Session FP1 - Poster Session III.
POSTER session, Tuesday morning, October 30
Exhibit Hall B,

[FP1.001] Reversed Field Pinch Experiments

[FP1.002] MST Progress and Plans

A variety of confinement physics results have accrued in MST. Improvement in programmed inductive current drive yields about a nine-fold increase in energy confinement (relative to standard RFP plasmas), thermal diffusivity ~ 5 m2/sec, beta ~ 14achieved. Fast electrons are confined up to 100 keV, implying a transition to well-formed magnetic surfaces. Core magnetic fluctuations are detected with FIR polarimetry, the first measure of core magnetic fluctuation reduction with current profile modification. Experiments without reversal (to remove the m = 0 mode which nonlinearly couples to m = 1 modes) reveal the importance of mode-coupling to sawtooth relaxation, ion heating, and momentum transport. These studies are aided by new measurements (magnetic field by motional Stark effect and polarimetry, ion temperature by CHERS and Rutherford scattering, electric field by heavy ion beam probe). With reduced Ohmic input power (improved confinement), auxiliary current drive and heating is now feasible in the RFP. Auxiliary systems being implemented in MST include neutral beam injection (feasibility tests underway), electron Bernstein wave injection (encouraged by emission of thermal levels of radiation), lower hybrid wave injection (low power studies underway), and oscillating field current drive (current penetration measurements underway).

Work supported by U.S.D.O.E.

[FP1.003] Improved confinement in the RFP and the role of the magnetic turbulence spectrum.

Control of magnetic turbulent transport in the reversed field pinch yields increased energy confinement, plasma temperature, and beta. Relative to standard toroidal induction with incumbent dynamo relaxation, added poloidal current drive in MST increases the energy confinement time nine-fold to 9 ms, increases beta from 9% to 14%, and permits electrons to exceed 1 keV despite decreased Ohmic heating, a clear demonstration of reduced transport. The electron heat diffusivity drops to \sim5 m^2/s, comparable to typical tokamak plasma values. Central to these improvements is a broad spectral reduction of tearing fluctuations associated with magnetic relaxation and dynamo, implying reduced magnetic stochasticity. The role of particular spectral features in determining transport will be emphasized. For example, the reduction of poloidal number m=1 modes resonant in the middle to outer region of the plasma is crucial to realize the best improved confinement.

[FP1.004] Fast Particle Confinement in MST

Measurements of the fast particle confinement in improved confinement MST dicharges demonstrate well defined flux surfaces and low fast particle diffusion. The hard X-ray flux spectrum (10-200 KeV) has been measured using a CdZnTe detector with up to 1 ms time resolution and \approx 5 KeV energy resolution during standard and improved confinement plasmas. The measured spectra give direct evidence for the existence of well defined flux surfaces with connection lengths (L_c > 10^4 m). Simulations of electron transport during improved confinement plasmas using the Fokker-Planck code CQL3D suggests diffusion of fast electrons is lower than Rechester-Rosenbluth scaling (<< v_th_i B_r^2) since the simulations require a low diffusion coefficient to predict the measured hard X-ray flux. This is also supported by measurements of fast ion (>10 KeV) confinement using a neutral beam and particle analyzer which show that diffusion is also lower than Rechester-Rosenbluth. density

[FP1.005] Time Evolution of Measured Energy and Particle Transport in the MST Reversed-Field Pinch

Time evolved measurements of thermodynamic profiles have been obtained in a variety of MST discharges (PPCD, F=-0.22, F=0, F=+0.02, F=+0.03), leading to the first measurement of radially resolved, time evolving heat transport in the MST. M=0 modes are absent in F=0 plasmas, and confinement is observed to improve, but degrades rapidly as F is raised above zero. In all cases, the heat flux is predominantly conductive over the majority of the plasma volume, though convective heat transport becomes significant in the edge. The observed heat and particle fluxes cannot be described by a diagonal transport matrix. However, including pressure gradient and electric field cross-terms can account for the observed fluxes. The radial electric field is calculated from ion momemtum balance and compared to measurements from a heavy-ion beam probe diagnostic. This work was supported by the U.S.D.O.E.

[FP1.006] Studies of Plasma Resistivity Through Measurements of Parallel Current Density and Electric Field in the MST Reversed Field Pinch

A two dimensional, toroidal equilibrium reconstruction code has been developed for the reversed field pinch. The parallel current density profile has been measured by incorporating several diagnostics into the code. A finite difference technique applied to a time series of smoothly varying equilibria determines the parallel inductive electric field profile. In the case of a rapidly changing equilibrium, realized in the Pulsed Poloidal Current Drive experiments on MST, the finite difference method is not ideal. A new technique, which solves the time derivative of the Grad-Shafranov equation constrained by the time derivatives of magnetic signals, is used to find the parallel electric field. During periods of low MHD activity, Ohm's law is expected to obey its simplest form and the ratio of E_||/J_|| determines the resistivity profile. These profiles are roughly consistent with the Spitzer and neoclassical modeled resistivity profiles based on electron temperature and upper bound Z_eff measurements. A method of determining the Ohmic power deposition profile based on Poynting flux has been developed for heat transport analyses. This is advantageous as it is independent of the very uncertain Z_eff profile and does not rely on Ohm's law to compute the deposition.

Work supported by USDOE.

[FP1.007] Condensation of the m=1 MHD mode toroidal spectrum in MST

Reversed Field Pinch (RFP) plasmas where the spectrum of m=1 modes condenses around an individual (m=1,n=n_0) mode, with n_0 \approx q(0)^-1, have been observed in the Madison Symmetric Torus (MST) device. This paper presents a survey of the most relevant features of these Quasi Single Helicity (QSH) plasmas. The ranges of global plasma parameters that make more likely the access to these plasmas will be discussed. A section of the paper is dedicated to the description of the behaviour of high frequency magnetic and kinetic fluctuations in QSH plasmas. A significant modification of thier properties is observed.

[FP1.008] Time-resolved Motional Stark Effect Measurements of Magnetic Field on the MST RFP

A spectral motional Stark effect (MSE) diagnostic is now in regular operation on the MST RFP. This is the first time MSE has been applied to measure the magnitude of the magnetic field in the core of a low-field (0.2 to 0.5 T) magnetic confinement device, an accomplishment made possible by a high quality diagnostic neutral beam and a carefully designed beam emission collection and detection system. Measurement of the core magnetic field provides a strong constraint for equilibrium reconstruction in MST. The diagnostic neutral hydrogen beam is short pulse (3 ms), intense (4 A and 0.4 A/cm^2), mono-energetic, and low-divergence. MSE measurements are made by recording the Doppler-shifted H-alpha Stark spectrum emitted by the beam with an imaging spectrometer and CCD camera. Signal-to-noise is sufficient to allow single-shot exposures of less than 100 microseconds using a ferroelectric liquid crystal shutter. An array of shutters will provide seven sequential exposures during a single neutral beam pulse to measure the evolution of on-axis magnetic field during fast equilibrium changes such as sawtooth crashes.

[FP1.009] Polarimetric Measurement of Equilibrium Magnetic Field and Current Profile on the Madison Symmetric Torus

The upgraded far-infrared polarimeter-interferometer system installed on the Madison Symmetric Torus (MST) enables measurement of either the line-integrated Faraday rotation or electron density profiles on a 1 \mus timescale. These quantities provide radial profiles of the poloidal magnetic field, plasma density, and current density essential for toroidal equilibrium reconstructions. Also, fast changes to the line integrated Faraday rotation signal due to the density and the poloidal field can be separated. Results will be presented using MSTFIT to show the evolution of the plasma profiles over a sawtooth crash, Pulsed Poloidal Current Drive (PPCD) and Oscillating Poloidal Current Drive (OPCD). The current profile is observed to peak during the application of PPCD and to flatten after a sawtooth crash. The diagnostic is currently being upgraded by the addition of a third FIR laser to permit simultaneous measurement of Faraday rotation and density. The status of this upgrade will be discussed.

Work supported by U.S.D.O.E.

[FP1.010] Internal Magnetic Field Fluctuation Measurement in the MST Reversed-Field Pinch

Internal magnetic field fluctuations and equilibrium poloidal magnetic field have been measured in the MST reversal field pinch by a 11 chord far-infrared polarimeter- interferometer system with frequency response up to 1 MHz. Fast time resolution and low phase noise of the polarimeter enable us to resolve m=1 resistive tearing modes as a precursor to the sawtooth crash. Turbulent magnetic field fluctuations up to 100 kHz have also been observed. The chord-averaged radial magnetic field fluctuation level is about 33 G or 1and is dominated by m=1 magnetic field fluctuations. By computing the coherence between two toroidally-displaced chords, one can determine the toroidal mode number and rotation speed. The phase of radial magnetic field fluctuations lags poloidal magnetic field fluctuations by 90 degrees which is in agreement with MHD computation for MST. Magnetic field fluctuations are reduced by a factor of four during a high confinement PPCD discharges, consistent with energy confinement improvement. Work Supported by U.S. DOE

[FP1.011] Non-reversed Discharges - A Tool for Understanding Reversed Field Pinch Physics

The reversed field pinch (RFP) configuration generally supports a full spectrum of coupled magnetic fluctuations. Three-wave interactions between two m=1 modes and an m=0 mode can have a strong influence on plasma rotation and current profile evolution. By removing m=0 modes from the plasma we perform a critical test of our understanding of their role in RFP physics. In MST, we remove m=0 modes by excluding their resonant surface through edge toroidal field programming. With m=0 modes suppressed, the relaxation of the flow profile and current profile during sawteeth is greatly reduced. Anomolous ion heating is also reduced and confinement is not as good as in standard reversed discharges. A common theme emerges which links many of the global relaxations in the RFP to the extensive mode coupling mediated by m=0 modes.

Work supported by U.S.D.O.E.

[FP1.012] Nonlinear coupling of magnetic fluctuations in MST

The nonlinear coupling of magnetic fluctuations in the reversed field pinch (RFP) is not fully understood. Large-scale magnetic fluctuations have been measured in the MST RFP using 64 edge magnetic coils in a toroidal array. Three-wave nonlinear mode coupling is measured through a sawtooth cycle using bi-spectral analysis. The bi-coherence (three-wave nonlinear coupling) between two m=1 modes and an m=0 mode is small between sawtooth crashes but becomes large during sawtooth crashes. Coupling between m=0 modes also increases at the crashes. Experimentally, we suppress m=0 modes by excluding their resonant surface through edge toroidal magnetic field manipulation. In these discharges, the bi-coherence between m=0 and m=1 modes appears to be similar to that in the standard case.

Work supported by U.S.D.O.E

[FP1.013] Ion Heating and Energy Transport on MST

Ion temperatures are often measured to be higher in Reversed Field Pinch (RFP) plasmas than can be accounted for by energy transfer from ohmically-heated electrons. Recently the Madison Symmetric Torus (MST) has been reconfigured to allow the time, space, mass, and charge dependences of this excess heating to be sorted out. Shot-to-shot temperature profiles are acquired, for the bulk majority (Deuterium) ions by Rutherford Scattering (RS), and for impurities (Carbon) by Charge Exchange Recombination Spectroscopy (CHERS). Anomalous ion heating occurs at Magnetic Reconnection Events (MREs). The relative importance of anomalous ion heating to the ion power balance is determined by comparison of measurements in standard and Pulsed Poloidal Current Drive (PPCD) discharges, which respectively do and do not experience MREs. [EOB]

[FP1.014] Neon gas injection experiments into the MST RFP plasmas

On the MST reversed-field pinch (RFP) device, the neon gas injection experiment into the moderate electron density, 250 kA standard plasmas, has been conducted. Both radiated and Ohmic input power increase with neon gas injection. The electron density increases, and the profile becomes hollow. However, the electron temperature decreases, and the profile flattens. Thus, the electron pressure profile flattens. As a result, Beta poloidal becomes similar to that in plasmas without neon injection. However, energy confinement time decreases due to the rise of the Ohmic input power. Neon gas also has been injected into the sawtooth free pulsed poloidal current drive plasmas. In this case also the electron temperature decreases. We have compared neon gas injection results on MST to those on TPE-RX at AIST. The location of the radiative neon ions is mainly distributed in the plasma edge. The overall plasma performance with gas injection is similar to the MST cases. Work supported by U.S. D.O.E.

[FP1.015] Non-Linear Analysis of Mode Locking in MST

A toroidally localized magnetic perturbation can form in the MST reverse field pinch just inside the reversal layer. Such perturbations have been shown to drift around the torus, following magnetic field lines in the direction of the ion diamagnetic drift velocity. We have determined that the mode trajectory follows the Sine-Gordon (SG) equation, i.e., it is a kink-type soliton. This represents the rotation angle around the minor axis that the mode makes in moving around the torus. The SG equation is obtained by summing the torques around the minor axis that are acting upon the island. Within the kink structure, local currents exist that can create a magnetic moment, which allows the kink to experience a potential well along the field lines proportional to grad B. When the center of the kink passes the poloidal gap, where the grad B force along the field lines is minimal, the radial magnetic field produced at the gap can change the force significantly, transforming the kink into a “breather” soliton and the mode locks.

[FP1.016] Measurement of Electrostatic Fluctuation Profiles on MST with a Heavy Ion Beam Probe

The core profile (r/a\approx0.27\sim0.72) of the electrostatic potential fluctuations and electron density fluctuations have been measured for the first time in the MST reversed field pinch with a heavy ion beam probe (HIBP). Traditional fluctuation measurements with Langmuir probes have been limited to the edge plasma region (r/a>0.8) at low current. The HIBP has been used to extend the MST measurements to the core region of the plasma and cover broader range of plasma parameters. The measured \tilde\phi is \sim30V (\tilde\phi/Te \sim 10-15%) for a standard 380kA discharge, while \tilden/n \sim 10-15%. The measured power spectra of both \tilde\phi and \tilden/n show a peak at the tearing mode frequency. Broadband fluctuations (>30kHz) are also found at frequencies higher than the core resonant tearing modes and their relationship is studied with bi-spectral analysis. Simultaneous measurements have been made at two sample locations, thus allowing us to estimate electrostatic fluctuations induced transport. The 2 sample volumes are nearly radially aligned and therefore do not provide information about k_pol. Correleration between the measured \tilde\phi from HIBP and \tildev from Ion Dynamic Spectroscopy (IDS) will also be discussed.

[FP1.017] The Measurement of the Equilibrium Radial Electric Field in MST-RFP

Measurements of the radial electric field (E_r) are being made for the first time in the core (0.3 < r/a < 0.7) of an RFP plasma. E_r has a magnitude of approximately 2.0-2.5 kV/m in a 380 kA discharge as measured by a Heavy Ion Beam Probe (HIBP). The electric field tends to increase over the period of a sawtooth cycle, peaking shortly before a crash. It is outward pointing which is consistent with stochastic magnetic fields. The measured E_r is close to the ExB rotation inferred from the n=6 mode rotation and estimates on poloidal rotation in the core. A key element of the measurements is the accurate determination of the sample locations within the plasma. This is accomplished by combining a particle trajectory code with magnetic field profiles generated by MSTfit code. Investigations of the current profile evolution as generated by MSTfit both with and without HIBP as a constraint have been studied. The electric field profiles of low current standard and edge biased electrode discharges will also be presented.

[FP1.018] Plasma Response to Oscillating Poloidal Current Drive in the Madison Symmetric Torus

A 1 MVA 500 Hz oscillator has been installed in the toroidal magnetic field circuit of the Madison Symmetric Torus (MST) Reversed Field Pinch. This device drives an alternating poloidal current in the edge of the plasma affecting the current profile and thus both the spectrum of the tearing mode fluctuations

and energy confinement. We find the amplitude of the dominant (m=1) modes rises and falls with the oscillating edge current while the m=0

mode amplitude appears to do just the opposite. Simultaneous with the decreasing m=1 amplitude is a rise in the electron temperature in the core. These effects and an entrainment of the sawtooth instabilities with the applied oscillation are shown. This oscillator will operate with a similar device installed in the poloidal magnetic field circuit to test Oscillating Field Current Drive (OFCD). Initial OFCD results may be reported.

[FP1.019] Penetration of Oscillating Magnetic Fields into a Reversed Field Pinch

An AC magnetic field is applied at the shell of the Madison Symmetric Torus (MST), and several methods are used to study the resulting field penetration into the plasma. The applied field is used in oscillating poloidal current drive (OPCD). Motional Stark effect (MSE) spectroscopy and 11-chord far-infrared (FIR) interferometry/polarimetry show that the oscillations can reach the magnetic axis much more quickly than a resistive diffusion time. An edge internal magnetic probe array shows that field penetration is related to background magnetic relaxation phenomena in MST. For example, the otherwise quasi-periodic sawtooth cycle is entrained onto the OPCD cycle. Both the phase lag and amplitude decay for penetrating fields are much less during an OPCD half-cycle that includes a sawtooth crash than during one that does not. The experimental results are compared with 1-D and 3-D MHD theory. Similar work with oscillating field current drive (OFCD) is planned.

[FP1.020] ELECTRON BERNSTEIN WAVE EXPERIMENT IN AN OVERDENSE REVERSED FIELD PINCH PLASMA

Experiments on the Madison Symmetric Torus (MST), a reversed field pinch operating at the University of Wisconsin, have observed blackbody emission in the ECRF from plasmas with ømega_pe\sim 3 ømega_ce. We speculate the emission is due to either OXB or XB mode conversion of electrostatic electron Bernstein waves (EBW) to electromagnetic waves at the plasma edge. The observation opens up possibilities for ECE diagnosis of the electron temperature, electron heating and current drive via the EBW, and potentially as a magnetic field diagnostic for the RFP. The present experiments have focused on emission and low power (several watt) coupling measurements of X mode for frequencies between 2.6 GHz and 3.9 GHz. An S-band waveguide horn attached to a directional coupler to monitor the forward and reflected power and the phase between them. These measurements have been used to calculate the surface impedence and reflection coefficient. Coupling results will be compared to a theoretical model. An 8 channel radiometer has been used and compared to the experiemental coupling measurements. These results show that EBW may be a feasible wave for heating overdense plasma. The work was supported bu U.S. DOE.

[FP1.021] A test of EBW heating and current drive on MST

Electron Bernstein waves (EBWs) show promise for providing localized heating and current drive in overdense plasmas such as RFPs and STs. Edge localized noninductive current drive may reduce the magnitude of the magnetic fluctuations and improve the confinement of the RFP. A medium power experiment (150 kW) experiment is underway using a TWT amplifier in the 3.1-3.6 GHz band. At these powers, numerical modeling has shown that detectable X-ray emission from fast electrons might be attained. This poster reviews the results of ray tracing and Fokker-Plank modeling of EBW current drive, extending previous calculations to include the effects of electron transport, and an examination of the scaling of current drive efficiency with power. Finally, initial results from the experiment will be presented.

[FP1.022] Tests of Interdigital Line Antennas for Launching LH Waves in MST

RF current drive has been proposed as a method for reducing the tearing fluctuations that are responsible for anomalous energy transport in the RFP. A system for launching lower hybrid slow waves at 800 MHz and n_||= 7.5 has been designed and implemented in MST. The antenna is an enclosed interdigital line using \lambda/4 resonators with an opening in the cavity through which the wave is coupled to the plasma. The power limit at which the first antenna reflects nearly all the input power is \sim3 kW. A boron nitride limiter assembly was added to the face of the antenna and resulted in only a slight increase in the observed power limit. The antenna shows no signs of damage after 16 months of operation in MST. A new antenna with design improvements to the vacuum feedthroughs and better impedance matching is being implemented.

[FP1.023] Active field error correction at the MST poloidal gap

Magnetic field errors at MST's poloidal gap can lead to increased plasma-wall interaction and "locking" of magnetic fluctuations to the wall, both of which can degrade MST discharges. The present pre-programmed compensation scheme can only partially correct these field errors. We are building an active correction system to suppress radial field errors at the poloidal gap. The gap radial field is monitored by 32 coils located at the poloidal gap inside MST's conducting shell. An existing analog summing circuit resolves poloidal mode signals from these coils; a similar circuit will map these mode-resolved signals to drive signals for up to 36 high-power IGBT switching amplifiers. The amplifiers will drive feedback coils wound through existing holes in the poloidal flange to cancel the radial fields sensed by the internal sense coils. We show results of initial feedback tests using a single drive coil and describe the design and expected performance of the full feedback system, currently under construction.

[FP1.024] Modeling Single Particle Transport in Stochastic Magnetic Fields

Single particle transport in a stochastic magnetic field is simulated via code ION and RIO. Developed in collaboration with a group in Novosibirsk, Russia, they simulate both single ion and multiple ion trajectories in a stochastic magnetic field. A sharp decrease in the relative diffusion of ions to magnetic field lines is seen in two gyro-radii regimes. One is explainable from the unbroken flux surfaces near the edge of the plasma. The other is thought to be due to a "gyro-averaging" effect that occurs when the gyro-radius exceeds the radial correlation length of the field lines. The simulations indicate a decrease in expected transport, most strongly as a function of gyro-radius, which will be tested experimentally with the MST neutral beam injector.

[FP1.025] 1.5 MW Neutral Beam Injector for MST Reversed Field Pinch

A high power injector of hydrogen atoms with total current of 60 A and energy of 25 keV has been manufactured by the Budker Institute of Nuclear Physics, Novosibirsk, Russia, for installation on the MST Reversed Field Pinch. The two goals of the beam injection experiment are: (1) to evaluate the confinement of fast ions in the presence of stochastic magnetic field, and (2) to determine the efficiency of beam power deposition into plasma. The beam duration is 1.5 ms, which is sufficient for the set goals. The beam will be injected tangentially into the plasma and the density and energy spectra of the fast ions will be measured with an NPA via analysis of CX spectra. The plasma (mostly electron) heating efficiency will be determined with the Thomson scattering diagnostic. The studies will allow us to formulate the requirements for a full scale, long pulse neutral injection system for plasma heating and beta-limit studies.

[FP1.026] Detailed analysis of m=1 coherent structure in RFX

The onset of Quasi Single Helicity states in the RFX reversed field pinch (RFP) device is accompanied with the formation of a helical coherent structure in the plasma core. This structure, whose pitch corresponds to that of the dominant m=1 instability, is imaged through soft x-ray tomography. The presence of the structure is related to the formation of closed magnetic flux surfaces in a region where initially there is a stochastic magnetic field. The coherent structure emerges also in cases where the total energy of the m=1 modes does not decrease and an enhanced level of islands overlapping would be expected. This aspect will be discussed from an experimental point of view. Helical structure with and without a separatrix are observed (the separatrix is defined assuming SXR images being representative of magnetic flux surfaces). The conditions which control the presence of the separatrix will be discussed, as the dynamical evolution of the helical structure topology. The influence of the secondary (i.e. all but the dominant) m=1 modes appears to be important in this respect.

[FP1.027] Various operating regimes in a reversed-field pinch, TPE-RX.

We will report the present status of TPE-RX, large reversed field pinch machine with R/a = 1.72/0.45 m. We have been exploring several operating conditions in addition to standard operation, and obtained Improved confinement states. Pulsed poloidal current drive (PPCD) experiment is one of them. The poloidal beta and the energy confinement time increase under a single-pulse PPCD experiments. In gas-puffing experiments, these confinement parameters increase as I/N (I is the plasma current, and N is line averaged electron density) decreases. We have also tried to rotate the toroidally localized magnetic structure called locked mode by means of an additional external rotating toroidal field. The same system is also used for a multi-pulse PPCD experiment. On the other hand, spatial profile of the electron density is speculated from the dual-cord interferometer system, and spatial profile of the radiation power loss is speculated from a bolometer system. Soft X-ray spectroscopy u! sing multilayer mirrors has been prepared in order to speculate electron temperature. We also report these diagnostics evaluating the confinement properties.

[FP1.028] Plasma Temperature Constraints Inherent to Electrode Current Drive

In electrostatic magnetic helicity injection, a current is driven between electrodes, parallel to the magnetic field in the edge plasma of a device. Plasma instabilities have the effect of distributing current throughout the plasma, thereby providing the desired toroidal current. In an electrode driven plasma, all parts of the plasma with sustained parallel currents are connected by parallel thermal transport to the electrodes. The effects of resistive plasma heating and parallel thermal and advective energy transport were studied with models ranging from the short mean-free path (MFP) classical Ohm’s law including Hall terms to a long MFP kinetic model. The advection of hot electrons into the anode is considered, even in the short MFP case. A robust relationship is demonstrated, whereby the peak electron temperature (in units of eV) is limited to being no more than 25% to 40% of the electrode voltage (in units of V). The addition of bremsstrahlung and line radiation to the model would further reduce the temperature-to-voltage ratio.

[FP1.029] Kinetic Simulation and Transport Modeling

[FP1.030] Hybrid MHD-Gyrokinetic Simulations of the Nonlinear Saturation of Energetic Particle Modes in Tokamaks

The nonlinear dynamics of shear-Alfvén modes in Tokamaks has been investigated by using the Hybrid MHD-Gyrokinetic simulation Code (HMGC). Toroidal Alfvén Eigenmodes (TAE's) have been shown to exist for low values of the energetic-ion pressure gradient, \beta_H'. Above a certain threshold value, the Energetic Particle continuum Mode (EPM) is strongly destabilized. Nonlinear TAE saturation appears to be due to the trapping of resonant energetic ions in the potential well of the wave. Saturation of the EPM is instead associated to a radial redistribution of the energetic particles, with potentially dramatic consequences on \alpha-particle confinement. The radial fragmentation of EPM coherent eddies and the spontaneous excitation of zonal flows by drift-Alfvén turbulence is discussed. Also the EPM destabilization and its nonlinear evolution in the presence of a non-monotonic q profile is presented: an unstable EPM grows at the radial position where the resonant drive (\propto \beta_H') is maximum. A rapid fast-ion radial transport is caused by that mode. The mode then becomes localized in frequency near the TAE gap and in space at the position of the q-minimum surface. These findings are discussed in relationship to their possible consequences of fast ion transport.

[FP1.031] Kinetic electron effects on plasma microturbulences

Kinetic electron effects on plasma microturbulences are studied using full torus gyrokinetic particle simulations of toroidal ion temperature gradient and trapped electron mode instabilities. Electron dynamics is treated using a fluid-kinetic hybrid electron model based on an expansion of the electron response using the electron-ion mass ratio as a small parameter. The model accurately recovers low frequency plasma dielectric responses and faithfully preserves nonlinear kinetic effects (e.g., phase space trapping). Maximum numerical efficiency is achieved by overcoming the electron Courant condition and suppressing high frequency modes. Nonadiabatic response by trapped particles and resonant electrons can affect the generation of zonal flows which control the level of turbulence and transport. Electron transport driven by ion scale fluctuations will also be discussed.

[FP1.032] Electron Dynamics in a Global, Gyrokinetic, Particle-in-Cell Code

Electron dynamics in the electrostatic and finite \beta regimes in magnetized toroidal plasmas are known to be key ingredients for microinstabilities. However because the electron dynamics is characterized by frequencies (e.g. transit and bounce frequencies) much higher than those of the modes of interest [e.g. ion temperature gradient-driven (ITG) mode], new numerical challenges become apparent. We discuss the extension of the shearless slab splight-weight scheme~(I. Manuilskiy and W.W. Lee, Phys. Plasmas 7), 1381 (2000) to general toroidal geometry. Unlike recent studies, we adopt a fully kinetic approach; the model does keep the correct electron dynamics in the layer around k_|| \mapsto 0. We will present recent advances in the implementation of the split-weight scheme into a global toroidal code. Issues related to finite-\beta effects and the generalization of the split-weight scheme will also be discussed.

[FP1.033] Comparisons of Kinetic Electron Models in Gyrokinetic Particle Simulations

Comparisons of the hybrid [1] and split-weight [2,3] schemes for the electron dynamics in gyrokinetic particle simulations has been carried out. The emphasis here is to understand the regions of validity for these schemes as well as their numerical properties with regard to noise, time step and accuracy. The numerical issues for the implementation of these schemes in our 3D global gyrokinetic particle code (GTC) in general geometry [4] will also be discussed. Work supported by US DoE.

[1] Z. Lin and L. Chen, Phys. Plasmas <8>, 1447 (2001).

[2] I. Manuilskiy and W. W. Lee, Phys. Plasmas <7>, 1381 (2000).

[3] W. W. Lee, J. L. V. Lewandowski, T. S. Hahm and Z. Lin, Phys. Plasmas (to appear).

[4] Z. Lin, T. S. Hahm, W. W. Lee, W. M. Tang and R. White, Science <281>, 1835 (1998).

[FP1.034] A New Split-Weight Perturbative Scheme for Shear-Alfvén Dynamics

A split-weight perturbative scheme for finite-\beta gyrokinetic particle simulation based on the generalized Ohm's law has been used successfully to produce shear-Alfven waves.(W. W. Lee, J. L. V. Lewandowski, T. S. Hahm and Z. Lin, Phys. Plasmas (to appear).) The scheme depends crucially on the calculation of \psi \equiv \phi + (1/c) \int (\partial A_\parallel / \partial t) dx_\parallel. To extend the scheme for general geometry, a new split-weight scheme utilizing the parallel canonical momentum in the direction of the external magnetic field, P_\parallel \equiv p_\parallel + (q/c) A_\parallel, is developed, where p_\parallel is the mechanical momentum and A_\parallel is the vector potential. The resulting electron distribution then becomes F_e = (1 + e\psi / T_e) F_0e + \delta h_e, where \delta h_e can be followed dynamically using the usual perturbative gyrokinetic particle simulation methods.(W. W. Lee, J. Comput. Phys. 72), 243 (1987).^,(S. E. Parker and W. W. Lee, Phys. Fluids B 5), 77 (1993). The resulting field equations include the modified Poisson's equation and Ampere's law as well as the equations for \partial \phi / \partial t and for \partial A_\parallel / \partial t. The last two equation require higher order velocity moments for closure. The applications of the scheme for finite-\beta modified microinstabilities with microtearing will be reported.

[FP1.035] Results on Toroidal ITG Turbulence from Nonlinear Gyrokinetic Simulations

Previous scans(A.M Dimits et al., Plasma Physics and Controlled Nuclear Fusion Research (IAEA, Vienna, 1995) Vol.\ 3, p.~457.) of ion thermal transport vs.\ external toroidal velocity shear V'_tor have been extended to show that in some cases the ion thermal transport first decreases as a function of V'_tor and then increases back to or beyond its V'_tor =0 value. Our simulations with external velocity shear are compared with other work.(Waltz, Garbet, and Dewar, Phys. Fluids 5), 1784 (1998). Differences in the transport levels and in the velocity-shear values needed to quench transport are observed. The relative contributions of various terms to the driving and damping of the zonal flows are investigated using a new diagnostic that distinguishes linear from nonlinear contributions to the rates of change of zonal-flow modes. The results are compared with theoretical expectations. The diagnostic further separates the nonlinear contributions into the Reynolds stress and diamagnetic contributions as well as contributions of moments of the distribution function higher than the pressure. Issues in the implementation of an analogous diagnostic in a tokamak experiment are examined.

[FP1.036] Nonlinear gyrokinetic PIC calculations of microturbulence characteristics in DIII-D-like discharges

We have been performing nonlinear, toroidal, three-dimensional, global gyrokinetic particle-in-cell calculations with parameters and profiles appropriate for DIII-D discharges, save for the fact that a circular cross-section, the electrostatic approximation, and adiabatic electrons are used. We will therefore be concentrating on the characterization of ion microturbulence in such discharges, with particular emphasis on those with quiescent double barriers. The array of calculations expected to be performed will include ones with and without zonal flow, and with the experimentally measured radial electric field included as an external field in the equations of motions for the gyrokinetic ions. The effect of the inclusion of each of these in turn and all of these at once in these DIII-D-like discharges will be examined on the calculated radial correlation lengths, mode spectra and frequency spectra, as available. Comparisons of these calculated quantities with the experimentally measured ones, particularly by microwave reflectometry, will also be carried out in order to isolate the important ingredients that affect the turbulence characteristics.

[FP1.037] Kinetic electron effects on Ion-Temperature-Gradient driven turbulence and transport

The effects of fully kinetic electrons, including both trapped and passing electrons, are studied with a three-dimensional toroidal gyrokinetic simulation that uses field-line-following coordinates in a flux-tube and the split-weight scheme for the electrons [Y. Chen and S.E. Parker, Phys. Plasmas 8(5) 2095 (2001)]. Ion-ion collisions and electron-ion collisions will be included. Although electromagnetic effects are included, the plasma beta is limited to \beta m_i/m_e\leq 1 due to the requirement of resolving the magnetic skin depth. Linear studies show that the growth rate of the ITG mode is significantly increased from the result of simulations with adiabatic electrons, typically by a factor of 1.5\sim 2. This is mainly due to the nonadiabatic effects of trapped electrons. Nonlinear runs indicate that the nonlinear ion heat flux is increased from that of adiabatic electron model by a factor of \geq 5. Analysis of the turbulence data shows that zonal flows are not as dominant as in the adiabatic electron simulations. We will also study the effects of kinetic electrons on the linear threshold of the ITG instability and the nonlinear shift of critical instability drive in terms of transport level [A.M. Dimits et. al., Phys. Plasma 7(3), 969(2000)]

[FP1.038] Electromagnetic gyrokinetic simulations

Fully electromagnetic simulations of plasma turbulence using the GS2 code have revealed significant enhancements in the electron heat flux, compared with the electrostatic case. Preliminary analysis indicates this is a purely nonlinear effect resulting from the generation of Alfven waves. (Linear growth rates of ion temperature-gradient modes and trapped electron modes are slightly reduced by the electromagnetic terms.) Conditions under which this effect occurs are explored, in particular its dependence on electron temperature gradient. Whether it bears on experimental observations of core electron thermal transport with small temperature gradient is investigated.

[FP1.039] Coarse-grained steady state in the collisionless ion temperature gradient driven turbulence

In studies of the anomalous transport in high temperature plasmas, the collisionless kinetic eauqtion of the velocity distribution function has been employed as the `first principle'. To find the steady and irreversible transport in the collisionless turbulence, one needs to consider a coarse-grained state of the distribution function that consists of a large-scale profile with fine-scale fluctuations. Development of the fine-scale fluctuations caused by the phase-mixing corresponds to growth of higher-moments, while lower-moments are kept constant. During this process, one observes linear increase of the entropy. Existence of the coarse-grained (quasi-steady) state is numerically confirmed by means of non-dissipative kinetic simulations of the ion temperature gradient driven turbulence with high velocity-space resolution. The results also provide some key informations for making collisionless kinetic-fluid closure models.

[FP1.040] The CURRAY Ray Tracing Code as an NTCC Module

The geometric optics code CURRAY(T.K. Mau, S.C. Chiu, R.W. Harvey, EPS Top. Conf. RF Heating and Current Drive of Fusion Devices, Brussels (1992) 181.) was developed jointly by UCSD and GA in the early 90's to model fast wave current drive experiments in DIII-D and to determine RF CD power requirements for the ARIES power plant study. It calculates the wave power deposition and current generation profiles in the ICRF and LH range of frequencies. Recently CURRAY was used to analyze HHFW heating and CD on NSTX. The code is presently being implemented as an NTCC library module. The modifications include making use of other high quality NTCC library modules in place of commercial libraries and developing the test cases and Make files for easy installation. The module will then be used to provide heating and CD source terms in transport codes, such as TRANSP, for NSTX discharge analysis through the XPLASMA interface to supply equilibrium and kinetic information. F90 features have also been included to increase portability between systems. Relevant test cases will be run to check conformance with NTCC standards before submission to the Review Committee.

[FP1.041] The TRANSP Neutral Beam Injection Calculation as a NTCC Library Module

The TRANSP transport analysis code contains a comprehensive computation of particle and source rates associated with neutral beam injection (NBI). The code to carry out this calculation, which has been developed over the past two decades, includes beam deposition, fast ion two-dimensional orbiting, beam driven current and momentum transfer, and accounts for particle collisions and charge exchange transport of beam particles. The TRANSP NBI code uses a Monte Carlo technique and has been tested with experimental data from most tokamaks since the early 1980s. In spite of the high degree of credibility, the TRANSP NBI code has not been used with other codes because of its high level of integration into the TRANSP code. The National Transport Code Collaboration (NTCC) Module Library and associated Library standards provides the framework for extracting the TRANSP NBI calculation as a NTCC module. The paper reports on progress with regard to the development of this module. Elements of an object-oriented approach are applied to the Fortran legacy code. Based on a static analysis, sets of Fortran-90 modules are developed to encapsulate data and relevant methods. The philosophy used in converting a complex legacy code into an NTCC module is discussed.

[FP1.042] Physics modules for transport in 2D and 3D toroidal plasmas

Several physics modules for studying transport in 2D and 3D toroidal plasmas using many of the modern features of Fortran 90 are now available through the National Transport Code Modules Library. These include: FRANTIC (neutral transport), NCLASS (neoclassical transport), CYTRAN (cyclotran radiation transport), AJAX (interface to 2D and 3D MHD equilibria), and TRACK (chord mapping through 2D and 3D equilibria). These modules take advantage of many of the efficient computational features of F90, while maintaining the ability to wrap them for calls from other languages. Physics features of each of the modules are illustrated. The TRACK and AJAX modules are fully 3D with essentially no computational penalty for treating axisymmetric plasmas. NCLASS has a comprehensive treatment of tokamak geometry for all neoclassical transport properties, while its electrical resistivity and bootstrap current are reasonable for quasi-axisymetric stellarators. The FRANTIC and CYTRAN modules illustrate cylindrical treatments of physics that have been used as approximations to source and sink terms in both 2D and 3D plasmas.

[FP1.043] Design of the National Transport Code

The National Transport Code is being redesigned for flexibility in configurability, use of multiple modules, data access and user interaction. The National Transport Code Collaboration (NTCC) Demonstration Project developed a code that (1) easily incorporates physics modules contributed by multiple researchers; (2) accesses data from multiple sources, including MDS plus; (3) can be invoked over the web and run interactively through a graphical user interface, or can be run interactively in a scripting environment, or can be executed non interactively like traditional codes. The demonstration code evolved temperatures and toroidal rotation with sources and sinks taken from data, while the transport coefficients were calculated from any of several models. The NTCC plan calls for upgrading this to a full-function, flexible predictive transport code. To facilitate this, the code is being re-designed. Arbitrary combinations of dynamic fields will be accommodated at the linear-solver level. Wrapping externally supplied modules is substantially simplified. All fields and transport coefficients are treated equally, so that any could be determined by evolution equations, data files, algebraic prescription, or an externally supplied module. Finally, the dynamic-field update accommodates interchangable solvers and solution algorithms

[FP1.044] Studies Planned with the NTCC Code

The focus of the ongoing restructuring of the NTCC project is to facilitate carrying out studies in collaboration with experimentalists associated with the NSTX, CMOD and DIII-D projects. For example, in NSTX neutral beam heated discharges, T_i>> T_e across the entire profile even though \hboxW_beam (80\kern2pt keV) >> T_e,crit (15\kern2ptkeV). In DIII-D, neon injection into \hboxL-mode edge plasmas caused bifurcation of toroidal rotation velocity shear, resulting in suppression of long-wavelength turbulence and reduction of ion thermal transport. In Alcator C-Mod with off-axis ICRF heating at 4.5\kern2ptT, internal transport barriers resulted in high central electron densities, n_ e(0)>\kern2pt 6 - 8 \times 10^20\kern2pt m^-3, with an H-mode edge. Use of the NTCC code to model such experiments should help in the identification of the key physics issues and in the determination of the best models to describe such phenomena. A plan is presented for an expanded NTCC project that addresses these and other needs of the major facilities. The plan includes a redesign and expansion of the NTCC code and the preparation of heating modules for the NTCC Module Library. These modules include the NBI module extracted from TRANSP, ECH (TORAY), and ICRF (TORIC and CURRAY).

[FP1.045] Physics Modules in the National Transport Code Modules Library.

The National Transport Code Collaboration (NTCC) Modules Library is being extended to incorporate major physics modules to describe neutral beam and RF heating of tokamaks. This builds on the work of the past three years, during which the modules library standards were defined, the code distribution system was built, and modules for portability, numerical methods, 2d tokamak MHD equilibria, and atomic/nuclear reaction rate databases were added. This poster reports on the status of the NTCC Modules Library project, with emphasis on large integrated physics modules, including but not limited to: the TRANSP Monte Carlo fast ion package and supporting systems, and, the RF packages CURRAY, TORAY, and TORIC.

[FP1.046] Progress report on upgrade to 3-D neutral code NUT.

We present a progress report on an improved, NTCC compliant modular subroutine version of the 3-D NUT algorithm. The existing 3-D semi-analytic neutral transport code NUT performs calculations of neutral penetration into 3-D plasmas much faster than Monte Carlo methods. The enhanced speed is due to the use of exact analytic solutions that are valid up to the scale lengths over which the plasma parameters and the corresponding neutral mean free paths vary. NUT has been used successfully in the analysis of spatially resolved H_\alpha data from TEXT and TFTR. NUT-II will feature: a) the capability to accept arbitrary or adaptive grids, b) a parallel architecture capable of efficiently using massively parallel machines, c) a more dynamic memory management structure, d) updated atomic cross section data, and e) new interfaces to plasma simulation grids, experimental geometry, and atomic physics databases stored in MDS plus database trees. The new NUT-II can fulfill a widely expressed critical need for popular codes such as TRANSP, for modeling 3-D data from many machines such as MAST, NSTX, DIII-D, and LHD, for modeling and designing divertors, and for modeling of plasma processing reactors. Ref: P. M. Valanju, J. Comp. Phys. 88, 114 (1990).

[FP1.047] Improvements to the National Transport Code Collaboration Data Server

The data server of the National Transport Code Colaboration Project provides a universal network interface to interpolated or raw transport data accessible by a universal set of names. Data can be acquired from a local copy of the Iternational Multi-Tokamak (ITER) profile database as well as from TRANSP trees of MDS Plus data systems on the net. Data is provided to the user's network client via a CORBA interface, thus providing stateful data server instances, which have the advantage of remembering the desired interpolation, data set, etc. This paper will review the status and discuss the recent improvements made to the data server, such as the modularization of the data server and the addition of hdf5 and MDS Plus data file writing capability.

[FP1.048] Predictive Modeling of Tokamak Configurations*

The Corsica code provides comprehensive toroidal plasma simulation and design capabilities with current applications [1] to tokamak, reversed field pinch (RFP) and spheromak configurations. It calculates fixed and free boundary equilibria coupled to Ohm's law, sources, transport models and MHD stability modules. We are exploring operations scenarios for both the DIII-D and KSTAR tokamaks. We will present simulations of the effects of electron cyclotron heating (ECH) and current drive (ECCD) relevant to the Quiescent Double Barrier (QDB) regime on DIII-D exploring long pulse operation issues. KSTAR simulations using ECH/ECCD in negative central shear configurations explore evolution to steady state while shape evolution studies during current ramp up using a hyper-resistivity model investigate startup scenarios and limitations. Studies of high bootstrap fraction operation stimulated by recent ECH/ECCD experiments on DIIID will also be presented. [1] Pearlstein, L.D., et al, Predictive Modeling of Axisymmetric Toroidal Configurations, 28th EPS Conference on Controlled Fusion and Plasma Physics, Madeira, Portugal, June 18-22, 2001. * Work 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.

[FP1.049] Kinetic Ballooning Mode Transport in Central Tokamak Plasma

Kinetic ballooning mode is likely to explain the deep core transport in tokamak that can not be accounted for by ion temperature gradient or trapped electron mode. The observed "soft" \beta limit on experimental pressure profiles also suggest the presence of kinetic ballooning mode in tokamak central region. In this work, we study the linear stability of kinetic ballooning mode in interior tokamak plasma with both complete kinetic dispersion relation and a fluid model. We derive the corresponding transport from kinetic ballooning mode in an attempt to improve the transport simulation of core tokamak plasma.

[FP1.050] Models for pedestal temperature at the edge of \mboxH-mode plasmas in tokamaks

Models are developed for predicting the temperature at the top of the pedestal at the edge of H-mode plasmas in tokamaks. The derivation of one model assumes that the pressure gradient is limited by ballooning mode stability and the width of the pedestal is based on the E \times B shear suppression of turbulence, together with the effects of the bootstrap current. The predictions obtained with this model are compared with experimental data from 182 discharges from DIII-D, JET and JT-60U. The pedestal temperature in these discharges ranges from 0.2 to 3.3\kern2ptkeV, and the comparison between the model and the data yields an RMS deviation of 0.41\kern2ptkeV. Models are also considered to include the effects observed in approximately 50 discharges from Alcator C-Mod. Pedestal temperature models are implemented to provide a boundary condition in the BALDUR integrated predictive modeling code. The resulting code is tested by comparing tokamak simulations with experimental data. In addition, this code, with edge temperatures provided by the pedestal model, is used to predict the performance of future tokamak designs such as FIRE, IGNITOR and ITER-FEAT.

[FP1.051] Conditions for ITB formation in tokamaks.

Analysis of the international ITB database is aimed at rigorous definition of ITB and finding general conditions for ITB formation observed in different tokamaks. Dependencies of the ITB power threshold on magnetic field, plasma current, density and q profile are compared. Magnetic shear is implemented in the database. Systematic reduction of the ITB power threshold is observed in the case of zero or negative magnetic shear. ITB is formed often near rational q surfaces when heating power is around the threshold power.

[FP1.052] Electron Transport in Lower Hybrid Current Drive Discharges in Tore Supra

The stability and turbulent transport for the lower hybrid current drive experiments (LHCD) in Tore Supra is analyzed using the electron distribution functions computed with a combined ray tracing/Fokker Planck code (DELPHINE). The distribution functions contain substantial parallel thermal fluxes with radial gradients that are greater than than those in the current and temperature profiles. Thus conventional drifting Maxwellian models are inadequate for the stability analysis. The electrostatic dispersion relation is computed in some detail to re-examine the growth rates of the electron temperature gradient ETG modes in these plasmas with substantial population of fast current carrying electrons. Electron Landau damping is reduced compared to its value in a Maxwell distribution and there is the substaintial radial gradient of the parallel thermal flux. These potential instability drives are controlled by the magnetic sheared induced electron Landau damping that becomes strong as the fluctuations propagate into regions of large parallel wavenumber away from the mode rational surfaces. An assesment of the these stability effects as function of the dimensionless Stix RF heating power parameter is carried out using the LHCD electron distribution functions.

[FP1.053] Analysis of the Critical Electron Temperature Gradient in Tore Supra

The Tore Supra database of fast wave electron heating (FWEH) discharges is analyzed with respect to the role of the critical electron temperature gradient. The experimental evidence for the linear theory critical gradient is presented from both (i) power balance thermal flux versus the temperature gradient extrapolated to zero flux and (ii) the fluctuation spectra versus the gadient extrapolated to the vanishing point. Case studies with the LOCO and BALDUR transport codes are used to investigate the impact of the critical gradient in both cases close to the Ohmic discharge with 0.75 MW of rf power and strongly heated discharges with up to 7.4 MW of RF power. The interpretation of the critical gradient as a heat pinch term is also explored with thermodynamic theory and the space-time symmetries of the underlying dynamical equations. There is a relationship between the critical electron temperature gradient and the particle pinch. The evidence for the two candidates to explain the electron transport: trapped electron modes (TEM), characterized by wavenumbers much longer than the ion gyroradius, and the smaller-scale electron temperature gradient modes (ETG) with wavelengths comparable and smaller than the ion gyroradius is presented in detail.

[FP1.054] Integrated Predictive Modeling Simulations of the Mega Amp Spherical Tokamak (MAST)

Integrated predictive modeling simulations are carried out using the BALDUR transport code for H-mode and L-mode discharges in the Mega-Amp Spherical Tokamak (MAST). Results obtained using either the Multi-Mode transport model (MMM95) or, alternatively, the mixed-Bohm/gyro-Bohm (JETTO) transport model are compared with experimental data. In addition to the anomalous transport, neoclassical transport is included in the simulations. The ion thermal diffusivity in the inner third of the plasma is found to be predominantly neoclassical. The computation of the trapped particle fraction and the rest of the variables needed by the neoclassical transport model have been generalized to account for physical effects at low aspect ratio and high beta. The sawtooth oscillations in the simulations have the effect of flattening the central temperature and density profiles as well as spreading the neutral beam injection heating profile across a broad radial sawtooth mixing region. The simulation results obtained in this study are compared with experimental data for electron temperature and density profiles (ion temperature profile data are not available for these discharges).

[FP1.055] Dusty and Strongly Coupled Plasmas

[FP1.056] A Non-local Study of the Rayleigh Taylor Instability in a Dusty Plasma

The nonlocal theory is employed to study the Rayleigh Taylor instability in an inhomogeneous dusty plasma in the presence of a uniform magnetic field. Typical space parameters are used to investigate the behavior of the instability. A comparison of the variation in the growth rate of instability in a pristine plasma versus dusty plasma shows that the growth rate in the dusty plasma is enhanced. A study of the effect of the dust parameters on the growth rate of the instability is made and the dependence of the growth rate on the dust charge, the dust mass and the dust density is studied.

[FP1.057] Role of Streaming Instabilities in Dusty Plasma Phase Transition

Experiments on crystallization in a dusty plasma indicate that below a critical neutral pressure, Pcrit, the dust component in a dusty plasma behaves as a weakly coupled fluid (H. Thomas and G.E. Morfill, Nature 379, 806, (1996)) with a dust temperature much higher than the neutrals, ions, or electrons. As the neutral pressure is increased such that P>Pcrit the dust grains crystallize forming a coulomb lattice. The transition from the solid to fluid state has been addressed, (V.A. Schweigert et al., Phys. Rev. E 54, 4155, (1996)) but an important outstanding issue is the physical process that governs the transition from the fluid state to the solid state, which is seen in both experiments and simulations. Our analysis indicates that a two-stream instability between the ions and dust is responsible for dust heating for P<Pcrit and thereby preventing the dust component from crystallizing. As the neutral pressure is increased the ion-neutral and dust-neutral collision frequencies increase. For P>Pcrit the collision frequencies are sufficiently large to stabilize the two-stream instability. This removes the heat source, which enables the conditions for strong coupling in the dust. Consequently, the dust component can now condense into a coulomb crystal. We will analyze the relevant streaming instabilities, and quantify Pcrit in terms of the background plasma parameters.

[FP1.058] Ionization Instabilities and Resonant Acoustic Modes at High Densities

Recent experiments on voids in dusty plasmas have shown that the occurence of the void is preceded by the excitation of the filamentary mode. These oscillations are typically broadband, peak around 100 Hz and are characterized by sudden onset as the charge on the grain exceeds a critical value. A recent theoretical analysis by Wang et al.[2001] explains the sudden onset of the filamentary mode in terms of a resonant coupling of the ion and dust acoustic modes. As it is known that the dust charge is a function of the dust density and furthermore, that the very nature of acoustic modes changes at high dust densities, it is important to revisit the scenario of Wang et al. at high dust densities characterized by the Havnes parameter P. In this paper, we will present detailed dispersion relations that take into account the coupling of acoustic modes at large P, and present results on instability thresholds and growth rates. A simple model for the nonlinear saturation of the filamentary mode and void formation will also be presented.

[FP1.059] Rotating Toroidal Dust Clouds in the NRL Large Volume Dusty Plasma Experiment: I. Initial Observations

Large volume toroidal dust cloud structures that rotate in both the azimuthal (i.e., about the cylindrical axis) and poloidal (i.e., about the toroidal axis) directions have been observed in the Naval Research Laboratory's Large Volume Dusty Plasma Experiment (DUPLEX). The DUPLEX plasma is an argon dc glow discharge operated at pressures ranging from 50 to 250 mtorr and voltages ranging from 500 to 1000 V. Detailed measurements of the rotational velocities of the 1 micron alumina particles are made using the Auburn University Particle Image Velocity (PIV) laser system [E.\ Thomas, Jr., Phys.\ Plasmas, 6, 2672 (1999)]. The size, shape, and rotational characteristics of the toroidal cloud depend sensitively on the neutral pressure and discharge voltage. Examples illustrating the morphology, formation of central voids, and disruption of the cloud will be presented.

[FP1.060] Rotating Toroidal Dust Clouds in the NRL Large Volume Dusty Plasma Experiment: II. Particle Velocity Analysis

Detailed analysis of Particle Image Velocity (PIV) observations [E.\ Thomas, Jr., Phys.\ Plasmas, 6, 2672 (1999)]. of large volume toroidal dust cloud structures made in the Naval Research Laboratory's Large Volume Dusty Plasma Experiment (DUPLEX) is presented. These structures form spontaneously in the DUPLEX dc glow discharge plasma. In general, the structures rotate in both the azimuthal and poloidal directions. The shape of the structures and the characteristics of the rotations depend sensitively on the characteristics of the background plasma. We present detailed analysis of the spatially resolved motion of the 1 micron alumina dust grains suspended in the argon plasma. Analysis codes written to calculate a number of variables, including the radial and angular velocities and angular accelerations of the dust grains will be described.

[FP1.061] Analysis of particle motion in the void region of rf glow discharge complex plasmas

Large, central void regions in rf-generated complex plasmas have previously been observed under microgravity conditions [G. E. Morfill, et. al., PRL, 83, 1598 (1999).]. A similar void region can be generated in laboratory experiments through the application of temperature gradient across the plasma volume. Through the use of particle image velocimetry (PIV) and laser flashing techniques, detailed measurements of particle transport in the complex plasma and in the void region have been made. Results to be presented highlight the spatial evolution of particle velocities in the void and the continuous acceleration of particles that transit the void volume and the rapid deceleration of particles near the void – particle cloud interface.

[FP1.062] Charged aerosol collection in the mesosphere during MIDAS/SOLSTICE 2001

A charged aerosol detector was flown during the MIDAS/SOLSTICE rocket campaign in June 2001 over Andoya, Norway. The probe is a graphite collection surface with a permanent magnet underneath to deflect electrons and light ions. This probe was first used on a sounding rocket over White Sands in 1998, where it measured a sharp positively charged layer at 86.5 km [Geophys. Res. Lett. 27, 3825 (2000)]. The first MIDAS launch was into a triple layered PMSE which extended from 82 to 90 km. On upleg, the probe measured a broad region of negatively charged particles inside a local ion and electron biteout at approximately 85 km. The second launch was into a very strong single-layered PMSE. On upleg, the probe measured another broad region of negative particles within the PMSE. On downleg, more negatively charged particles were seen in the PMSE, this time in a broad region with a very sharp upper boundary. During both flights the probe also measured a positively charged background which was well-correlated with an onboard positive ion probe. Also seen was a photoelectron signal due to solar UV. These other signals show that the probe functioned as expected during flight. Further analysis is underway.

[FP1.063] Particle Levitation in a Plasma Sheath Above a Surface

Dust grains suspended above the lunar surface have been observed on multiple occasions. Virtually all small, airless bodies in the solar system are coated with a dusty regolith; therefore, charged dust particle levitation and transport may also occur in planetary ring systems, on Mars, Mercury, planetary satellites, asteroids, and comets. The interaction between charged dust particles and a photoelectron layer or plasma sheath above the surface is the most likely explanation for these dust dynamics. We report the results of experiments on the levitation, dynamics, and charging of dust particles in an Ar plasma sheath above a flat plate. Types of particles tested include hollow and solid glass microballoons (<45 microns in diameter), polystyrene DVB beads (5 microns and 10 microns in diameter), and JSC-1, a lunar regolith simulant (<25 microns in diameter). Plasma and sheath characteristics are determined through Langmuir probe and floating potential probe sweeps. An agitator under the surface provides a disturbance to inject dust into the sheath. Dust particles levitating above the surface of the plate are illuminated by an Ar laser and observed by a video camera.

[FP1.064] Experimental search for Debye shielding by orbiting ions

Theoretical and numerical models [Lampe et al., PRL 86, 5278 (2001); Goree, PRL 69, 277 (1992)] have shown that the Debye shielding cloud around negatively charged dust particles may contain a large fraction of trapped ions. In a search for these ions, we have modified an experiment in which the charge on dust particles is measured in a Faraday cup after they have fallen through a plasma in a single-sided DP device [PRL 75, 838 (1995)]. The particles are spherical glass of 0.1 mm diameter and the plasma density is adjusted for a 1 mm Debye length of the cold ions from charge exchange collisions. The pressure of argon is adjusted to give many of these collisions in the time of flight but few in the Faraday cup. A bias voltage on the entrance tube to the cup is used to locally remove the sheath at the wall. A bias voltage on the Faraday cup may or may not be applied to remove the orbiting ions. Negatively charged dust has been detected and the charge on grains measured, however, an experimental difficulty is the generation and diagnosis of plasma with argon pressure below 10(-5) Torr. Additional experiments are in progress.

[FP1.065] Dust charging on surfaces

Dust particles are a nuisance in plasma processing of semiconductors, cause a rich variety of phenomena in laboratory plasmas, and play an important role in planetary and space physics. Experimental investigations have been made of the charge on dust particles resting upon a metal surface in vacuum [Sternovsky et al., JVST A, in press]. The surface is agitated so that the particles drop though a small hole and a Faraday cup beneath measures the charge on each particle. The surfaces are metals (Hf, Zr, V, W, Co, Ni, Pt and stainless steel) and the dust grains are both metallic conductors (Zn, V, and stainless steel) and insulators (silica, alumina, Martian regolith simulant, and lunar dust simulants) in the size range of 35 – 200 microns. The basic charging mechanisms studied are contact charging, electrical induction in presence of an electric field, and charging under UV illumination. The contact charge is proportional to the difference in work functions. Investigations of the charging by contact with different metal surfaces allow the determination of the effective work function of dust samples (5.8 eV and 5.65 eV for the Lunar and Martian dust simulants, respectively). An electric field above the surface induces an additional charge on metallic grains consistent with Gauss's law. The contact charge and the induced charge on insulating grains increased with repeated agitation of the surface. UV irradiation may increase or decrease the dust charge depending upon the relative importance of photoemission and photoconductivity.

[FP1.066] Dust wave instabilities in collisional plasmas

Low frequency dust waves associated with the dynamics of charged dust grains in a plasma can be excited by the drift of plasma particles relative to the dust. Collisions of charged particles with neutrals can affect the properties of such waves and instabilities in weakly ionized dusty plasmas in laboratory or space environments. In this study, two electrostatic dust wave instabilities in collisional dusty plasmas are considered using linear kinetic theory. First, an ion-dust streaming instability with frequency less than the dust-neutral collision frequency is investigated. Under certain conditions, this instability may be excited by streaming ions even when the dust acoustic wave is heavily damped. Second, low frequency drift instabilities are studied in the case of a magnetized collisional dusty plasma with gradients in the dust and electron density. Applications of these instabilities to dusty plasmas in the laboratory or lower ionosphere are discussed.

[FP1.067] Beam-Plasma Instabilities in Strongly Coupled Plasmas

Strongly coupled dusty plasmas under laboratory conditions are permeated by streaming ions: in this scenario beam- plasma instabilities may be excited. The strong coupling between the dust grains, however, fundamentally affects the condition for instability and renders the conventional Vlasov treatment entirely inadequate. Based on the Quasilocalized Charge Approximation [1,2,3] we develop an analysis of instabilities generated by the relative streaming of a weakly coupled and a strongly coupled plasma. The central role in this formalism is played by the Dynamical Matrix D(k), a functional of the equilibrium correlation function, determined in our earlier work [2,3]. Novel physical effects generated by strong coupling alter both the beam resonance condition and the coupling between the beam and the plasma modes. Our analysis covers both resonant and non-resonant, as well as resistive instabilities.

[1] Kenneth I. Golden and Gabor J. Kalman, Phys. Plasmas, 7, 14 (2000) [2] M. Rosenberg and G. Kalman, Phys. Rev. E 56, 7166 (1997) [3] G. Kalman, M.Rosenberg and H. E. DeWitt, Phys. Rev Lett. 84, 6030 (2000)

[FP1.068] Shocks and solitons in a 2D strongly coupled complex (dusty) plasma.

Shock and solitary waves were studied in a 2D strongly coupled dusty plasma. The experiments were performed in an rf parallel plate discharge. A monolayer dust lattice was formed from monodisperse plastic microspheres and levitated in the electrode sheath. Linear compressional pulse waves were launched in the lattice. The wave structures were directly imaged with a high speed video camera and analyzed with particle tracking software. Waves formed by strong excitation resembled shocks destroying the crystal structure. Weak excitation produced soliton-like waves i.e. those that propagated with no or little change of the shape.

The theory describing the experiment is based on a set of equations of motion written for a monolayer hexagonal lattice. It takes into account damping, dispersion and nonlinearity. The resulting KdV equation allows determination of the relation between the particle charge and the screening length of the lattice in the Yukawa approximation.

[FP1.069] Mach cones and wakes in a 2D dusty plasma crystal

A dusty plasma is an ionized gas containing small particles of solid matter. Highly charged polymer microspheres were suspended in a capacitively-coupled parallel-plate rf plasma. The particles settled in a horizontal monolayer, arranged in a hexagonal lattice. They were imaged using a video camera. A 2D plasma crystal sustains two kinds of sound waves, compressional and shear (transverse). When these waves are excited by a moving supersonic disturbance, the superposition of the waves creates a Mach cone, i.e., a V-shaped wake. In our experiments, the disturbance was a moving spot of argon laser light. By moving the laser spot faster than the shear sound speed c_t, but slower than the compressional sound speed c_l, we excited a shear wave Mach cone. Alternatively, by moving the laser spot faster than c_l, we excited both cones. We also observed a wake structure that arises from the compressional wave’s dispersion, for laser spot speeds both higher and lower than c_l. We compare our results to Dubin’s theory (Phys. Plasmas 2000) and to molecular dynamics simulations.

[FP1.070] Recent Results from the Idaho Dusty Plasma Experiment*

Experiments on the Idaho Dusty Plasma facility continue to focus on the study of dust dynamics. Plasma formation is from a DC Argon glow. Plastic spheres in the 20 micron diameter range are suspended in the plasma sheath beneath the DC glow plasma. The dust manifests solid and liquid phases depending on the neutral background pressure. Dust kinetic energies are typically in the range of the measured electron, 5 eV. Experimental results indicate that the dust energy source is fluctuations in the electric field in the sheath. Fluctuation studies of the dust kinetic properties have been carried out. * This work was supported by the University of Idaho and the Idaho NASA Space Grant Consortium.

[FP1.071] Influence of Supra-thermal Electron on the Dynamic Behavior of Dust Plasma in the Ion Sheath with a Low Gas Pressure

In our recent experiments, it is clearly observed that dust particles trapped in the ion sheath under a low gas pressure of 1-5 mtorr oscillate vertically in a parameter range depending on the plasma density ne and the neutral gas pressure P. The excitation mechanism of unstable vertical motion is discussed based on the finite charging time of dust particles moving over an complex sheath structure[1]. The ion sheath structure and space distribution of dust charge in the ion sheath is drastically changed by presence of supra-thermal electron in the plasma, so that the supra-thermal electron have a great influence on the dynamic behavior of dust particles. In this presentation, we will show the experimental result of dynamic behavior of dust particles, and discuss the influence of supra-thermal electron upon the instability of dust particles trapped in the ion sheath. [1] S. Nunomura, T. Misawa, N. Ohno and S. Takamura: Phys. Rev. Lett. \bf83 (1999) 1970.

[FP1.072] Molecular Dynamics (MD) Calculation of Viscosities in Yukawa Systems

We have estimated the shear viscosity of Yukawa system, using molecular dynamics (MD) simulations of its thermodynamical equilibrium states. The Yukawa system is a collection of particles interacting with the Yukawa (screened Coulomb) potentials. This system serves as a model for homogeneous dusty plasmas. The shear viscosity is of special interest as it characterizes the dispersion relations of shear waves in strongly-coupled states. Our simulation results show that the shear viscosity has strong dependence on both the system temperature and potential screening length. Our MD simulation code includes the Ewald sum (i.e., the infinite potential sum due to all periodic images) appropriate for the given periodic boundary conditions, which enables us to simulate the system with weak screening, including the limit of no screening. [The Yukawa system with no potential screening is called the one-component plasma (OCP)]. As the screening becomes weak, we have demonstrated that the shear viscosities of Yukawa systems smoothly converge to the previously known estimates of the OCP shear viscosity. In the presentation, we shall also discuss the scaling of shear viscosity as well as our recent estimate of the bulk viscosity of Yukawa system.

[FP1.073] The Classical analogue of the Casimir effect in a dusty plasma

The analogy betwen the Casimir force acting on two parallel plates in vacuum and a similar force, of classical nature, acting on two parallel plates immersed in a turbulent plasma is explored. Like in the vacuum case, the force is attractive in a plasma, but its magnitude depends now on the energy level of the turbulent fluctuations, as well as on the type of turbulence. In the absence of plasma the force reduces to the normal vacuum Casimir force. Application to dust attraction will be studied and compared to other attraction forces such as the shadow force produced by particle bombardment and the attractive force due to wakefield potentials induced by relative flow.

[FP1.074] Excitation of large amplitude dust grain oscillations in a plasma sheath

In the last few years several experiments have been performed towards studying the behaviour of dust crystals under low pressure regimes (1-10 mTorr). We analyzed the parameters that can excite vertical grain oscillations and the threshold conditions that can lead to the large amplitude vertical oscillations observed by Nunomura . The stabilizing effects of plasma density as well as the destabilizing effects of dust size are analyzed.

Under low pressure regimes, a modulation in plasma density can drive unstable such vertical oscillations even when the grain conditions would be stable. We followed the grain dynamics under different pressure regimes and studied the influence of friction in the process of energy gain from the modulation in plasma density. The grain dynamics at saturation may in fact reach two different equilibrium regimes: the grain may oscillate in phase with the plasma density or with a \sim \pi phase difference. The parameters leading to the saturation equilibrium situations are explored.

[FP1.075] Structural Properties of Strongly Coupled, Two-Di\-men\-sio\-nal Yukawa Plasmas

The study of two-dimensional (2D) systems has a long experimental history using, among other methods, colloidal systems, semiconductors, and liquid-helium surfaces. Due to the natural trap formed by gravity and the sheath electric field 2D screened plasmas are also readily formed in dusty plasma experiments. The dusty plasma system has numerous experimental advantages, such as very long-lived states, easy manipulation of single particles, and a macroscopic size allowing for video imaging each particle's trajectory. Dust grains imbedded in the background plasma can be modeled as a screened Coulomb system, which is also referred to as a Yukawa system. Our focus so far has been on structural properties of the fluid phase for which screening occurs more weakly than for the 3D counterpart. We have explored this through varying degrees of sophistication in the structure calculation through mean-field, hypernetted-chain (HNC), and self-consistent HNC (SC-HNC) approaches. The SC-HNC equations are obtained through a Gaussian ansatz for the 2D bridge function with parameters chosen to ensure thermodynamic consistency.

[FP1.076] Two-Dimensional Strongly Coupled Plasma on a Liquid Surface

A simple two-dimensional (2-D) strongly coupled one-component plasma system has been developed by confining charged fine particles on oil surface. The fine silica (SiO2) particles placed on the bottom of concave electrode where silicon oil is filed are charged up by applying a voltage with respect to the upper glass electrode which is coated by electro-conductive film (ITO). The particles move upward by the electric field between the electrodes and confined on the oil surface by the surface tension. This new system has several advantages in studying 2-D Coulomb solidification compared with the colloidal suspension in a water: the charge of the particles can be easily controlled by the voltage, the number of impurity ions which shield out the Coulomb potential can be reduced in the insulating oil. We observed the long time evolution of Coulomb crystal growth and the development of defect and dislocation in the crystal. We also study the mixture of two different ionic state of particles confined on the same oil surface.

[FP1.077] Analysis of Experimental Production of Dense Titanium Plasma

As part of the stockpile stewardship program, we are developing the capability to produce strongly coupled plasmas at very high density and modest temperature. In this experiment, we desire a cylindrical shell of titanium plasma with ion density \approx 0.1 times solid density and ion temperature of a few eV. The shell has a radius of 1 cm, a length of 4 cm, and a shell thickness of 0.2 cm. The plasma is produced by using \approx 1 MA of current (2.5 \mus risetime) from the LANL Colt capacitor bank to ohmically heat a 100 \mum thick titanium cylindrical foil to the desired conditions. Plasma pressure causes the titanium to expand to the desired thickness, with nylon tamps preventing further expansion. Magnetic force at the foil is reduced by splitting the return current between the axis and outside the foil. The primary diagnostic was two radial x-ray radiographic systems. Analysis of these data indicate the Titanium foil turns to plasma from the outside surfaces inward, rather than a bulk transition to plasma. The data indicate that after 4.8 \mus, roughly one half of the foil mass has been turned into plasma, which has expanded to fill the gap between the nylon tamps.

[FP1.078] High Power Microwave and Laser Sources

[FP1.079] Measurements of the Angular Distribution of Tunable, Monochromatic X-rays in the Laser Synchrotron Source Experiment

The Laser Synchrotron Source (LSS)\footnote P. Sprangle et al, J. Appl. Phys. 72, 5032 (1992) experiment at the Naval Research Laboratory has generated monochromatic, tunable x-rays by Thomson backscattering of laser photons from a relativistic electron beam. Laser photons from a 5 J, 10 nsec Nd:Glass laser at 1.053 um wavelength are backscattered from a 4 MeV, 300 mA electron beam generated by a one and a half cell S-band RF gun. Recent experimental results include the generation of 1X10^7 x-ray photons per pulse at 372 eV with a bandwidth of 7 eV. Results will be presented on the present experiment, which involves characterization of the angular distribution of the backscattered x-ray photons

[FP1.080] Cherenkov Radiation in the THz Frequency Range from a Magnetized Plasma

Large amplitude electrostatic (es) plasma wave are excited by laser beams or particle bunches in plasma accelerators. These es waves couple poorly to vacuum electromagnetic (em) waves, and are dissipated in the plasma. A fraction of the es plasma wave can be converted into an em wave by a applying a static magnetic field transverse to the propagation direction of the laser or particle bunch. The laser or particle bunch couples to the left branch of the extraordinary (XO) mode of the magnetized plasma. This process can be described as Cherenkov radiation by the laser pulse or particle bunch in the magnetized plasma. The radiation is emitted essentially in the forward direction, and in practical cases, at the plasma frequency. In the UCLA Neptune laboratory 30 GV es fields are generated through resonant excitation of the plasma wave using the beating of a two-frequency CO2 laser beam in the plasma beatwave accelerator (PBWA). The plasma frequency is around 1 THz. We plan to apply a magnetic field in the 7 kG range transversely to the plasma, and to detect the short pulses of THz radiation. Analysis of the amplitude of the Cherenkov radiation emitted by the plasma will provide information about the plasma wave amplitude and lifetime. Analysis of the radiation frequency will provide information about the plasma density. This radiation will be used as an additional diagnostic for the PBWA experiment. The experimental results will be use to investigate the possibility of using this radiation scheme as a new high-power, short-pulse, THz radiation source. The experimental set-up, 2-d and 3-d PIC simulations and preliminary experimental results will be presented.

[FP1.081] Experimental Observation of Temporal Evolution of Radiation Frequency in DARC

In the DARC (DC to AC Radiation Converter) experiments, the short, tunable radiation in the microwave range from the periodic electrostatic field with a wavelength of 2\,cm is generated via a laser-produced relativistic ionization front with a pulse width of the ionizing laser, 6\,ns. The frequency of the emitted radiation is observed by the time of flight diagnostic method using the delay waveguide line. The center frequency from 15 to 42\,GHz in the K- and Q-band waveguide type DARC is observed and is in a good agreement with the expected theoretical values. The temporal evolution of the radiation frequency is measured by using the local oscillator in the K-band range. The radiation signal is separated into two by the power divider. One is the radiation power detected by the crystal detector, and the other is for the frequency measurement by combination of both the local oscillator and the mixer. The observed frequency of the emitted radiation is from 18\,GHz to 32\,GHz with the rise time of 6.5\,ns. This change of the radiation frequency corresponds to the temporal evolution of the plasma density.

[FP1.082] Genaration of Short Microwave Pulse by Interaction both Periodic Electrostatic Field and Superluminous Ionization Front

A short microwave pulse is generated from a periodic electrostatic field by an underdense ionization front called "DARC". In the " DARC" scheme, thefrequency of the emitted radiation is given by both of the dispersion relation of the electromagnetic wave and the phase continuity between the initial field and the radiation. The front velocity is defined by the same direction as the wavenumber vector of the initial field, that is, v_f = d z/ dt = c/\cos \theta. As the front velicity is greater than the speed of light, this is called superluminous ionization front. The radiation frequency depends on the plasma density and the front velocity, i.e., the incident angle of the ionization front. An emitted radiation frequency is given by \ omega = (c^2 / v_f^2 - 1)^-1 \ (c^2k_0/ v_f) \pm [ ømega_p ^2 + c^2k_0^2 - (c ømega_p/v_f)^2 ]^1/2\, where \ omega_p, c and v_f are the plasma frequency, the speed of light and the front velocity, respectively. In our experiments, the ionization front is created by 4 ømega_0 light (266\,nm, 100\,mJ, 6\,ns (FWHM)) of the Nd:YAG laser. The wavelength of the initial electrostatic field is 1.4\,cm. A working gas is TMAE (U_i = 5.36\,eV). Incident angle of ionization front is from 0 through 90 degrees. The radiation frequency is observed by the time of flight method. The observed radiation frequency is from 14 through 28\,GHz range and is in good agreement with the expected values. lines footnote

[FP1.083] Experimental Observation of Emitted Radiation from Cerenkov Wakes in Perpendicularly Magnetized Plasma

A proof-of-principle experiment has been demonstrated the radiation generation from the Cerenkov wakes excited by an ultrashort and high intensity laser pulse in a perpendicularly magnetized plasma. The laser system for this experiment consists of a mode locked Ti:sapphire laser with the pulse duration of 100 fs (FWHM), operating at 800 nm wavelength with a maximum energy of 100 mJ and a repetition rate of 10 Hz. The laser pulse focused by a lens of f/5 is irradiated into the gas-filled interaction region with a static magnetic field from 0 to 6 kG. The emitted radiation is observed by microwave components. The frequency of the emitted radiation is in the millimeter range (up to 0.2 THz) with the pulse width of 200 ps which is estimated to be close to the plasma frequency. The signal intensity of the emitted radiation is proportional to the strength of the applied static magnetic field as is expected >from the theory. The polarization and spatial distribution of the emitted radiation are also observed.

[FP1.084] Cusp Gun Second-Harmonic TE21 Gyro-TWT Amplifier

A second-harmonic TE21 gyro-TWT with an axis–encircling beam is being constructed at UCD that is predicted to double the efficiency of our previous 200 kW, 12% efficient, MIG second-harmonic TE21 gyro-TWT [Q. S. Wang, et al., Phys. Rev. Lett. 75, 4322 (1995)]. The new device will avoid the loss in efficiency due to off-axis electrons interacting with a linearly polarized mode. The device is predicted by our large-signal code to produce 50 kW at 30 GHz with 20% efficiency, 30 dB saturated gain and 3% bandwidth. The 70 kV, 3.5 A axis-encircling electron beam will be produced by a Cusp gun delivered by Northrop Grumman. The device has been designed using linear theory to provide a 30% safety margin from absolute instability. Loss is added to the first 30.5 cm of the circuit walls to ensure the device is stable to harmonic gyro-BWO. The last 11.5 cm of the circuit is lossless to avoid attenuating the high power wave. The device also employs a sliced mode-selective circuit to suppress odd order azimuthal modes.

[FP1.085] Stability of Heavily Loaded W-Band TE01 Gyro-TWT

The validation of the stability and small-signal results of a high performance, high power, W-Band TE01 gyro-TWT driven by a 100 kV, 5 A, MIG electron beam with vt/vz=1.0 and axial velocity spread of 5% will be presented. The single-stage amplifier is heavily loaded [K.R. Chu, et al., Phys. Rev. Lett. 81, 4760 (1998)] for stability and is predicted by our large-signal simulation code to generate 140 kW with 28% efficiency, 50 dB saturated gain and 5% bandwidth. Linear theory has been used to determine that the loss needed to stabilize the gyro-BWO interactions is 70 dB at 93 GHz for the circuit with a cutoff of 91 GHz, which was achieved by coating the walls with Aquadag. The TE01 input coupler has less than 2 dB insertion loss over a 3% bandwidth and the edges of the MIG emitter are coated with molybdenum to suppress edge emission.

[FP1.086] Ka-Band Second-Harmonic Peniotron with Fundamental Mode Interaction

The peniotron interaction has been proven to have a fundamentally higher electronic conversion efficiency than the gyrotron interaction [T. Ishihara, et al., IEEE Trans. on Electron Devices 46, 798 (1999)]. The UC Davis peniotron is designed for high device efficiency while providing immunity to mode competition from gyrotron interactions. The four-vane slotted circuit is designed for fundamental mode resonance at 34.0 GHz for interaction at the second-harmonic cyclotron frequency of an axis-encircling electron beam produced by a Northrop Grumman Cusp gun. The electronic efficiency for interaction with the 70 KV, 3.5 A beam is predicted as 58%. The cavity incorporates diagnostic side-wall couplers and a changeable output coupling iris. Overcoupling of the cavity is predicted to yield a device efficiency of 47%. The device easily couples to the TE11 mode of the circular output waveguide. The diffraction coupling heavily loads higher order axial modes while avoiding mode conversion at the output.

[FP1.087] Cusp-Gun W-Band Sixth-Harmonic Slotted Gyrotron

A high-harmonic slotted gyrotron has been constructed at UC Davis to be driven by a 70 kV, 3.5 A, axis-encircling electron beam from a Northrop Grumman Cusp gun. The 94 GHz, slotted sixth-harmonic gyrotron is predicted to generate 50 kW with an efficiency of 20%. The test-stand is being reconfigured to more nearly approximate Northrop Grumman’s magnetic Cusp profile. In a Cusp gun, the electrons acquire their azimuthal velocity from the vzBr force in the magnetic reversal region.. The design will also be presented for an 8th-harmonic W-band gyrotron whose magnetic field can be supplied by a lightweight permanent magnet.

[FP1.088] Operation of a Second Harmonic, Ku-Band Gyroklystron

At the University of Maryland, we have designed a 4-cavity Ku-Band second harmonic gyroklystron tube which is expected to produce peak powers in excess of 80 MW with gains above 50 dB. The first cavity is driven at about 8.57 GHz in the TE011 mode and the remainder of the cavities operate at twice the drive frequency in the TE021 mode. In this paper we describe the details of the design, cold-testing, and hot-testing of the microwave tube. We also present results for a new output waveguide system which is designed to transform the output signal of the gyroklystron into a configuration which can be used to drive a compact linear accelerator structure at 17.136 GHz. The transformation occurs in stages, with the TE02 mode converted to a TE01 mode via a rippled-wall converter, followed by a conversion to the TE20 mode in rectangular waveguide, and completed with a bifurcation and linear taper to (2) standard WR62 rectangular waveguides. Details of the theoretical and cold-test results for each subsection will be presented.

[FP1.089] Theory of the gyro-TWT with distributed losses

Both small and large signal theories describing a two-stage gyro-traveling-wave tube (gyro-TWT) with the first stage having distributed losses are developed. In the framework of the small-signal theory, the small-signal gain of the operating forward wave and the self-excitation conditions for parasitic backward waves are analyzed. The large-signal theory contains the analysis of the gain saturation, the velocity spread effect and the influence of the input section downtaper on the operation of the device. For evaluating the correctness of the simplified theory based on the use of of gyro-averaged equations for electron motion, the results are compared with simulations done by using the self-consistent code MAGY. The operation of the Ka-band gyro-TWT designed at NRL was analyzed. Results show that the small-signal gain calculated by both methods is essentially the same. Nonlinear calculations indicate that both methods predict practically the same wave growth rate in the uniform part of the waveguide. However, there is a certain inconsistency in analyzing the interaction in the input downtaper and output uptaper by these two methods.

[FP1.090] Initial Testing of a 140 GHz 1 MW Gyrotron

CPI has completed the fabrication of a 140 GHz 1 MW CW gyrotron to be used on the W7-X stellarator at IPP Greifswald. Testing of the initial build of this gyrotron had just begun when this abstract was prepared, and was expected to finish in September, at which time a planned rebuild of the device was scheduled to begin. This poster will summarize the gyrotron design, present results of initial testing, and outline any design changes planned as a consequence of these results. This gyrotron's design employs a number of advanced features, including a diode electron gun for simplified operation, a single-stage depressed collector to enhance overall efficiency, a CVD diamond output window, an internal mode converter that converts the excited TE28,7 cavity mode to a Gaussian output beam, and a high-voltage layout that locates all external high voltage below the superconducting magnet system without requiring an oil tank for insulation. Similar features are being used for an 84 GHz 500 kW system being built for the KSTAR tokamak program and for a 110 GHz 1.5 MW system being designed in collaboration with MIT, UMd, UW, GA, and Calabazas Creek Research with funding provided by DOE.

[FP1.091] Beam Confinement in Periodic Permanent Magnet Focusing Klystrons

The confinement of a tightly bunched electron beam is studied in a periodic permanent magnet (PPM) focusing klystron. By analyzing the Hamiltonian dynamics of a train of collinear periodic point charges interacting with a conducting drift tube, an rf field, and an applied PPM focusing field, a space-charge limit is derived for the radial confinement of lightly bunched electron beams, and is shown to be significantly below the well-known Brillouin density limit for an unbunched beam. Several state-of-the-art PPM klystrons developed at SLAC are found to operate close to this limit, shedding some light on the origin of observed beam losses.

[FP1.092] Confinement and Equilibrium of Bunched Annular Beams

The azimuthally invariant fluid equilibrium is obtained for a periodic strongly bunched charged annular beam with arbitrary radial density profile inside of a perfectly conducting cylinder and an external constant magnetic field. The electric and magnetic fields, which are utilized in the equilibrium solution, are computed self-consistently using an electrostatic Green’s function technique in the longitudinal rest frame of the beam. An upper bound on the maximum self-field parameter, which allows beam equilibrium is obtained. As an application of the model, we find annular beam equilibrium for the Relativistic Klystron Oscillator experiment at Phillips Laboratory and the Backward Wave Oscillator experiment at the University of New Mexico. In addition, we compare the self-field parameters of these with the maximum theoretical values. This work is supported by AFOSR.

[FP1.093] Intense Sheet Beams for Klystron and Accelerator Applications

It is shown that intense corkscrewing and periodic wobbling elliptic beam equilibria exist in piecewise uniform magnetic fields. The envelope stability analysis reveals that the corkscrewing elliptic beams in a uniform magnetic field are stable, whereas the periodic wobbling elliptic beams are stable in certain regions in the parameter space. These results are useful not only in beam matching, but also in producing large-aspect-ratio sheet beams for high energy accelerators such as the next linear collider as well as for use in high-power rf sources such as sheet-beam klystrons.

[FP1.094] Investigations of Intermodulation in a Klystron Amplifier.*

The growing importance of high data rate microwave transmission is driving a need for compact, high power, efficient amplifiers with good linearity. The requirement for linearity is becoming even more crucial by the desire to operate individual amplifiers under multi-tone excitations. An experiment has been configured to investigate the fundamental physics of intermodulation product (IMP) generation in klystron amplifiers (KLAs). The experiment employs a 1kW CW Varian 4K35L, 4-cavity KLA, with instantaneous bandwidth of 10 MHz. A primary purpose of this experiment is to compare experimental measurements with predictions from a new general theory of IMP generation [1].

* This work was supported in part by AFOSR, and by DUSD(Samp;T) under the Innovative Microwave Vacuum Electronics Multidisciplinary University Research Initiative (MURI) program, managed by AFOSR. Significant assistance by M. Blank and CPI are gratefully acknowledged.

[1] Y.Y. Lau, D.P. Chernin, C. Wilsen, R.M. Gilgenbach, IEEE Trans. Plasma Sci., Vol.28, 959 (2000).

[FP1.095] Investigations of Non-Linear Spectral Growth in a Broadband Traveling Wave Tube Amplifier.*

The growing importance of broadband microwave transmission has prompted the need for compact, high power, efficient amplifiers with good linearity. A high degree of linearity is crucial for non-distorted, simultaneous amplification of multiple carrier signals. Experiments are underway to investigate the fundamental physics responsible for the non-linear behavior of traveling wave tubes (TWTs). A novel 2-6 GHz TWT, jointly designed by UW and Northrup-Grumman researchers, includes multiple output taps along the helix, providing a unique means to observe the evolution of the rf spectrum at various axial positions, as well as the cold versus beam-loaded dispersion characteristics. The growth rates of the fundamental tones, the harmonics and the intermodulation products (IMPs), in addition to the dispersion characteristics, are measured and compared with model predictions.

* This work was supported in part by AFOSR, and by DUSD(Samp;T) under the Innovative Microwave Vacuum Electronics Multidisciplinary University Research Initiative (MURI) program, managed by AFOSR.

[FP1.096] Stationary Solutions of the Magnetron Equations

Previous studies of crossed-field electron vacuum devices such as magnetrons and crossed-field amplifiers (CFA) have centered on their initial growth, as an indication of their operating modes. In such an analysis, one assumes a growth rate, and solves the equations for the density profile, the operating frequency, and the growth rate^1. However, once the RF fields have saturated, continued operation as a growing mode is no longer possible, and the RF fields must then either switch into some purely stationary operating mode, or some other mode, which would in general be an non-operating mode. To study this regime of operation, we return to the classical planar magnetron equations^1 and study what solutions are possible when the RF modes saturate, with the consequencial vanishing of the growth rate. What we find is that upon saturation, the growth rate can indeed vanish, but its effect in the RF equations will be replaced by certain second-order velocity terms, and in particular, the second-order vertical velocity. This effect is important only at the diocotron resonance, and if they occur, at the two magnetron resonances. What we will do here is to derive and present the magnetron equations when the growth rate vanishes, and discuss the solutions of the RF equations about the these three resonance points. Lastly, we shall consider what implications these results may have for relativistic and nonrelativistic crossed-field devices.

1. D.J. Kaup, Phys. of Plasmas 8, 2473-80 (2001).

[FP1.097] Experiments on Relativistic, 100's MW Magnetrons and Low Voltage, kW Magnetrons

Experiments on relativistic 100's MW magnetrons are underway. Research is in progress to understand the roles that e-beam space charge and impedance plays in the output efficiency and microwave production of magnetrons. Mode diagnosis and control are also crucial issues. A Titan L-Band, 6-vane, relativistic magnetron is driven by the MELBA accelerator. Experimental parameters are: voltage = -400 kV, current = 1 - 10 kA and pulselengths = 0.1 - 0.6 microseconds. Power measurements and time frequency analysis of the heterodyned microwave signal find power levels of up to 150 MW at 1.04 GHz and 0.95 GHz. Recent cathode designs have helped reduce endloss current from 50% of the total current to 15%. Commercial kW magnetrons are being studied to understand noise generation mechanisms in crossed field devices.

* Research supported by AFOSR and by DUSD (S amp; T) under the Innovative Microwave Vacuum Electronics MURI program, managed by the Air Force Office of Scientific Research under Grant F49620-99-1-0297. Simulation support provided by the AFOSR-sponsored MAGIC User Group, administered by Mission Research Corp. Support also received from Northrop Grumman Corp.

[FP1.098] Computational Investigation of the Los Alamos Magnetically Insulated Line Oscillator

Numerical modeling has been performed using the particle-in-cell code ISIS to interpret earlier experimental investigations of a Magnetically Insulated Line Oscillator (MILO)*. Improvements in the experimental performance of the MILO attributed to interposing a plasma opening switch between the MILO structure and the downstream diode has been studied computationally. Pulsed power operating conditions examined were in the range of 400-600 kV, 50-150 kA, in 1 microsecond electrical power pulses. A downstream post-hole convolute diode load controlled the overall operating impedance of the system to match the average E/B velocity of the vacuum flowing electrons to the MILO resonance conditions. Two nominal MILO operating impedances were studied: 3 and 6 ohms. The 6 ohm geometry was a better impedance match to the the Marx pulse power generator and provided RF pulses greater than 600 ns.

*R. M. Stringfield,et al, Bul. Am. Phys. Soc. Vol 33, No. 9, October, 1988, p1959.

[FP1.099] An Investigation of Beam Loading in a Cavity

Beam loading remains a problem of primary importance in high power vacuum electronics. The DC and AC space charge alters the resonant frequency, the quality factor Q, and the gap transit time factor. These effects need to be accurately accounted for in the evaluation of the intermodulation products in a klystron driven by a multi-frequency signal, and in high power microwave sources. We report simulation results using MAGIC [1]. Emphasis will be placed on the loading at various combinations of DC and AC beams. Our data show that beam loading depends mainly on the beam perveance, and is virtually independent of the degree of beam neutralization and of the AC current that the beam carries when it enters the cavity. Comparison with theories and experiments will be reported.

This work was supported by the AFOSR, and by the DUSD (S+T) under the Innovative Microwave Vacuum Electronics MURI Program, managed by AFOSR under grant F49620-99-1-0297, and by the Northrop-Grumman Industrial Affiliates Program.

[1] L. Ludeking et al., “MAGIC User’s Manual”, Mission Research Corporation, March 1999.

[FP1.100] Cathode Improvements for a Magnetically Insulated Line Oscillator (MILO)

Recent experimental and computer simulation results on MILO have indicated that the large beam current density emitted from each end of the cathode leads to anode plasma formation via stimulated desorption. The presence of bipolar space-charge flow in the anode-cathode (A-K) gap causes significant microwave power reduction and RF pulse-shortening on a 600-ns time scale. We report on implementation of the miniature Pierce focusing electrode concept of Umstattd et al. [R. Umstattd et al., "Design and implementation of a new UHV threshold cathode test facility," Proc. SPIE, vol. 4031, pp. 185-194, 2000.] on MILO as a way to minimize the beam current density at both ends of the cathode. XOOPIC and ICEPIC simulations for two focusing electrode configurations are presented. Also, preliminary experimental results using the same focusing electrodes in a smooth-bore anode configuration (no RF) are shown.

[FP1.101] High Power Microwave Generation in a Backward Wave Oscillator with a Central Conductor

High power microwave radiation generated by Backward Wave Oscillators (BWOs) and Travelling Wave Tubes has been extensively studied for many years. Both of these devices use the interaction of an electron beam (typically an annular beam) with a rippled waveguide to generate microwaves. In this paper, the effects of placing a central conducting rod inside the annulus of the annular electron beam of a BWO will be presented. This geometry was simulated using the PIC code MAGIC. In preliminary simulations, the radiation frequency increased linearly with increasing inner conductor radius. In addition, the amplitudes of the higher harmonics also tended to increase with increasing inner conductor radius.

[FP1.102] High-power, pulsed, electronically tunable broadband microwave source

Recent progress on the development of a broad-band, tunable source of microwaves that can generate a power output of 50 MW and pulse duration of 500 ns will be discussed. The microwave source is based on the plasma Cherenkov maser and an advanced field emission cathode technology, both developed by PlasmaIOFAN. The microwave source will have a bandwidth of about 1 GHz and be easily tunable, electronically, from 2 to 6 GHz. We will discuss the theory of operation of the device and the basic design, construction and experimental results obtained, to date, on the system.

[FP1.103] Experimental Observation of Coherent Synchrotron Radiation in an X-Band Photoinjector

A high-gradient, high-brightness X-band rf photoinjector is used to produce sub-picosecond electron bunches at energies in the 1.5 - 2 MeV range, which are propagated in a short section of corrugated waveguide, with an inner diameter of 10 mm, a corrugation period of approximately 2.5 mm, and a modulation depth of 1.5 mm. Extremely short rf pulses are measured, in the 100 ps - < 5 ns range, corresponding to the slippage time between the electron bunch and the radiation produced by a coupling to the slow-waves supported by the waveguide structure. Coherence is measured by varying the bunch charge and verifying a quadratic scaling of the rf power; W-band radiation is produced using this new radiation mechanism, which may have very important potential applications in coherent millimeter-wave and THz radiation generation. We will discuss the technical details of the X-band rf gun and slow-wave circuit, as well as our experimental results, and a theoretical description of the interaction. A proposed millimeter-wave, chirped-pulse free-electron laser experiment using the X-band rf gun will also be discussed.

[FP1.104] Experimental Characterization of a High-Brightness X-Band Photoinjector

We report detailed experimental results on an X-band (8.547 GHz) rf photoinjector, currently being commissioned at UC Davis on the LLNL site. The device operates with an average accelerating gradient in the 50-100 MeV/m range, and produces relativistic (1.5-2 MeV) photoelectron bunches, with a measured relative energy spread \gamma / \gamma_0 < 1.8%, at the resolution limit of our energy spectrometer. The measured quantum efficiency is 2 \times 10^-5, at the maximum of the Schottky curve, and the 90% normalized emittance is measured at 1.65 \pi-mm mrad for 20 pC of charge. The bunch duration is extremely short, under the 700 fs resolution of our streak camera. Timing jitter has also been measured directly from the delay between the UV light diffused by the photocathode, and Cerenkov radiation produced by the photoelectrons; a systematic delay is found, corresponding to the time-of-flight difference between the UV photons and the accelerated electrons, and the additional timing jitter is estimated at 5 ps, or 15 degrees of rf phase. Finally, coherent synchrotron experiments are underway, as the photoelectron bunches radiate coherently up to the THz region.

[FP1.105] Free-Electron Laser with Ion-Channel Guiding

A theory of a planar wiggler free-electron laser with ion channel guiding is developed. An analysis of the quasi-steady-state electron trajectories is first obtained by solving the equations of motion for an electron in the ion channel electrostatic field and the wiggler magnetostatic field. Next a sixth-degree polynomial dispersion equation for electromagnetic and space-charge waves in the wiggler is derived. Numerical solutions of the polynomial equation yield the complex wave number as a function of the frequency of the waves. These results are used to illustrate the dependence of growth rate-frequency curves on the ion-channel frequency and the peak growth rate and corresponding wave frequency as functions of the ion-channel frequency.

[FP1.106] On the beam cross-section of high-gain Compton free-electron lasers

We investigate the linear stability properties of a high-gain Compton free-electron laser(FEL), relevant to the the beam cross-section matter. In common practices the FELs are usually operated in a configuration that the beam diameter is similar to the optical beam extent. In the present work we imagine a situation that the beam cross-section is at our control to a certain extent, and consider the stability property as a function of the beam diameter and the transverse density profile. Identified is a characteristic radius that scales the beam radius and controls growth/diffraction. In a thin beam region, where the electron beam radius is less than the characteristic radius, we find that the radiation radius remains almost unchanged at the characteristic radius, and that the growth rate increases in proportion to the logarithm of the diminishing electron beam radius. The concrete density profile at a given beam radius introduces a minor effect on the growth rates. A center-humped profile is found to have a higher growth rate than a center-depleted profile. Their growth rate difference, however, is within a 10% range.

[FP1.107] Three-Dimensional Theory of Compton Scattering

A complete, three-dimensional theory of Compton scattering is described, which fully takes into account the effects of the electron beam emittance and energy spread upon the scattered x-ray spectral brightness. The radiation scattered by an electron subjected to an arbitrary electromagnetic field distribution in vacuum is first derived in the linear regime, and in the absence of radiative corrections; it is found that each vacuum eigenmode gives rise to a single Doppler-shifted classical dipole excitation. This formalism is then applied to Compton scattering in a three-dimensional laser focus, and yields a complete description of the influence of the electron beam phase space topology on the x-ray spectral brightness; analytical expressions including the effects of emittance and energy spread are also obtained in the one-dimensional limit. Within this framework, the x-ray brightness generated by a 25 MeV electron beam is modeled, fully taking into account the beam emittance and energy spread, as well as the three-dimensional nature of the laser focus; its application to x-ray protein crystallography is outlined. Finally, coherence, harmonics, and radiative corrections will also be briefly discussed.

[FP1.108] Nonlinear Laser Acceleration in a Magnetostatic Field and the Chirped-Pulse Inverse Free Electron Laser

A novel, high-gradient acceleration process is introduced, whereby electrons are interacting in vacuum with a femtosecond laser operation at relativistic intensities, and a static magnetic field. The interaction is best described as a three-stage process: first, the electron energy is boosted by the radiation pressure; at the peak energy, a transverse, static magnetic field is applied, which optimally dephases the electron and the wave; finally, the dephased electron is further accelerated in the second half of the pulse. The final energy is \gamma_+\infty/\gamma_-\infty \approx (1+A_0^2)^2, where \gamma_-\infty is the initial energy and A_0 is the normalized laser potential. The inverse free-electron laser(IFEL) interaction is studied both theoretically and numerically in the case where the drive laser intensity approaches the relativistic regime, and the pulse duration is only a few optical cycles long. A computer code which takes into account the three-dimensional nature of the interaction is in development and results will be presented as available.

[FP1.109] Ion and Electron Beams

[FP1.110] High-Power Testing of 11.424-GHz Dielectric-Loaded Accelerating Structures

Argonne National Laboratory has previously described the design, construction, and bench testing of an X-band traveling-wave accelerating structure loaded with a permittivity=20 dielectric (P. Zou et al., Rev. Sci. Instrum. 71, 2301, 2000.). We describe a new program to build a test accelerator using this structure. The accelerator will be powered by the high-power 11.424-GHz radiation from the magnicon facility at the Naval Research Laboratory ( O.A. Nezhevenko et al., Proc. PAC 2001, in press). The magnicon is expected to provide up to 30 MW from each of two WR-90 output waveguide arms in pulses of up to 1 microsecond duration, permitting tests up to a gradient of 40 MV/m. Still higher power pulses (100-500 MW) may be available at the output of an active pulse compressor driven by the magnicon ( A.L. Vikharev et al., Proc. 9th Workshop on Advanced Accelerator Concepts.).

[FP1.111] Development of a Durable, Large Area Cathode for Repetitive, Uniform Electron Beam Generation

Electra is an electron beam pumped, repetitively pulsed krypton fluoride (KrF) laser that will be used to develop the technology necessary for an inertial fusion energy (IFE) power plant. The laser is pumped with two opposing electron beams each with parameters of 500 kV, 100 kA, and 100 ns flat-top pulse duration. They are emitted from 27 x 97 cm^2 cathodes in a vacuum diode that is immersed in an external magnetic field of 0.14 T (about 2.5 times the beam self-field). The repetition rate is 5 Hz. Our goals are to develop a cathode that has: i) a current rise time of no more than 40-50 ns, ii) a diode gap closure velocity of less than 1 cm/\mus, iii) a beam current density variation of less than 10initial evaluation of emitters has concentrated on cold cathodes due to their inherent simplicity. The emission characteristics of eighteen different cold cathode materials were evaluated on Electra. Three cathodes - velvet dielectric fiber, carbon fiber, and carbon flock - that initially exhibited viable rise time and uniformity, were studied in detail using time resolved electrical and optical diagnostics. A fourth metal/dielectric-type cathode was also examined in light of its promising lifetime capability. Emission characteristics were observed for up to 10,000 shot runs to evaluate long term behavior.

[FP1.112] Influence of Low-Density Plasma on Multiple-Pulse X-Ray Radiography

In multiple-pulse X-ray radiography, an intense electron beam is focused onto a small area of a high-Z Bremsstrahlung converter causing vaporization and partial ionization of the local material. The plasma thus created can have deleterious effects on subsequent beam pulses due to beam-plasma instabilities, especially the interaction in the low-density region at the front of the expanding plasma plume. We employ an electron Monte Carlo code to compute the energy deposition in the target by the first beam pulse, which then serves as the initial condition for a subsequent hydrodynamic calculation of the target expansion. The profile of the low-density blow-off from the expanding plasma is estimated by a semi-analytic approximation. A particle-in-cell calculation is performed to model the plasma interaction with the second beam pulse. The transverse velocity spread of the beam is found to increase somewhat, but the beam stability and average spot size are not significantly affected. Other factors may also compromise the beam quality and are currently under investigation.

[FP1.113] The University of Maryland Electron Ring (UMER) - Progress Report

A detailed understanding of the physics of space-charge dominated beams is vital for many advanced accelerators that desire to achieve high beam intensity. In that regard, low-energy, high-intensity electron beams provide an excellent model system. The University of Maryland Electron ring (UMER), currently under construction, has been designed to study the physics of space-charge dominated beams with extreme intensity in a strong focusing lattice with dispersion. The tune shift in UMER will be more than an order of magnitude greater than in exiting synchrotrons and rings. The 10-keV, 100 mA, UMER beam has a generalized perveance in the range of 0.0015, and a tune depression of 0.12. Though compact (11-m in circumference), UMER is a very complex device, with over 140 focusing and bending magnets. We will report on the unique design features of this research facility, the beam physics to be investigated, simulation studies and first experimental results.

[FP1.114] Experiments on Energy Spread Evolution in Space-Charge Dominated Electron Beams

The University of Maryland electron ring (UMER) experiment, currently under construction, will be used to investigate the dynamics of space-charge dominated low energy (~10keV) electron beams. One point of interest is the energy distribution in the beam. We use an energy analyzer of cylindrical geometry to determine the longitudinal energy of the electron beam. The data can be resolved temporally, which gives an indication of the axial energy distribution in the beamlets. We also have the capability of scanning the beam in the transverse direction so that a lateral distribution of the longitudinal energy may be determined. We present results of energy spread measurements from the injector region of UMER under various conditions. Later efforts will include the effects of bending on the energy spread in the beam.

[FP1.115] Computer Simulation of the UMER Electron Gun

An important feature of the University of Marlyand Electron Ring Experiment (UMER) has been the continuing use of comparison between computer simulation and experimental measurement. Two- and three-dimensional simulations of the UMER electron gun are being conducted with the aims of understanding the space-charge-dominated physics in the gun and interpreting experimental data. These simulations will also be used to determine a realistic initial beam distribution for simulating beam dynamics during injection into the main ring. Preliminary simulations have revealed, even under relatively ideal conditions, a surprisingly rich set of features which appear to be characteristic of the beam distribution at the gun exit. Examples of the simulations and comparisons to experimental observation will be presented.

[FP1.116] Diode Perveance Study for High-Power Relativistic Electron Beam

One of the important issues in vacuum diodes is the increase of the perveance of the electron beam for various applications, including high-power microwave generation, transport of high current electron beams and production of intense bremsstrahlung. High perveance means high concentration of beam electrons, which in turn generates intense self-electric and magnetic fields. These intense self fields may produce high-power radiation by charge and current oscillations. In recent years, there is renewed interest on studies of the electron beam perveance. The fundamental limitation for the perveance is the space-charge-limited current. The electron beam current cannot flow beyond this limited current in normal conditions. The space-charge-limited current in a diode including the relativistic effects of beam electrons on it is investigated in connection with development of an analytical model of the perveance for high-power relativistic electron beams. A simple scaling law for perveance in high-power diode is obtained and compared with experimental data, which agree reasonably well with theoretical predictions.

[FP1.117] A Simple Theory on the Two-Dimensional Child-Langmuir Law

This paper presents a simple analytic theory for the two-dimensional (2D) Child-Langmuir Law. For electron emission over a strip of width W in a planar gap of gap separation D, the 2D limiting current density, J(2), is found to be related to the one-dimensional (1D) Child-Langmuir value, J(1), by J(2)/J(1) = 1 + D/(pW), under the assumption W/D >> 1. This scaling law turns out to be in excellent agreement with the simulation results even when W/D is as small as 0.1 [1]. For electron emission restricted to a circular patch of radius R, a similar analytic theory yields J(2)/J(1) = 1 + D/(4R). These results indicate that electrons emitted from a cathode over only a restricted area may have a current density much exceeding the classical (1D) Child-Langmuir value.

This work was supported by DUSD (S+T) under the Innovative Microwave Vacuum Electronics MURI Program, managed by AFOSR under grant F49620-99-1-0297, and by the Northrop-Grumman Industrial Affiliates Program.

[1] Luginsland, Lau, Gilgenbach, Phys. Rev. Lett. 77, 4668 (1996).

[FP1.118] Recent Results in High Current Density Surface Ionization and Plasma Ion Sources for Heavy Ion Fusion.*

Experimental results will be presented in the field of high current density (~100 mA/cm2), pulsed (multi micro-second scale), heavy mass (cesium, argon, etc.) ion sources. The goal of this research is to demonstrate and provide data required for the design and construction of heavy ion source and injectors to be used in future heavy ion fusion (HIF) experiments and drivers. Interest in high current density ion sources has originated from preliminary driver designs using the merging of multiple beamlets to make the required high current, space-charge dominated beams necessary in HIF. The paper presents current density yields obtained for surface ionization sources with different ionization substrates (tungsten, iridium) and ion species (potassium, cesium). Relevant pulse characteristics and considerations regarding neutral loss and its impact on lifetime will be discussed. Results regarding pulse characteristics (risetime, pulse length and temporal uniformity) obtained from a preliminary noble gas plasma source experiment will also be presented, as well as technical details and preliminary results from the next series of plasma source experiments.

* Work supported by US DOE under contract number DE-AC03-76SF00098 (LBNL) and W-7405-ENG-48 (LLNL).

[FP1.119] Heavy-Ion Injector for the High Current Experiment

We report on progress in development of the Heavy-Ion Injector at LBNL, which is being prepared for use as an injector for the High Current Experiment (HCX). It is composed of a 10-cm-diameter surface ionization source, an extraction diode, and an electrostatic quadrupole (ESQ) accelerator, with a typical operating current of 0.6 A of potassium ions at 1.8 MeV, and a beam pulse length of 4.5 microsecs. We have improved the Injector equipment and diagnostics, and have characterized the source emission and radial beam profiles at the diode and ESQ regions. We find improved agreement with EGUN predictions, and improved compatibility with the downstream matching section. Plans are to attach the matching section and the initial ESQ transport section of HCX. Results will be presented and compared with EGUN and WARP simulations.

[FP1.120] Progress in developing high current injectors for Heavy Ion Fusion

For heavy ion fusion (HIF) induction linac drivers, a typical injector requires total beam current of 50-100 A and is comprised of many individual beams of ¡Ö 0.5 A each. For many years, the HIF program has been exploring using a single large diameter source for each beam. The research has focused on alkaline metal surface ionization sources because of the ion temperature constraint. However, simulation and design studies in the last two years have shown that the method of merging an array of high current density beamlets can significantly reduce the size, and likely the high cost, of a HIF driver scale injector and therefore creates an affordable development path for HIF drivers. Furthermore, it widens the choice of ion sources to include plasma sources, and also possibly relaxes the beam pulse rise time control. Experiments are being planned to test this new concept on a new 500 kV injector test stand at LLNL. We will report the recent progress on injector experiments at the VNL, including both single source per beam and many beamlets per beam research.

[FP1.121] Progress on the High Current Experiment for Heavy-Ion Fusion

The High Current Experiment has been designed to test transport issues near the low energy end of a typical induction linac driver for Heavy Ion Fusion. The experiment is being assembled in phases and employs a single coasting beam characteristic of recent conceptual design drivers. A major goal of the experiment is the evaluation of the maximum acceptable aperture filling factor and transverse phase space evolution, as influenced by the initial state distribution function, beam steering and quadrupole alignment sensitivity, envelope mismatch, halo, and secondary electrons. A thorough measurement, within the limits of the diagnostics available, of the beam distribution function produced by the injector and matching section will be presented along with early commissioning results from the electrostatic transport quadrupoles and diagnostics.

[FP1.122] PIC Simulations for the High Current Experiment

Experiments are beginning on the Heavy Ion Fusion High Current Experiment (HCX). The HCX will explore transport issues using one driver-scale 600 mA, 3-7 microsecond, 1.8 MeV coasting K+ beam transported through 40 electrostatic quadrupoles followed by 4 magnetic quadrupoles. The beam physics is dominated by beam space charge forces, and, at the large beam-pipe filling factors of interest, by image forces and focusing field nonlinearities. We present the results of 2D transverse PIC simulations. Simulations have been done both for an idealized (semigaussian) distribution function, and for a self-consistent distribution simulated from the anode. Effects addressed will include nonideal features of the distribution function, physical processes which limit the amount of usable aperture, matching of the beam to the focusing lattice, and generation of beam halo due to mismatch.

[FP1.123] Expectations for electron buildup in HCX

We evaluate likely levels of electron contamination in the magnetic quadrupole sections of a heavy-ion accelerator, with particular emphasis on estimates for HCX. Sources include ionization of background gas, and secondary electron emission and ionization of neutrals emitted when stray beam ions strike the wall. Other relevant factors include trapping due to nonadiabatic effects, electron-electron two stream instability, ExB drift out of the quadrupole, ionization and charge exchange rates for neutral gas on the ion beam, and wall conditioning. We find that ionization of emitted neutrals is the most important source, and the ionization and charge exchange rates and wall conditioning play key roles in determining the electron density. HCX should be a good vehicle for testing electron physics, since without special preparation a large fractional electron population is predicted, but wall conditioning and other mitigation techniques should result in substantial reduction. The importance of ExB drift as a loss mechanism depends on the fate of electrons in fringe fields; studies of this using the WARP code will be shown.

[FP1.124] Electron interactions with a heavy-ion beam in HCX quadrupole magnetic fields – experimental planning

Barely trapped electrons can originate from beams or UV radiation impinging on walls, and deeply trapped electrons can originate from beam ionization of gas. Electrons are electrostatically trapped by the beam and can reduce its potential from the nominal ~5 kV. Since the quadrupole transport is primarily balancing the beam space charge, a change in potential causes a mismatch in the beam transport which could vary both spatially and temporally. We describe plans for diagnostics to measure the electrons, and their effects, and for experiments to determine the scaling and relative importance of the sources and of the effects on ion beam dynamics using a 1.8 MeV, 0.6 A K+ heavy-ion beam. Proposed diagnostics include a low-frequency interferometer, capacitive probes, ion/electron collectors, Faraday cups, gas injection and pressure measurement, energy distribution of expelled ions from gas ionization, and electron trapping from the change in energy of electrons during a single pass.

[FP1.125] Nonlinear \delta F Simulation Studies of Intense Charged Particle Beams with Large Pressure Anisotropy

In plasmas with strongly anisotropic distributions (T_||b/T_\bot b\ll 1) a collective instability may develop if there is sufficient coupling between the transverse and longitudinal degrees of freedom. Such anisotropies develop naturally in accelerators and may lead to a deterioration of beam quality. The kinetic and fluid descriptions predict instability for different ranges of beam intensity. In this paper, a 3D nonlinear perturbative particle simulation code (BEST) is used to systematically study the stability properties of intense nonneutral charged particle beams with pressure anisotropy \left( P_\perp b >P_\parallel b \right) . The most unstable modes are identified and their eigenfrequencies and radial mode structure are determined for axisymmetric perturbations with \partial/\partial \theta=0. The simulation results clearly show that moderately intense beams with s_b=\widehatømega _pb^2/2\gamma_b^2ømega_\beta b^2 \geq 0.5 are linearly unstable to short wavelength perturbations with k_z^2 r_b^2 \geq 1 , provided the ratio of longitudinal to transverse temperatures is smaller than some threshold value. In the nonlinear saturation stage, the total distribution function is still far from equipartitioned, and free energy is available to drive an instability of the hydrodynamic type.

[FP1.126] Effects of Dispersion During Drift Compression in a Heavy Ion Fusion Driver

Immediately prior to final focus onto a fusion target, heavy ion driver beams are compressed in length by typically an order of magnitude. This process is simultaneous with bending through large angles to achieve the required target illumination configuration. The large increase in beam current is accommodated by a combination of decreased lattice period, increased beam radius, and increased strength of the beamline quadrupoles. However, the large head-to-tail velocity tilt (up to 15%) needed to compress the pulse results in a very significant dispersion of the pulse centroid from the design axis. A principal design goal is to minimize the magnitude of the dispersion while maintaining approximate first order achromaticity through the complete compression/bend system. Configurations of bends and quadrupoles which achieve this goal while simultaneously maintaining a locally matched beam-envelope have been analyzed. Approaches using dispersion matching and adiabatic turn-on of the bends are compared.

[FP1.127] Self-pinched chamber transport of heavy ion beams

Self-pinched heavy ion beams are being examined as a chamber transport scheme for heavy-ion-driven inertial confinement fusion. In this scheme, beam-impact-ionization of a low-density background gas provides neutralizing electrons. For certain ranges of background gas pressures the beam is essentially charge-neutralized but incomplete current-neutralization allows the self-magnetic field of the beam to act as a pinch force, confining the beam divergence. Equilibrium transport modes for a Pb^+65 ion beam propagating through low density Xe gas are being studied with particle-in-cell simulations using the LSP code [1]. Time dependent evolution of the beam net current and pinched beam radius as a function of Xe chamber pressure from the simulations is examined.

[1] T. P. Hughes, R. E. Clark, and S. S. Yu, Phys. Rev. ST-AB 2, 110401 (1999); D. R. Welch, D. V. Rose, B. V. Oliver, and R. E. Clark, Nucl. Inst. Meth. Phys. Res. A 242, 134 (2001).

[FP1.128] ECR Plasma Source for Heavy Ion Beam Charge Neutralization

Highly ionized plasmas are being considered as a medium for charge neutralizing heavy ion beams in order to focus beyond the space-charge limit. Calculations suggest that plasma at a density of 1 - 100 times the ion beam density and at a length \sim 0.1-2 m would be suitable for achieving a high level of charge neutralization. An ECR source has been built at PPPL to support a joint Neutralized Transport Experiment (NTX) at LBNL to study ion beam neutralization with plasma. The ECR source operates at 13.6 MHz and with solenoid magnetic fields of 5-20 gauss. The goal is to operate the source at pressures \sim 10^-6 Torr at full ionization. The initial operation of the source has been at pressures of 10^-4 - 10^-1 Torr. Electron densities in the range of 10^8 - 10^11 cm^-3 have been achieved. Low-pressure operation is important to reduce ion beam ionization. A cusp magnetic field will be considered to improve ionization and reduce field on the beam axis.

[FP1.129] Simulation of "Foot" Pulses for Heavy-Ion Fusion

Indirect-drive targets for heavy-ion fusion must initially be heated by "foot'' pulses that precede the main heating pulses by tens of nanoseconds. These pulses typically have a lower energy and perveance than the main pulses, and the fusion-chamber environment is different from that seen by later pulses. The preliminary particle-in-cell simulations of foot pulses reported here examine the sensitivity of the focal-spot radius to beam perveance, background-gas pressure, and pre-neutralization by a plasma near the chamber entry port.

[FP1.130] Multiple electron stripping of Heavy Ion Beams

An approach being explored as a route to fusion energy uses heavy ion beams focused upon an indirect drive target. Such beams will lose electrons as they pass through background gas in the target chamber. It is necessary to assess the rate at which the charge state of the incident beam evolves while passing through the background gas. Present accelerators utilize already highly stripped ions in order to achieve high energies economically and no accelerators capable of producing heavy ion beams of 10 to 20 Mev/amu with charge state 1 currently exist. Hence, theoretical models have been developed to predict the performance of heavy ion fusion driver beams. We have investigated the stripping of 3.4 Mev/amu Kr 7+ and Xe 11+ in nitrogen. The multi-electron loss events have been observed, namely the average number of electrons lost per single collision is 1.87 for Kr 7+ and 1.97 for Xe 11+. The experimental data agree with theoretical predictions.

[FP1.131] Electromagnetic Darwin Model for High Intensity Beams

This paper explores the advantages of using the canonical momentum, P \equiv p + q A/c, in the laboratory frame for the electromagnetic Darwin model of the Maxwell-Vlasov system, for which the transverse displacement current is neglected in Ampere's law. Here, A is the vector potential and p is the mechanical momentum. With the new formulation, we not only can eliminate the high-frequency transverse electromagnetic effects from the system as intended by the Darwin model, but also can get rid of the numerically troublesome \partial A / \partial t term for particle pushing.~(C. W. Nielson and H. R. Lewis, Methods in Computational Physics) (Academic Press, New York, 1976), Vol. 16, p. 976. The formulation is most useful for studying space-charge effects in high-intensity multi-species beams using perturbative particle simulation methods,~(W. W. Lee, Q. Qin and R. C. Davidson, Phys. Lett. A 230), 347 (1997). such as two-stream~(H. Qin, R. C. Davidson and W. W. Lee, Phys. Rev. Special Topics on Accelerators amd beams 3), 08441 (2000); 3, 109901 (2000). and filamentation instabilities, that may cause deterioration of the beam quality in heavy ion fusion drivers and the fusion chamber. Moreover, the new formulation provides an easy link to the electrostatic model for a single-species beam in the beam frame. Initial numerical results will be reported.

[FP1.132] Control of beam halos through nonlinear transport

Beam halo formation is an important issue in various applications of high intensity charged particle accelerators. This is because the halo particles hit the walls of the accelerator causing radioactivation. In this presentation we discuss the use of nonlinear focusing as a method to control beam halos. Results will be shown from simulations performed by a one dimensional PIC code with continous focusing. Comparisions are made for equivalent beams under linear and nonlinear focusing. Our results show that nonlinear focusing can mitigate halo formation. We will discuss results for three types of initial distributions The first a regular Gaussian distribution in phase space, the second a mismatched KV beam and finally a beam close to thermal equilibrium where the distribution function is given by f = f_o\exp(-H/H_o). Our simulation results also show that the beam ocillations get damped in the presence of nonliner focusing.

[FP1.133] Symmetric Plasma Trapping, Acceleration, and Transport

Symmetric plasmas composed of positive and negative ions of equal mass lend themselves to applications in fusion energy research, plasma processing, and ion propulsion. Reported here are new aspects of the trapping, acceleration, and transport of such plasmas. For trapping and acceleration, electric quadrupoles driven at radio frequencies (RFQs) are employed. The trapped particle density scales inversely with quadrupole aperture size; the plasma density is increased by shrinking the quadrupole, and total beam currents are increased by paralleling small quadrupoles in place of a large one. The acceleration process breaks the beam into alternating bunches of positive and negative ions; an additional RFQ can debunch and neutralize the beam. The trapping scaling theory as well as experimental and computational results are presented. Also, a theoretical basis for the ability of a sufficiently dense symmetric plasma beam to transport across a magnetic field and pertinent simulation results are presented. This work is supported by DOE through contract DE-FG03-00ER54575 and the Fusion Energy Sciences Fellowship.

[FP1.134] Self-focusing of a modulated electron-positron beam in dense plasmas

We have found the necessary conditions for the electrostatic equilibrium of a stratified electron-positron beam, which is modulated at a specific frequency smaller than that of the surrounding plasmas. The beam self-focusing effect [1-5] occurs for both electron and positron species. It remains valid for the case of a strong axial magnetic field as well. In the presence of a radial magnetic field Lorentz force drags particles off the potential wells to the beam periphery. However, rising polarization field stabilizes it. As a result, axial transport of the electron-positron beam becomes much more favorable if compared to the case of an equivalent sequence of pure electron bunches in vacuum.

[1] V.B.Krasovitskii, Sov. JETP Lett., v.9, 422, 1969; [2] F.Winterberg, Bull. Amer. Phys. Soc., v.11, 1453, 1970; Atomkernenergie, v.22, 142, 1973; [3] V.P.Kovalenko and P.N.Yushmanov, Sov. J. Plasma Phys., v.3, 714, 1977; [4] V.D.Dorofeenko and V.B.Krasovitskii, Sov. Phys. JETP, v.72, 278, 1991; [5] O.V.Batishchev, V.B.Krasovitskii et al., Plasma Phys. Rep., v.19, 379, 1993.

[FP1.135] Hamiltonian Formalism for Solving the Vlasov-Poisson Equations and its Application to the Coherent Beam-Beam Interaction

A Hamiltonian approach to the solution of the nonlinear Vlasov-Poisson equations has been developed. Based on a nonlinear canonical transformation, the rapidly oscillating terms in the original Hamiltonian are transformed away, yielding a new Hamiltonian that contains only slowly varying terms. The essential feature of this formalism is the use of mixed canonical variables (old coordinates and new canonical momenta). The key point is that we transform the Vlasov equation in the mixed set of variables only, leaving unchanged the Poisson equation. This makes the method simpler and more efficient in a computational sense. Unlike the standard canonical perturbation technique, the approach developed here avoids many lengthy and cumbersome calculations. The formalism has been applied to the coherent beam-beam interaction, and a stationary solution to the transformed Vlasov equation has been obtained.

Part F of program listing