coulomb-scattering events. Within uncertainties of \stackrel\sim<15%, the data confirm the classical theory.

[LP1.044] Recent Progress with Reflectometer Electron Density Profile Measurements on DIII-D

Over the past several years, UCLA has been utilizing a continuous frequency-modulated reflectometry system to provide edge and core electron density profiles with high temporal and spatial resolution in support of physics research on DIII-D. Recent improvement in data analysis makes it possible to implement automated profile analysis capability for edge and core X-mode measurements. During the experimental campaign this year, the system provided valuable data for physics research in areas such as the edge harmonic oscillation in QDB discharges, far-SOL plasma transport, density pedestal structure, and other phenomena.

[LP1.045] Characteristics of DIII-D Electron Temperature Profiles

There are two points of view on electron thermal transport in tokamak plasmas. The first view is one where the thermal diffusivity is determined by local plasma parameters and the diffusivity determines the temperature profile. The other view is that the temperature profile shape is more significant and thermal transport will adjust itself accordingly to give a resilient temperature profile shape. The formation of electron transport barriers in the plasma interior in which the profile undergoes drastic changes further complicates the issue. Some of these ideas rely on active drift wave turbulent modes or the suppression of these modes. A study of the electron temperature profiles obtained in DIII-D for the different plasma regimes, L-mode, H-mode and plasmas with internal barriers will be presented. Specific attention will be given to the electron temperature gradient scale length and whether a critical gradient length value seems to be limiting the profile shapes.

[LP1.046] Oblique Electron Cyclotron Emission as Diagnostic for Fast Electrons in the DIII-D Tokamak

The DIII-D tokamak is equipped with multi-mega-watts of ECH power for heating and current drive. The wave power deposition is highly localized based on ray tracing calculations, and a small fraction of energetic electrons are produced. These fast electrons can be used as test particles for electron transport studies provided that a diagnostic with good spatial, temporal and velocity resolution is available. We explore the possibility of using the existing high power ECH launchers as receiving antennae for the electron cyclotron emission from these electrons. The signal should be strong enough, but the unfolding procedure is not straight forward due to plasma refraction and absorption of the emitted waves. The DIII-D plasma is optically thick for second harmonic X-modes and viewing dumps are not needed. The right hand cutoff strongly affects the wave trajectories, and ray tracing becomes an essential part of the analysis. The upper hybrid resonance layer is inside the plasma at high densities, X-modes converted from EBW have to tunnel through an evanescent layer to reach the receiving antennae. An algorithm for data analysis will be presented, and the hardware for various experiments will be described.

[LP1.047] Dipoles, Mirrors and Other Plasma Configurations

[LP1.048] A New Velocity Inversion Method for the ZaP Flow Z-Pinch Project

Sheared flow stabilization of the kink instabilities is being studied on the ZaP Flow Z-Pinch Project at the University of Washington.

Operating regimes have been found where a quiescent period 700 times the static growth time of the m=1 (kink) mode data is seen. A 20 chord gated, image intensified ccd spectrometer (ICCD) is used to measure the emission from the plasma. The instrument function of the ICCD varies for each chord. A new inversion technique is used to calculate the radial dependence of the emmisivity, velocity and temperature. This method includes the effect of the instrument function in the calculation of the local emissivity, velocity, and temperature. Emissivity, velocity and temperature data before, during and after the quiescent period seen in ZaP for C III, O IV, and O V will be presented.

[LP1.049] Holographic Interferometry on the ZaP Flow Z-Pinch

The ZaP Flow Z-Pinch Project investigates the effects of sheared axial flow on the m=0 sausage and m=1 kink instabilities. Holographic interferometry is used to measure the behavior of these density perturbations in a sheared-flow Z-pinch plasma. Chord-integrated phase information is recorded during a plasma pulse using a pulsed ruby laser and holographic techniques. Deconvolution of the chord-integrated phase information is used to determine radial and axial electron density profiles. The experimental methods and results will be presented, along with their correlation to the stability of the sheared-flow Z-pinch.

[LP1.050] Status of the Maryland Centrifugal Experiment (MCX)*

MCX, formerly known as the Maryland Centrifugal Torus, is well into the construction phase. The purpose of MCX is to test centrifugal confinement of plasmas and velocity shear stabilization of MHD interchange instabilities. The geometry of the magnetic field is that of a solenoid with axisymmetric mirror end fields. Biasing of an inner core relative to the outer wall produces a radial electric field which will drive supersonic ExB azimuthal rotation and the resulting centrifugal force will contain plasma to the solenoidal portion. In the past year MCX underwent a redesign of the magnetic geometry and vacuum vessel to produce a significantly larger plasma at midplane. The solenoidal field is now produced by two 25 inch ID coils with a maximum strength of 0.5 T and the mirror fields are produced by four 18 inch ID coils with a maximum field of 1.8 T; the mirror ratio is variable from 3-10. The system incorporates a toroidal magnetic field with a strength at midplane up to 0.3 T. The vacuum vessel is 55 cm diameter at midplane with an overall axial length of 410 cm. The plasma will have a radial width of 15-20 cm at midplane and an elongation of 5 -10. A 10 KV, 0.4 MJ capacitor bank has been constructed as the power source, with upgrade capability to 18 KV and 1.5 MJ. Plasma densities in the range (1-10) 10 13 cm-3 , temperatures from 10-50 eV, and Mach numbers greater than 2 are expected. *formerly known as MCT

[LP1.051] NMCX: Numerical Maryland Centrifugal Experiment

The Maryland Centrifugal Experiment (MCX*) is simulated using a 3D MHD + transport code (UMMHD). The solenoid + mirror magnetic configuration is used. The discharge is initiated with the vacuum field, constant "room temperature", and constant fill pressure. A simulated voltage is applied between the inner and outer boundaries in cylindrical radius. The plasma is seen to spin up toroidally. There is a concomitant detachment of the plasma from the mirror throats and towards the central solenoid section, demonstrating centrifugal confinement. There is also a rise in the temperature from viscous heating. The system reaches a steady state after several confinement times at Mach 4 with a pressure drop of ~ 90 (the Reynolds numbers are ~ 500). The final state is laminar, even though the code is 3D and random noise was initiated: simple mirrors are grossly flute unstable, more so if there is an outward centrifugal force. We next artificially turned off the velocity shear terms in the code, though retaining all centrifugal and Coriolis force terms. The plasma was seen to go flute unstable in about 6 rotation periods, effectively destroying the discharge. We then restored the deleted terms and laminarity was restored. We conclude that the MCX may be laminar, even without a toroidal field, if high Mach numbers can be achieved. This numerical experiment opens up the possibility of a magnetically confined plasma for thermonuclear fusion that has a particularly simple coil configuration. *formerly MCT

[LP1.052] Convective Cells in Asymmetrically Sourced Magnetized Plasmas

The convection of a magnetically confined plasma resulting from heat and particle sources is studied. It is assumed that the convection is low-level in that the system stays stable to ideal interchange instabilities. A Z-pinch plasma with asymmetric particle and heat sources is considered. It is found that there is no convection if there are no particle sources, independent of the distribution of the heat sources. Particle sources result in convection which in turn influences heat transport. The central temperature, however, may go up or down in response to this convection, depending on the distribution of the source function. For the LDX experiment, these results imply that nonaxisymmetric heat source placements may not be a concern as far as convective cells go. Nonaxisymmetry in particle sourcing, however, must be taken into account.

[LP1.053] Overview of the Levitated Dipole Experiment

The Levitated Dipole Experiment (LDX) [http://www.psfc.mit.edu/ldx/] will be the first experiment able to study high-beta plasma confined by a magnetic dipole with near classical energy confinement. LDX consists of three superconducting magnets and illustrates the role of innovative magnetic technology that makes possible explorations of entirely new confinement concepts. We describe the LDX machine design and detail the fabrication status of the superconducting floating-coil, charging-coil, and levitation-coil. In addition, we summarize (1) our procedure to cool, to inductively charge, and to levitate the 1.3 MA floating coil, (2) our initial diagnostic set, and (3) our experimental and physics plans that answer the key questions of high-beta stability and confinement in the dipole fusion concept.

[LP1.054] Multi-Frequency ECRH on LDX

High \beta plasmas will be produced in the Levitated Dipole Experiment (LDX) using ECRH. In order to control the pressure profile, the LDX experimental plan calls for the use of multiple frequency electron cyclotron heating (MFECH). MFECH was observed to result in increased pressure for the trapped, high energy electrons in mirror experiments.(B. Quon et al, Phys. Fluids 28), 1503 (1985). Initial experiments will establish the parameter regimes for stable plasma operation. We will also study the effects of MFECH on the density and pressure profiles, and the possible formation of convective cells due to non-axisymmetric RF launch. We have several source frequencies available, and we plan for at least two of them, 6 GHz and 10 GHz, to be available for first plasma. To aid our experimental efforts, work is in progress on modifying existing ray tracing codes to include the LDX equilibrium profiles.(A. K. Ram and A. Bers, Phys. Fluids B 3), 1059 (1991).

[LP1.055] The Levitation Control System for the Levitated Dipole Experiment

The confining field in the Levitated Dipole Experiment (LDX) is provided by a 1/2 ton levitated superconducting dipole magnet. This floating coil is charged with 1.5 MA current and will be levitated continuously for the eight hour experimental run day. Earnshaw's theorem states that there exists no statically stable configuration for levitation of magnets. In LDX, the floating coil is levitated by a smaller dipole levitation coil 1.5 meters above. This configuration is unstable vertically, but stable in tilt or horizontal motion. The position of the coil will be monitored with a set of eight laser position detectors giving redundant measurements of the five degrees of freedom of the floating coil. The levitation will then be stabilized by feedback control of the current in the levitation coil. The feedback system is a digital system running on a real time operating system platform. This system is programmed, monitored, and controlled by a second computer using Matlab Simulink. The system is currently being tested on a small model and a larger test is planned before LDX operation. Results from these tests and optimizations will be presented.

[LP1.056] Magnetic Diagnostics in LDX

The Levitated Dipole Experiment (LDX) will investigate the stability and equilibrium of a high-beta plasma confined in a dipolar magnetic field. One of the principal goals of the experiment is to understand the effect of the plasma pressure profile on its global stability. We describe our plan to characterize the pressure profile by iteratively fitting solutions of the Grad-Shafranov equation to the magnetic inputs obtained from various magnetic sensors. Equilibrium fields will be measured using coils and loops placed outside the vacuum vessel. A poloidal array of pairs of orthogonally oriented B_p-coils will be placed at nine positions, and ten poloidal flux loops will be placed at optimal locations. Hall probes will be placed outside of the vessel to supplement the coil measurements. A toroidal array of Mirnov coils will be installed inside the vessel to measure fast (MHz) plasma oscillations caused by instability and to deduce their mode numbers. A detailed design of the proposed diagnostics and the current progress on their construction will be presented. Calibration test results of the B_p-coils will also be given. A general overview of the reconstruction algorithm will be shown.

[LP1.057] Drift Frequency Interchange Modes in a Dipole Confined Plasma at Varying Collisionality

Gross plasma stability can derive from plasma compressibility in the bad curvature regions in closed field line systems such as in a dipole field. In this situation MHD theory predicts that the maximum pressure gradient that is stable is proportional to \gamma, the ratio of specific heats.

We have examined low \beta electrostatic modes using a kinetic approach in various collisionality regimes including the collisional regime (J. Kesner, Phys Plasmas 7 (2000) 3837.)^, (A.N. Simakov, P.J. Catto, R.J. Hastie, submitted to Phys Plas (2001).), the collisionless ion regime expected in the LDX experiment and collisionless ion and electron (J. Kesner, Phys Plasmas 5 (1998) 3675.), which might be expected in a reactor. We show that near marginal stability there is a coupling between the MHD-like mode and a low frequency ``entropy'' mode. The maximum sustainable pressure gradient was found to be dependent on the ratio of the temperature and density gradients (\eta\equiv (n/T)(\nabla T/\nabla n)) as well as on the curvature drift frequency. For \eta=2/3 the MHD stability condition is reproduced. When \eta< 2/3 the mode changes character and the stability criterion becomes more stringent in all collisionality regimes.

[LP1.058] Microstability and Turbulence in Magnetic Dipole Configuration

We present linear and nonlinear gyrokinetic simulations of LDX plasmas. Our simulations include trapped particles, finite Larmor radius, Lorentz pitch-angle scattering, and electromagnetic perturbations (including magnetic perturbations both perpendicular to and along the equilibrium magnetic field). No assumption is made regarding the frequency of the instabilities compared to the bounce time. We have not attempted a more sophisticated collision operator that is needed to see some collisional destabilization effects (discussed by Catto and Simakov). The weak magnetic field on the outboard side of the dipole configuration means that finite beta modifications of the equilibria quickly become significant, even at low average beta. We focus on target LDX equilibria using an equilibrium code developed by Mauel and Garnier, for equilibria with beta as high as O(10). The dipole configuration is an interesting laboratory for the study of the interactions of turbulence and zonal flows, since dipole zonal flows are purely azimuthal and therefore linearly undamped in the absence of collisions. On the other hand, the lack of magnetic shear makes the instability threshold of the tertiary instability (which tends to break up zonal flows) very low. The nonlinear simulations we present will focus on these effects.

[LP1.059] Particle-In-Cell Simulations of Plasma Confinement in a Levitated Dipole

Stability for a plasma confined by closed B field lines is P \times (V^g) = Const., with P = pressure, V = flux tube volume, g = ratio of specific heats (g=Cp/Cv = 5/3). Kesner (J. Kesner, Innovative Confinement Concepts Workshop, Mar. 3-6, 1997) proposed a levitated current ring with the plasma stabilized so as an alternate fusion reactor. We have constructed a 3D electromagnetic particle code for investigating the kinetic physics of plasma confinement in a levitated internal conductor device. We have started by studying plasma confinement by the B field surrounding a straight conducting wire. Because the code is a full particle code we have limited ourselves to electron-positron plasmas. We have also confined our studies to electrostatic interactions. We have used systems which are 128x128x64 with about 50 million particles. A current carrying wire runs down the center and a circular limiter is imposed at the outside. Various temperature and density profiles have been run. Plasma profiles satisfying the theoretical marginal stability condition, obtained by imposing conservation of flux and energy, are stable over the length of our runs. If the density and or temperature profiles are such that we would predict instability, we see instability with growth rates that exhibit the right trend. We find that density gradients produce stronger instabilities than do temperature gradients. If the density is inversely proportional to the flux tube volume and the temperature gradient is such that there is an instability, strong convective cells develop but the density profile is not changed.

[LP1.060] On Equilibria of Finite Pressure Plasma in Magnetic Multipoles

We have considered the finite plasma pressure equilibria in multipolar magnetic configurations by using separable forms of the flux function, thereby drastically simplifying the equations and even allowing us to analytically obtain limiting forms of their nonlinear solutions. We show that these solutions are only stable against interchange mode for quadripolar and sextipolar magnetic configurations. Both low and high pressure forms of the solution in sextipole configuration are explicitly displayed and squeezing of magnetic flux surfaces of the region occupied by plasma, similar to that found for high beta plasma equilibria in dipolar magnetic field [S. I. Krasheninnikov, P. J. Catto and R. D. Hazeltine, Phys. Rev. Lett. 82, 2689 (1999)], is demonstrated.

[LP1.061] Kinetic Stability of Electrostatic Modes in Closed Line Magnetic Fields

Kinetic stability of electrostatic modes is investigated in closed field line devices for ømega_b > \nu > ømega \sim ømega_d \sim ømega_*, where ømega_b, \nu, ømega, ømega_d, ømega_* are the bounce, collision, mode, magnetic and diamagnetic drift frequencies. In such systems two classes of modes are found^1,2, namely a drift frequency entropy mode and a high frequency MHD mode. Both modes are flutes and their stability can be described in terms of \eta =d\ln T/d\ln n and d = ømega_*(1 + \eta)/ømega_d. Ion collisional gyro-relaxation^3 effects are studied. These describe relaxation of the perturbed distribution function towards a Maxwellian and can destabilize a stable entropy mode. They are of order (ømega_di/\nu_ii), would be contained in the ion viscosity tensor in a fluid description, but are not present in Braginskii's equations. Despite collisional destabilization large stable regions are found in d, \eta space. ^1B. B. Kadomtsev, Sov. Phys. - JETP 10, 780 (1960). ^2J. Kesner, Phys. Plasmas 7, 3837 (2000). ^3A. B. Mikhailovskii and V. S. Tsypin, Sov. Phys. - JETP 32, 287 (1971).

[LP1.062] Kinetic Stability of Electromagnetic Modes in Closed Line Magnetic Fields

Kinetic stability of electromagnetic modes is investigated in closed field line configurations for ømega_b > \nu > ømega \sim ømega_d \sim ømega_*, where ømega_b, \nu, ømega, ømega_d, ømega_* are the bounce, collision, mode, magnetic and diamagnetic drift frequencies. A second order integro-differential ballooning equation is derived. This reduces to the ideal MHD ballooning equation when ømega > ømega_d, ømega_*, but also describes a drift frequency entropy mode^1,2 when ømega \sim ømega_d, ømega_*. The entropy mode remains electrostatic and flute-like for arbitrary beta. Stability of the modes depends on \eta =d\ln T/d\ln n, d = ømega_*(1 + \eta)/ømega_d and \beta. The most important collisional effects, ion gyro-relaxation effects^3, are investigated for the entropy mode and found capable of destabilizing it. The results are applied to Z-pinch and point dipole configurations. ^1B. B. Kadomtsev, Sov. Phys. - JETP 10, 780 (1960). ^2J. Kesner, Phys. Plasmas 7, 3837 (2000). ^3A. B. Mikhailovskii and V. S. Tsypin, Sov. Phys. - JETP 32, 287 (1971).

[LP1.063] Experimental Studies Of Hot Startup Plasma In Central Solenoid Of AMBAL-M Mirror Machine

The status of completely axisymmetric ambipolar trap AMBAL-M is presented. Hot plasma is produced by gas discharge source with longitudinal magnetic field. At the initial stage of experiments in the central solenoid MHD-stable plasma has been obtained with the length 6 m, diameter \sim 35 cm, density \sim10^13 cm^-3, ion energy \sim300 eV, and electron temperature \sim70 eV. The measured transverse particle transport corresponds to the transverse lifetime of \sim30 msec. At the present time the experiments on further increase of plasma density and temperature in central solenoid by hydrogen puffing and ICR heating are being prepared.

[LP1.064] Plasma Production with the high harmonics ICRF and central cell NBI in the GAMMA 10 Tandem Mirror

The production of higher density plasma with the plug and thermal barrier potentials is the main issue on the GAMMA 10 Tandem Mirror. Various techniques were employed to obtain the higher density, so far. The radial loss during the potential confinement was reduced with the conducting plates located in the anchor transition region [1]. The improvement of the axisymmetry of the heating patterns by ECRH led to the effective potential confinement [2]. The high harmonics ion cyclotron range of frequency waves (RF3) were introduced for the plasma production and heating. The preliminary experimental results with RF3 and NBI in the anchor cell resulted in the peak density of 4\times 10^12 \,cm^-3 [2,3]. To improve the situation further, NBI in the central cell was newly introduced and the modified conducting plates will be installed soon. Currently, the experiment with RF3 and NBI in the central cell is in progress to optimize the experimental parameters.

[1] K. Yatsu, et al., Nuclear Fusion 39 (1999) 1707. [2] T. Saito, et al., Fusion Eng. Design 53 (2001) 267. [3] M. Ichimura, et al., Phys. Plasmas 8 (2001) 2066.

[LP1.065] Initial Results From INS-E

The INS-E experiment has achieved its first plasma. Operation has been achieved with voltages ~ 1 kV and currents ~ 100 mA. Fluctuation data is presently being analyzed.

[LP1.066] A Self-consistent Discharge Simulation for Hanbit RF Heating Experiment

We develop a theoretical model of RF (radio requency) heating mechanism in a cylindrical plasma discharge sustained by slot and half-turn antennas. In this model, the Maxwell-Boltzmann equations with source current profiles are solved by the mode analysis method and the discharge impedance is calculated from the electromagnetic field quantities by field definition. Utilizing the calculated discharge impedance, we devlope a self-consistent discharge model which is a complete circuit model including the impedance matching network. We inspect heating characteristics for two cases of perfect and imperfect matching conditions using the developed circuit model.

[LP1.067] Proton-Boron Beam Fusion Device Concept with Cooling Electron Beams

One of the main problems with colliding ion beam fusion is the growth of the beam in the direction transverse to the main beam velocity. This beam or emittance growth is due to scattering of the colliding beams and is at a rate in which the particles are lost to the walls of the device before sufficient fusion events can occur for overall energy gain. Another major problem is to efficiently keep the ion beams at the energy of the peak fusion reaction cross section. To address these concerns, a novel proton boron ion beam fusion device concept is explored which uses electron beams at the same longitudinal velocity as the ions to cool the transverse ion beam temperature, and keep the longitudinal ion velocity close to the desired peak fusion reaction cross section. A racetrack configuration is used with a mass and energy analyzer at each end of the racetrack to select, collect and/or reinject the ion and electron beams.

[LP1.068] Proton Collimator for Fusion Energy Extraction

A proton collimator concept is under study for use with various fusion devices such as an inertial confinement fusion (IEC) reactor. G. H. Miley, et al., IEEE Trans. on Plasma Science, 25 (1997), 733. It consists essentially of a pair of coils anti-parallel to an external magnetic channel. Spacing of the coils is equal to the coil radius, forming a "Helmholtz Coil". To eliminate the attractive force between pair coils, a stabilization coil is installed anti-parallel to pair coils. The resulting magnetic configuration is cylindrically symmetric. Currents on each coil are chosen to chancel the magnetic field at the center, forming a hexa-pole magnetic configuration. With the zero-field region near the plasma center, an inertial confinement fusion (ICF) reactor or the IEC could be operated without interference. Isotropic fusion protons, as well as leaking fuel components, from D-D or D-He3 fusion will be collimated by the outer magnetic field. This stream can then be lead to a traveling wave direct energy converter, TWDEC, H. Momota, et al., Fusion Technology, 21 (1992), 2307-2323. or to a thruster for space propulsion. H. Momota, et al., AIAA Joint Propulsion Conf., Huntsville Al. Bombardment of particles on structural devices can largely be avoided by optimizing current ratios on the pair and stabilization coils. Another property of this design is that it scatters charged particles into random directions near the center, providing a separation of low-energy leaking unburned fuel components from energetic fusion products. Such separation is essential for use with a TWDEC or for a space thruster to avoid unwanted waste of costly fuel components. A quantitative discussion of these features will be presented.

[LP1.069] ILLIBS, a RF-Driven Ion Source for IEC Fusion

A unique RF-driven ion source ("ILLIBS") is under development for use with the U of Illinois gridded Inertial Confinement Fusion experiment G. H. Miley et al., IEEE Trans. on Plasma Science, 25 (1997), 733. The ILLIBS unit is attached to an IEC vacuum vessel port, using a magnetic entrance nozzle to direct ions through a mechanical restriction that maximizes the differential pressure between the IEC vessel and source volume. The nominal RF input power is 250 Watts at 13.5 MHz, with a reflected power ~ 20 mainly accelerated to fusion energies by the IEC grid structure analogously to normal IEC operation. The injected ion beam (~3-mm diam.) is focussed at the plasma core region of the IEC. Reflection by the potential structures provides multi-passes through the core. The resulting source-driven plasma discharge differs significantly from prior Paschen-limited discharge operation. Steady-state D-D neutron rates approaching 10*7 n/s are achieved, with a significant reduction of charge-exchange losses, hence improved power efficiency. The ILLIBS source design and parametric studies of IEC operation with it will be presented.

[LP1.070] Stellarator and SOL Physics

[LP1.071] Ideal MHD Ballooning Modes in Three-Dimensional Equilibria: Study of Stability Boundaries

A new set of ideal three dimensional MHD equilibria may be generated by varying plasma profiles in the vicinity of a magnetic surface associated with an existing equilibrium [C.C. Hegna and N. Nakajima, Phys. Plasmas 5, 1336 (1998)]. Two free profile functions, in particular the variation of pressure and rotational transform gradients, appear in the description of the new class of equilbria. In this work, we examine the effect of the profile variations on ideal MHD ballooning stability boundaries as described by generalized s-\alpha curves. In particular, configurations of the magnetic field that are relevant to the HSX stellarator are studied and the field line dependence of the eigenvalues of the ballooning equation is investigated.

This work has been supported by DOE grant DE-FG02-9954546.

[LP1.072] Effect of plasma flows on equilibrium in non-symmetric configurations

Properties of the MHD equilibrium equations in presence of plasma flows are studied for general toroidal configurations. Existence of good magnetic surfaces in quasi-symmetric [1] and quasi-helical [2] toroidal equilibria have been shown recently. It depends essentially on the possibility to satisfy all periodicity requirements for the current density in a sheared magnetic field. We show that the same argument applies to the continuity and the parallel-equilibrium equations in the presence of plasma flows due to an ambipolar potential. In the general ideal case it is impossible to satisfy these two equations simultaneously under the periodicity requirements, so that sheet flows or density layers may result, which would destroy the equilibrium near rational magnetic surfaces. However, inclusion of small cross-field transport of particles and momentum softens restrictions on possible quasi-equilibria without actually destroying the flux surfaces. The full set of equilibrium eq! uations is also analysed by expans ion near the magnetic axis. [0.1cm] [1] M. Tessarotto, J.L.Johnson, L.J.Zheng, Phys.Plasmas 2, 4499 (1995) [0.1cm] [2] J.Nuhrenberg and R.Zille, Phys.Lett A 129, 113 (1998) \smallskip\

[LP1.073] Elimination of Unnecessary Constraints on Stellarator Coils

Engineering optimization of stellarator coils is important for both the design of experiments and the assessment of designs for stellarator power plants. Coil optimization is limited by the constraint of good physics properties for the plasma. Good physics properties can be obtained using a few tens of Fourier coefficients to define the shape of an outermost plasma surface, so no more than a few tens of constraints on the coils should come from physics. When coils are designed by existing methods to support a given plasma shape, the few tens of constraints are turned into hundreds by forcing the normal magnetic field to be zero at a large number of points on the plasma surface. In effect one is constraining the coils to not only reproduce the known Fourier coefficients of an optimized plasma shape but to also zero all the unknown Fourier coefficients. By determining the relation between small changes in the plasma shape and small changes in the field due to the coils, one can determine the constraints on the coils that are required to reproduce all the known Fourier coefficients of the shape. Our work on a code that defines the minimal number of coil constraints will be discussed as well as the insights the method offers on issues of optimized stellarator design.

[LP1.074] Global Search Methods for Stellarator Design

We have implemented a new variant Stellopt-DE of the stellarator optimizer Stellopt used by the NCSX team.(A. Reiman, G. Fu, S. Hirshman, D. Monticello, et al., EPS Meeting on Controlled Fusion and Plasma Physics Research, Maastricht, the Netherlands, June 14-18, 1999, (European Physical Society, Petit-Lancy, 1999).) It is based on the ``differential evolution'' (DE) algorithm,(R. Storn, K. Price, U.C. Berkeley Technical Report TR-95-012, ICSI (March, 1995).) a global search method which is far less prone than local algorithms such as the Levenberg-Marquardt method presently used in Stellopt to become trapped in local suboptimal minima of the cost function \chi. Explorations of stellarator configuration space z to which the DE method has been applied will be presented. Additionally, an accompanying effort to understand the results of this more global exploration has found that a wide range of Quasi-Axisymmetric Stellarators (QAS) previously studied fall into a small number of classes, and we obtain maps of \chi(z) from which one can see the relative positions of these QAS, and the reasons for the classes into which they fall.

[LP1.075] Reduction of islands in full-pressure free-boundary stellarator equilibria.

In three-dimensional plasma confinement devices such as stellarators, the lack of a continuous symmetry implies that magnetic islands will generally be present in the magneto-hydro-dynamic (MHD) equilibrium. For improved confinement it is required that the magnetic islands are small as possible.

The largest islands are associated with low order rational-rotational transform surfaces and are formed by resonant radial magnetic fields, which may be caused by both the plasma currents and the confining coil currents.

To optimize stellarator coil design, the resonant radial fields in MHD equilibria consistent with a given coil set may be computed using the free-boundary PIES code. Derivatives of selected resonant fields in the MHD equilibrium consistent with a coil geometry, with respect to changes in the coil geometry, are calculated and a Newton procedure is used to find a coil design for which the MHD equilibrium has reduced resonant fields and thus reduced island size.

This `coil-healing' procedure is applied to a coil set relevant for the National Compact Stellarator Experiment (NCSX). The coil set is healed for the full pressure equilibrium case. The healed coil set reduced island content for a variety of equilibria at different pressure and plasma current relevant for start-up

[LP1.076] Recent Developments in the PIES Code

The PIES code calculates three-dimensional MHD equilibria without making assumptions about the existence of flux surfaces. We discuss recent improvements to the code, including the incorporation of a new set of island diagnostics based on the theory of quadratic flux minimizing surfaces. Applications to equilibria of interest to the National Compact Stellarator Experiment (NCSX) are discussed.

[LP1.077] Linear and Nonlinear Gyrokinetic Calculations in Stellarator Geometry

The nonlinear gyrokinetic code GS2 has been extended to treat nonaxisymmetric stellarator geometry. Electromagnetic perturbations and multiple trapped particle regions are allowed. Collisions are treated with a pitch angle scattering, Lorentz collision operator. Linear, collisionless, electrostatic simulations of the quasi-axisymmetric, three field period stellarator design QAS3-C82 have been successfully benchmarked with the eigenvalue code FULL, with generous help from G. Rewoldt.\footnote[2] G. Rewoldt, L. P. Ku, and W. M. Tang, Phys. Plasmas \textbf6, 4705 (1999). Geometric quantities needed for the gyrokinetic simulations are calculated using the VVBAL code of W. A. Cooper.\footnote[3] W. A. Cooper, Plasma Phys. Controlled Fusion \textbf34, 1011 (1992). Furthermore, a comparison between the variation of the critical temperature gradient for axisymmetric and nonaxisymmetric configurations for various parameters, including density gradient and collision frequency, has been explored. Finite beta, collisional, and nonlinear electromagnetic effects on stellarator geometry have also been investigated, using recent NCSX design points. Limitations of the application of flux tube simulations to nonaxisymmetric equilibria are discussed.

[LP1.078] Drift Mode Calculations in Nonaxisymmetric Geometry

Fully kinetic assessments of the stability properties of toroidal drift modes have been obtained for cases for the Large Helical Device (LHD) and for cases obtained in the course of the design process for the quasiaxisymmetric National Compact Stellarator Experiment (NCSX). This calculation employs the comprehensive linear microinstability code FULL, as recently extended for nonaxisymmetric systems. The code retains the important effects in the linearized gyrokinetic equation, using the lowest-order ``ballooning representation'' for high toroidal mode number instabilities in the electrostatic limit. These effects include those of trapped particles, FLR, transit and bounce and magnetic drift frequency resonances, etc., for any number of plasma species. Results for toroidal drift waves destabilized by trapped electrons and ion temperature gradients are presented, using numerically-calculated three-dimensional MHD equilibria. For LHD these are reconstructed from experimental measurements. The effects of helically-trapped particles and helical curvature are investigated.

[LP1.079] HINT Computations of LHD Equilibria with m/n=1/1 Island

A Large Helical Device (LHD) equilibrium with islands is numerically studied by using the three dimensional MHD equilibrium code HINT with a full torus calculation. Especially, we report on recent HINT computations of the LHD equilibrium with an m/n=1/1 island formed at the edge region. Here, m is a poloidal mode number and n is a toroidal mode number. An LHD equilibrium with the m/n=1/1 island is required in the local island divertor (LID) experiment. The LID is a divertor which uses the m/n=1/1 island formed at the edge region. The LID has been proposed to control the edge plasma of LHD. The island is useful for control of heat and particle fluxes. Control of the edge plasma by means of the LID is expected to realize the high temperature divertor operation which leads to a significant energy confinement improvement. As a result of the HINT computation, we see that good flux surfaces are kept in the case of low beta.

[LP1.080] Theory Modeling of Second Harmonic Electron Cyclotron Plasma Breakdown in a Quasi-Helical Plasma

An analytic model is presented to describe the breakdown processes of a Stellarator plasma produced by second harmonic cyclotron waves. The important wave-particle process involves the nonlinear interaction of ECRH waves with cold electrons trapped near the resonance region. The nonlinear interaction causes the initially cold electrons to undergo excursions in energy which can ionize the plasma through collisions with neutrals when enough wave power is provided. To describe the plasma population growth, a three-group model (energetic trapped electrons, energetic passing electrons and cold untrapped electrons) is presented that accounts for the nonlinear wave-particle interactions, pitch angle scattering and coalitional slowing down on neutrals, ionization processes and particle loss rates due to transport. The density evolution equations for each group can be solved and the population growth rate is computed. In this work, we emphasize the role of the magnetic configuration on the plasma loss rates and detail its effect of the plasma breakdown rate. We apply this model to various magnetic configurations available to the HSX experiment.

[LP1.081] Detachment: synergy of the perpendicular transport and atomic physics effects

The conventional physical picture of plasma detachment phenomena per se in fusion related devices relies solely on parallel plasma transport and atomic physics effects including the plasma recombination, although we should note that perpendicular transport in the scrape off layer (SOL) was considered to be important for enhancement of impurity radiation loss. Such a physical picture was, to some extent, supported by experimental observations of the plasma recombination in detached regimes. However, recent experimental and theoretical developments show that cross-field plasma transport in the SOL is actually much faster than it was previously assumed to be. It was shown that the SOL transport has a spiky convective character where plasma blobs are quickly moving towards the main chamber wall, forcing the plasma to recycle there rather than to flow into the divertor. These findings can significantly alter our understanding of the plasma detachment mechanisms including the effects of the recombining plasmas. Here we analyze the impact of the plasma cooling due to the plasma-neutral interaction and resulting plasma recombination on the blob motion, particle balance, detachment, and plasma first wall recycling in both tokamaks and divertor simulators.

[LP1.082] Numerical Simulation of Blob Propagation through SOL and Divertor Plasmas

In recent years blobs (plasma filaments) generation and propagation is considered as a possible mechanism for a fast radial transport in scrape-off layer and divertor plasma tokamak regions[1,2]. We use BOUT code[3], which is based on 5-field fluid Braginskii model to describe transport and turbulence phenomena in realistic tokamak divertor geometry, to simulate cross field blob transport in scrape-off layer of tokamak plasma. At this stage our consideration is restricted by a circular cross section plasma with belt limiter. We consider evolution of a seeded plasma density perturbation (blob) and its propagation through the SOL plasma. Influence of initial and boundary conditions as well as SOL plasma turbulence on blobs density, temperature and propagation are studied. [1] S.I.Krasheninnikov, Phys. Let. A 283(2001),368; [2] D.A.D'Ippolito, J.R.Myra, S.I.Krasheninnikov, LRC Rep.LRC-01-83; [3] X.Q.Xu, R.H.Cohen, Contrib.Plasma phys. 38(1998),158.

[LP1.083] Cross-Field Blob Transport in Tokamak SOL Plasmas

Non-diffusive, intermittent transport of particles can play a major role in the SOL of fusion experiments, as indicated by accumulated experimental evidence showing flat profiles out to the walls. A possible mechanism for fast convective plasma transport is related to the structures (“blobs”) observed in the SOL plasma with fast cameras, optical diagnostics (BES) and probes. Physical arguments suggesting the importance of blob transport(S.I. Krasheninnikov, Physics Letters A 283, 368 (2001).) have been extended by calculations(D.A. D'Ippolito et al., LRC Report LRC-01-83, July, 2001.) using a 3-field fluid model and treating the blobs as coherent propagating structures. The properties of density, temperature and vorticity blobs, and methods of averaging over blob ensembles to get SOL profiles, will be discussed. Qualitative features of the data (e.g. flat profiles and transport coefficients increasing toward the wall) are shown to emerge naturally from the blob transport paradigm. Quantitative comparisons of the theory with DIII-D SOL turbulence data(J.A. Boedo et al., Phys. Plasmas, 2001.) will also be presented.

[LP1.084] Curvature driven nondiffusive transport in the scrape off layer

Many recent experimental results have indicated that particle transport in the scrape off layer is charaterized by fast convective radial transport that produces poloidally localized blob-like structures. In order to gain some insight as to the cause of this transport, we investigate the transport cased by the \nabla B driven drift instability in the scrape off layer. In this simple model, the perpendicular current consists of the ion polarization drift and the curvature drift, while the parallel current is given by the well known Debye sheath condition. Results of a nonlocal linear analysis show that the strongest growing modes occur for small radial and large poloidal mode numbers. Fully nonlinear simulations of this system show that this mode does indeed produce large, anomalous levels of particle transport, qualitatively similar to that observed in experiments.

[LP1.085] UEDGE code modelling of non-diffusive cross-field transport in DIII-D edge plasmas

We use the 2D multi-fluid code UEDGE [1] to simulate the effect of intermittent plasma transport that have been recently observed experimentally on various plasma devices (e. g. [2]) and studied theoretically in [3]. This transport, which is non-diffusive, is simulated by UEDGE in terms of anomalous cross-field convective velocity Vp directed and increased outward the plasma column. The 2D profile of Vp is adjusted until the simulated radial profiles agree with measurements in the scrape-off layer and in the divertor as well as with measurements related to the impurities and the recycling of neutral particles. Our study of several DIII-D dischages shows that Vp of about 100 m/s at the chamber wall is required to match the experimental data in both the L and H modes. Based on modelling of double-null magnetic configuration, we discuss the effect of anomalous plasma convection on inner divertor detachment and on in/out asymmetries of divertor plasma parameters.

[1] T.D. Rognlien et al, J. Nucl. Mat. 196-198 (1992) 347. [2] J. Boedo et al, to appear in Phys. Plasmas (2001). [3] S.I. Krasheninnikov, Physics Letters A 283 (2001) 368.

[LP1.086] Effect of Dipole Perturbation on Area of Footprint of Field Lines on Collector Plate Using Maps

Symmetric Simple Map is employed to represent unperturbed magnetic topology of single-null divertor tokamak. Dipole Map represents the effects of externally applied high MN perturbation. Purpose of the study is to investigate if the area of footprint of magnetic field lines on the divertor plate can be increased by dipole perturbation. Results of this study are presented.

[LP1.087] Assessment of issues for the MAST divertor biasing experiment

A biasing experiment is being undertaken in the MAST scrape-off layer; the goal is to induce intense convection by a toroidally alternating biasing of divertor tiles. This would lead to a thickening of the SOL and a reduction of the heat load on the divertor plates. In addition, by studying the reaction of a plasma to a varying bias, one can collect new information regarding pre-existing SOL turbulence. We consider the following issues: 1. The bias amplitude required to produce significant SOL broadening; 2. Excitation of shear-flow turbulence in convective cells; 3. The role of magnetic shear; 4. Effects of electrostatic sheaths at the divertor plates; 5. Redistribution of heat fluxes during biasing. We show that a significant effect of the biasing on the SOL structure can be reached at relatively small bias voltages ~ 30 V. We also show that the potential perturbations will be limited to a zone between the X-point and the biased tiles, and will be essentially decoupled from the main SOL plasma. Preliminary experimental results may be shown.

[LP1.088] Neutral Gas Transport Simulations of Gas Puff Imaging Experiments on NSTX and Alcator C-Mod

A series of experiments using visible imaging of gas puffs to characterize edge plasma turbulence has been carried out in the NSTX and Alcator C-Mod devices. Their objective was to provide data that can be compared with edge plasma turbulence codes. However, simulations of the transport of the puffed gas and the neutral atomic physics are needed to relate the observed light fluctuations to the local density (and perhaps temperature) fluctuations. The results would also permit an assessment of a ``shadowing'' effect in which a localized density peak near the outer edge of the emitting region sufficiently ionizes the puffed atoms to affect the light fluctuations at smaller radii. The DEGAS 2 Monte Carlo neutral code is used to generate radial emission profiles of the D_\alpha or 5876 Å\ He line that can be matched against observations. Nominal plasma profiles for NSTX are taken from runs of the UEDGE code. Midplane reciprocating probe data from Alcator C-Mod directly specify the average plasma parameters in the region of interest. The sensitivity of the size and location of the emitting region to variations and these profiles will be assessed. Simulations of the view seen by the fast visible camera are also possible.

[LP1.089] Radiation Transport in an Edge Plasma Code

Plasmas in edge regions of tokomaks can be very optically thick to hydrogen lines [1]. The strong line radiation introduces a non-local coupling between different regions of the plasma and can significantly affect the ionization and energy balance [2]. Current edge plasma codes use a local atomic physics model, which cannot model these effects in any realistic manner. We are working to include the effects of a strong, spatially-varying radiation field by coupling an edge plasma code to a non-local thermodynamic equilibrium radiation transfer code. We explain the numerical model being used, discuss the underlying assumptions and approximations of the model, and present preliminary numerical results.

This work was performed under the auspices of the U.S. Department of Energy by the University of California, Lawrence Livermore National Laboratory under Contract No. W-7405-Eng-48.

[1] Terry JL et al, Phys Plasmas 5 (1998) 1759. [2] Scott HA et al, J Nucl Materials 266-269 (1999) 1247.

[LP1.090] Effect of Neutral atoms on Tokamak Edge Plasmas

Neutral atoms can affect tokamak confinement by influencing the plasma behavior just inside the separatrix. Even a small minority of neutrals can have a significant effect due to their high mobility across the magnetic field. Previous work [1], assuming short mean-free path neutrals and ions, has shown that the ion-neutral coupling through charge-exchange can cause the neutrals to determine the radial electric field and, at higher neutral densities, to directly affect the neoclassical flow. However, the mean-free path of the neutrals is not always short in the edge pedestal of a tokamak. To allow for strong temperature and density variation we use a self-similar solution for the neutral distribution function with the ratio of the neutral mean-free path to the scale length a constant [2]. The self-similar treatment broadly confirms the results from the short mean-free path treatment and demonstrates that only the neutral heat flux is sensitive to the the mean-free path expansion. [1] T. Fülöp, P. J. Catto and P. Helander, Phys. Plasmas 5, 3969, (1998) [2] P. Helander and S. I. Krasheninnikov, Phys. Plasmas 3, 226, (1995)

[LP1.091] Z-Pinches and ICF Diagnostics

[LP1.092] Observation of Non-0-D Wire Array Trajectory and Pinch Precursor on the Z Accelerator

We report on the first experiments to measure the implosion dynamics of wire arrays on Z (M.E. Cuneo, et al., 3rd Wire Array Workshop, Abingdon, England, UK, April 2001). We use chordally- and axially-resolved visible and x-ray self-emission diagnostics. These experiments show implosion trajectories somewhat delayed (10-20 ns) from 0-D behavior near the initial array radius, but intercepting 0-D by half the initial radius. We observe a precursor pinch on the axis of the wire array at least 50 ns prior to the final stagnation, consistent with early acceleration of wire corona material. These experiments may indirectly indicate the presence of trailing mass at final stagnation. Comparisons of the data with a variety of 0-D models and 1- and 2-D radiation MHD simulations will be presented. These observations may impact our understanding of array power scaling and methods for array radiation pulseshaping for ICF. Sandia is a multiprogram laboratory operated by Sandia Corp., a Lockheed Martin Company, for the USDOE under contract DE-AC04-94AL85000.

[LP1.093] Theory and simulation of Z wire dynamics in the r-\theta plane

The radial acceleration of low density (relative to the wire cores) coronal plasmas, towards the axis of wire-array Z pinches, is observed in experiments at Imperial College and Sandia National Laboraties (S.V. Levedev et al., Phys. Rev. Lett. 85), 98 (2000); M.E. Cuneo et. al., these proceedings. It is anticipated that the dynamics of the coronal plasmas play an important role in the distribution of both current and density prior to the run in phase of the full array. Resistive MHD simulations of the corona and core dynamics in the r-\theta plane are conducted to aid study of the mass and current distributions prior to array implosion. Simulations are conducted with the Alegra-MHD code. Initial conditions for the 2-D simulations are inferred from high resolution, 1-D, wire initiation/breakdown simulations. Solutions to 1-D, resistive-MHD equilibria, suggest that the asymptotic flow velocity v_max is related to the mass flux from the wire core \Gamma_0 and the maximum field strength B_0 via v_max = B_0^2/8\pi\Gamma_0. Comparison of the theory to the simulations and detailed discussion of the coronal evolution will be presented.

[LP1.094] Tungsten wire number scan on the Z accelerator: experiments and simulations

The standard 20-mm tungsten wire array configuration that is currently used on the Z accelerator is a design that was developed to optimize radiated power based on height and mass only. Over the years, a number of experiments have been conducted which indicate that wire array power should also be optimized based on wire number. In particular, experimental results show that higher number wire arrays provide increasingly higher powers for broadband emissions. A number of effects may together influence this observed trend and focus on the wire initiation/pre-acceleration evolution. In particular, the ability to initiate the wires themselves, the amount of plasma precursor produced, and the global magnetic field structure, are all dependent to some degree on the wire number. To this end, a recent experimental campaign was carried out to investigate the effect of wire number on a 20-mm diameter, 10-mm high tungsten wire array, keeping the array mass at ~ 5.8 mg. Wire numbers ranged from 50 to 500 wires, corresponding to wire diameters between 8.8 \mum and 27 \mum, and interwire gap spacings of 1.23 mm down to 0.12 mm. Numerical simulations investigating the role of wire number for this configuration are in progress and will be presented along with a summary of the experimental results.

[LP1.095] Wavelet Denoising of and X-Ray Power Extraction from Bolometry Data from fast Z Pinches

Bolometry data consists of X-ray energy vs time as released during a Z-pinch implosion. It is typically quite noisy and hence the extraction of power vs time is challenging (since it involves differentiating a noisy signal). Wavelet analysis can overcome these difficulties by nonlinear thresholding techniques. Both largest coefficient and level thresholding techniques are demonstrated on X-ray energy signals which denoise the data and allow power extraction.

Mathematica notebooks dedictated to the performance of these tasks will be demonstrated and various Sandia Z machine bolometry data from single and double shell implosions analyzed using these new techniques. When the data is excessively noisy, the successive application of mild low pass filtering and then wavelet largest coefficients thresholding works best. Future applications of these tools to radiography data analysis and combined low order Legendre polynomial and wavelet analysis will also be discussed.

[LP1.096] Characteristics of Molybdenum Plasmas Created on the Z-Accelerator at Sandia National Laboratory

Recent experiments on the Z Accelerator have used molybdenum wires in imploding arrays to create hot dense plasmas that efficiently radiate multi-keV energy x-rays. These molybdenum (Z=42) plasmas create a broad spectrum of x-rays ranging from 2.3 to 3.5 keV when stripped to the sodium-, neon- and fluorine-like states. Spectrscopic measurements of these x-rays will be presented, including measurements of a low energy continuum in the range of 1 to 2 keV, representative of 0.4 keV electron temperature. This data as well as other soft x-ray measurements will be used to generate synthetic spectra, from which determination of plasma temperatures and densities will be made. These analyses will be presented.

[LP1.097] Comparative 2D Radiation MHD Simulations of Argon Gas Puff Z-pinch Plasma Experiments on the Sandia Z Machine Using the Radiative Diffusion and CRE Transport Models

The recent development of the computationally efficient tabulated collisional radiative equilibrium (TCRE) radiation transport model(J.W. Thornhill, J.P. Apruzese, J. Davis, R.W. Clark, A.L. Velikovich, J.L. Giuliani, Jr., Y.K. Chong, K.G. Whitney, C. Deeney, C.A. Coverdale and F.L. Cochran, Phys. Plasmas 7, 3480 (2001).) has made possible full multidimensional radiation MHD simulations of hot dense Z-pinch plasmas with a realistic description of the non-LTE ionization dynamics and radiation transport physics. In this study, we focus on the implementation of the TCRE radiation transport model in the Mach2 2D radiation MHD code. An application of the model is made through a full dynamical simulation of an argon gas puff pinch driven by a circuit model of the Z generator. An analysis of the simulation, in particular, the K- and L-shell radiation yields, as well as the spectral and spatial characteristics of the radiation will be presented. In addition, a comparison of this multidimensional transport method will be made with the existing radiative diffusion model.

[LP1.098] Numerical Simulation of MHD instabilities in classical and twisted z-pinch plasmas

The spatial uniformity and therefore performance of the z-pinch is known to be determined by the development of the turbulent structures. We present the results of numerical MHD simulations in the cylindrical coordinates for the magnetically driven plasma instabilities in classical and twisted z-pinch configurations. The simulated picture of plasma near the stagnation phase is compared with the x-ray pinhole images of z-pinch driven by the 20 MA, 100 ns rise-time Z-accelerator at Sandia National Laboratories^1. The role of stabilizing axial component of the magnetic field in twisted pinch, zippering effect, and overall spatial non-uniformity will be discussed.

^1Chandler et al, Bull. Am. Phys. Soc., Vol.45 (2000).

[LP1.099] Dynamic Hohlraum Z-Pinches

Experiments on the Sandia Z machine have produced peak radiation temperatures of about 250 eV inside an on-axis foam cylinder target. This ``dynamic hohlraum'' (DH) is driven by either a nested or single tungsten wire array z-pinch striking the target and then compressing it. Surprisingly, such targets have resulted in radial (off-axis) radiation pulsewidth reductions of about a factor of two (nested arrays) or three (single arrays) when compared to identical implosions without a target, indicating substantially reduced effects from the expected magnetically driven Rayleigh-Taylor instabilities. On-axis (axial) radiation output from the DH through a 2.4 mm diameter radiation exit hole (REH) peaks at about 10 TW and is remarkably reproducible in pulseshape and especially in the rise of the pulse. Recent efforts at Los Alamos in simulating this high-performance z-pinch application will be shown with comparisons to other code results. In addition, some novel possible applications of this source will be shown.

[LP1.100] Spectral Analysis of Tracer Emission and Absorption Lines in Z-Pinch Dynamic Hohlraum Experiments

Spectroscopic analysis of tracer absorption and emission lines is reported for a nested tungsten (W) wire array z-pinch containing an on-axis CH foam with an embedded, thin Al-Mg tracer. The absorption and emission of radiation from the W plasma, the tracer, and the foam are studied with the goal of determining plasma parameters, such as electron temperatures, W composition along the spectrometer (approximately on-axis) line-of-sight, optical depths, and effects of photoionization on atomic populations and spectra. Collisional-radiative calculations are performed using the SPECT3D and NLTERT spectral analysis codes. We will present results of our simulations and comparisons with experimental data.

[LP1.101] Magnetic Diffusion in Aluminum and Copper on the Z-Accelerator

Recent experiments on the Z-Accelerator at Sandia National Laboratories were performed to characterize magnetic diffusion in aluminum and copper at currents of 2-4MA/cm and pulse widths of 200ns. Low, medium, and high pressure shots were taken in each material to establish a magnetic diffusion versus current scaling curve and to characterize relative propagation times between the pressure wave front and the onset of magnetic field diffusion at varying depths in the material. Principal diagnostics include unique B-dot probes and VISAR laser interferometry. The diagnostics were fielded on square short circuit loads designed to produce a symmetric current on each of the diagnostic surfaces. Linear current values were scaled by varying the anode surface area and charge voltage of the machine. Target pressure values of 500kbar, 1.0Mbar, and 2.0Mbar were achieved in this manner.

As the current ramps up in the load a pressure wave is launched into the conductor and magnetic and thermal diffusion waves are also initiated in the conductor. The pressure wave front compresses the material to densities greater than ambient solid resulting in dissipative energy, and magnetic field diffusion produces ohmic heating which is also dissipative. As larger accelerators are designed and built it is necessary to understand the losses in the conductors and to determine which conductors make the best construction materials. Additionally, recent experiments on the Z-Machine have demonstrated the ability to perform isentropic compression experiments (ICE) which require clear time separation between the pressure wave and the magnetic diffusion front for unambiguous data acquisition. These magnetic diffusion experiments document that time separation and also demonstrate a significant difference in rate of magnetic field diffusion between aluminum and copper.

[LP1.102] Mach2 Simulations of Kr Shell Deuterium Target Z-Pinches

Mach2 Simulations of Kr shell onto Deuterium target Z-pinches were performed and are compared to UC Irvine's experimental results of such load configurations. The two load configurations tested were: a 4~cm dia.\ Kr hollow gas shell liner with either a 1.7~cm dia.\ D_2 hollow gas shell or solid gas jet target, with axial magnetic fields ranging from 0 to 2~kG.

[LP1.103] Deuterium fiber extrusion and handling system for neutron production experiment

A frozen D_2 fiber fragment extrusion and handling system has been developed at AFRL to provide a central target for a wire array implosion on SNL's Z machine. The system, though, can be modified for use in Magnetized Target Fusion research. As presently configured, it extrudes a 0.5~mm diameter fiber, cuts the fiber to a length of 7~cm, and drops the fiber fragment into an LN_2 refrigerated support structure where the fiber remains intact for about 7~minutes. A heavy hydraulically actuated blast shutter protects the extrusion system after the fragment is dropped. Design and performance information, including detailed images of the fiber during the various phases of operation, will be provided.

[LP1.104] Simple Estimation of Material Thickness Ablated by Intense X-ray with Spectrum and Energy Level of Sandia Pinch Source

Before we design advanced version of IFE reactors, lots of technical data must be gathered. One of the important data is the thickness of reactor chamber wall material by intense X-ray which is produced by IFE target implosion. So that, we showed the results of the ablated thicknes by X-ray associated with a typical laser target implosion in the former meeting. Only the numerical estimation was possible for the high total energy of X-ray similar to the reactor level. On the contrary, we present here the estimation of the ablated thickness by X-ray associated with typical wire-arrayed pinch discharges at Sandia National Laboratories. The energy spectrum obeys the Plank's law. Various temperature of the black body radiaion and various total energy of X-ray were assumed which were very realistic with the recent experiments at Sandia National Laboratories. The first candidate material was solid carbon, and the surface ablation thickness was within the micron range with the total X-ray energy within the MJ range. Our numerical results will be compared with the results of our scheduled future experiments under the US-Japan collaboration Project.

[LP1.105] Characterization of Hydrogen Desorption from Rapidly-Heated Metal Foils and Wires

Experiments have been conducted to characterize the parameters of hydrogen plasma and gas produced from rapidly heated metal foils and wires. Several metals with important applications in pulsed power systems have been studied, including tungsten, titanium, and tantalum. Both surface hydrogen and interstitial hydrogen from electrochemically doped and undoped samples have been investigated. The samples were rapidly heated using a focused 600 mJ, 25ns KrF (248nm) laser. The laser's energy and beam size were varied to produce fluences over a range of 1-10 J/cm^2. Various plasma diagnostic techniques, including intensified, gated CCD optical emission spectroscopy, dye-laser-resonance-absorption photography, and interferometry, have been used to measure electron temperatures of 1-2 eV and electron densities of 1x10^16 cm^-3. Streaming velocities for hydrogen are in the range of 1-2x10^6 cm/s, while streaming velocities for the metals are slower, 0.8-1.7x10^6 cm/s, scaling with their mass and laser fluence. Comparisons are being made with LSP code results and plasma plume expansion models. These types of hydrogen desorption plasmas may play an important role during the initiation phase of wire Z-pinch experiments.

*This research is supported by Sandia National Laboratories subcontract BE-6069 to the University of Michigan

[LP1.106] ZPMHD-a Rep-rated Z-Pinch Power Plant Direct Conversion Concept*

We have performed a preliminary conceptual study of ZPMHD, a z-p