Quiet direct simulation Monte Carlo (QDSMC) is a new particle simulation technique that is applicable to a broad range of applications where the underlying system dynamics obey Fokker Planck equations. These include hydrodynamics, radiation transport, magnetohydrodynamics, diffusion, and collisional kinetic plasmas. At the beginning of each time step in QDSMC, the weights and abscissas of Gaussian-Hermite quadrature are used to deterministically create particles to sample the random process. At the end of the time step, particles are gathered to the computational mesh to obtain updated distributions of conserved quantities on the mesh and then the particles are destroyed. The creation and destruction of particles allows arbitrary dynamical range to be accessed quiescently with only a small number of particles per computational cell. The application of QDSMC to the simulation of Coulomb collisions is considered in this report, and the method is demonstrated on problems involving the collisional relaxation of non-Maxwellian distributions.

A new approach to particle simulation, called “quiet direct simulation Monte Carlo” (QDSMC), is described that can be applied to many problems of interest, including hydrodynamics, magnetohydrodynamics (MHD), and the modeling of collision plasmas. The essence of QDSMC is the use of carefully chosen weights for the particles (e.g., Gauss–Hermite, for Maxwellian distributions), which are destroyed each time step after the particle information is deposited onto the grid and reconstructed at the beginning of the next time step. The method overcomes the limited dynamical range and statistical noise typically found in particle simulations. In this article QDSMC is applied to hydrodynamics and MHD test problems, and its suitability for modeling semi-collisional plasma dynamics is considered.

The direct simulation Monte Carlo method of modeling fluids requires sampling one or more random variables every time step for each particle. In this paper a "quiet Monte Carlo" technique is proposed that eliminates the random sampling and the noise it produces by deterministically generating a small number of computational particles. The technique is applied to particle equations of motion appropriate for modeling an Eulerian fluid. Results indicate that strong one- and two-dimensional shocks with large dynamic ranges are accurately represented with only a few particles per cell.

Three material properties experiments that are to be performed on the Atlas pulsed power facility are described; friction at sliding metal interfaces, spallation and damage in convergent geometry, and plastic flow at high strain and high strain rate. Construction of this facility has been completed and experiments in high energy density hydrodynamics and material dynamics will begin in 2001.

CO chemisorption on Bi-modified Ni(100) surfaces, along with the structure and growth of vapor-deposited Bi adlayers on Ni(100), was characterized by Auger electron spectroscopy (AES), temperature-programmed desorption (TPD), low-energy electron diffraction (LEED), energy loss spectroscopy (ELS), UV photoelectron spectroscopy (UPS), work function measurements, and high-resolution electron energy loss spectroscopy (HREELS). Bi growth on Ni(100) at 500 K proceeds via a layer-plus-island (Stranski - Krastanov) growth mode and the gradual formation of a c(2 x 2) structure near monolayer coverage. Desorption of Bi from the first monolayer on Ni(100) occurs with an activation energy E-d = 290 = 240 kJ mol(-1). Bi desorption from Bi multilayers has E-d = 200 kJ mol(-1). Adsorbed Bi changed the work function of the Ni(100) surface only slightly, indicating an initial dipole moment of only -0.5 D and thus relatively little charge transfer between Bi and Ni compared to other modifier adlayers. CO chemisorption was used to probe the reactivity of Ni(100) surfaces modified by preadsorbed Bi adlayers, denoted as Bi/Ni(100). Only a small decrease (4 kJ mol(-1)) occurs for the CO adsorption energy as determined by CO TPD. Site-blocking effects dominate over electronic (ligand) effects on the surface chemistry of CO on Bi/Ni(100). A comparison of these results to those on Bi/Pt(111), where Bi has been used as a model inert site-blocking agent, indicates that Bi modifies the electronic structure of Ni(100) even less than on Pt(111). Therefore. Bi adatoms may allow useful probing of adsorption and reaction ensemble requirements on Ni surfaces that contain modifiers as adatoms.

The enhancement of energy loss of short pulses of electrons in plasma has been experimentally observed. An enhancement of up to 3.5×104 over single-particle losses was observed when a 15 ps electron bunch was injected into a preformed target plasma with an electron density of approximately 1013 cm−3. This matches the theoretical prediction that the energy loss should be enhanced when the temporal duration of the electron bunch is approximately equal to π/ωpe, where ωpe is the plasma frequency of the target plasma. 2D numerical simulations are in good agreement with the observations.

We quantify a model which incorporates observed features of contaminant particle growth in plasma processing reactors. According to the model, large ‘‘predator’’ particles grow by adsorbing smaller, typically neutral, ‘‘prey’’ protoparticles. The latter are supplied by an assumed constant mass injection of contaminant material. Scaling laws and quantitative predictions compare favorably with published experimental results.

Ion interpenetration, stagnation, and energization processes are studied in colliding laser-produced plasma configurations relevant to Trident [R. G. Watt, Rev. Sci. Instrum. 64, 1770 (1993)] experiments using four different numerical methods: one-dimensional Monte Carlo and Lagrangian multifluid codes, and one- and two-dimensional hybrid (particle ions, fluid electrons) and single-fluid Lagrangian codes. Results from the four methodologies are compared for plasmas generated with gold and deuterated polyethylene (CD2) targets. Overall, the various codes give similar results concerning the initial expansion of the plasmas and their collisional interaction, the degree of stagnation, stagnation time, and amount of ion thermalization for gold targets, while multispecies techniques indicate a much softer stagnation for CD2 plasmas than the single-fluid model. Variations in the results of the calculations due to somewhat different initializations and parameters, as well as to different physics in the codes, are discussed.

A new method is presented to model the intermediate regime between collisionless and Coulomb collision dominated plasmas in particle-in-cell codes. Collisional processes between particles of different species are treated through the concept of a grid-based “collision field,” which can be particularly efficient for multi-dimensional applications. In this method, particles are scattered using a force which is determined from the moments of the distribution functions accumulated on the grid. The form of the force is such to reproduce the multi-fluid transport equations through the second (energy) moment. Collisions between particles of the same species require a separate treatment. For this, a Monte Carlo-like scattering method based on the Langevin equation is used. The details of both methods are presented, and their implementation in a new hybrid (particle ion, massless fluid electron) algorithm is described. Aspects of the collision model are illustrated through several one- and two-dimensional test problems as well as examples involving laser produced colliding plasmas.

We identify, analyze, and propose remedies for a numerical instability responsible for the growth or decay of sums that should be conserved in Monte Carlo simulations of stochastically interacting particles. ''Noisy'' sums with fluctuations proportional to 1/root n, where n is the number of particles in the simulation, provide feedback that drives the instability. Numerical illustrations of an energy loss or ''cooling'' instability in an Ornstein-Uhlenbeck process support our analysis.

Colliding Au, CD, and Ti-CR plasmas have been generated by illuminating two opposing foils each with an approximately 100J, 0.5 nsec, 2(omega) Nd-glass laser beam from the Trident laser facility at Los Alamos. The plasmas are being used to study plasma interactions which span the parameter regime from interpenetrating to collisional stagnation. X-ray emission during the laser target interaction and the subsequent collision is used to diagnose the initial plasma conditions and the colliding plasma properties. X-ray instrumentation consists of a 100 ps gated x-ray pinhole imager, a time-integrated bremsstrahlung x-ray spectrograph and a gated x-ray spectrograph used to record isoelectronic spectra from the Ti-Cr plasmas. The imager has obtained multiframe images of the collision and therefore, a measure of the stagnation length which is a function of the ion charge state and density and a strong function of the electon temperature. Other isntrumentation includes a Thomson scattering spectrometer with probe beam, neutron detectors used to monitor the CE coated foil collisions, and an ion spectrometer. We will describe the current status of the experiments and current results with emphasis on the x-ray emission diagnostics. We will also briefly describe the modeling using Lasnex and ISIS, a particle-in-cell code with massless fluid electronics and inter-particle (classical) collisions.

The drift of plasma ions relative to charged grains in a dusty plasma can give rise to a dust/ion acoustic instability. We investigate the linear properties of the instability by numerically solving an appropriate linear dispersion equation and examine the nonlinear behavior through one‐dimensional electrostatic particle simulations, in which the plasma and dust ions are treated as discrete particles and the electrons are modeled as a Boltzmann fluid. The instability saturates by trapping some of the plasma ions. The instability is slightly weaker when the dust particles have a range of sizes, and corresponding range of charges and masses. It is argued that due to dust particles that comprise planetary rings, this process can contribute to ion heating and diffusion observed in the inner magnetosphere of Saturn.

Most of the proposed Extreme Ultraviolet Lithography (EUVL) sources utilize the interaction of some high energy density source, i.e. a laser, electron beam, or arc discharge, with a solid target. EUV can be efficiently generated by these techniques, but debris from the interaction region impacting on the reflective optics can substantially degrade the performance of the condenser system. Los Alamos and Grumman are cooperatively investigating a debris free source 1,2 for EUV lithography. This source utilizes the predicted anomalous energy loss 3,4 of a short pulse electron beam in a preformed plasma to heat and ionize the ions to a charge state where efficient radiation at 130Å occurs. Accelerators 5 developed for the free electron laser program at Los Alamos are used as the electron bunch source. These accelerators use a laser driven photocathode to produce 15 psec electron bunches containing 4 nC of charge with an energy of 15.5 MeV. These micropulses are produced at a repetition rate of 108 MHz and continue for the length of the rf macropulse energizing the accelerator cavities, typically 1-10 μsec. The weakly ionized preformed plasma is created by purely classical collisional ionization caused by the initial few electron micropulses within the macropulse. When a critical electron density is reached, n e ≅ 1.6x10 15 /τ 2 cm -3 , where τ is the FWHM, in psec, of the electron bunch, the plasma responds collectively to the electron micropulse generating a large amplitude plasma wave. The plasma wave efficiently slows the high energy electron beam while heating the background plasma electrons. The initial electron population rapidly heats and then equilibrates with the bulk ion and electron populations in a few 10's of picoseconds. With neon as the dominant ionic species, an efficient filamentary radiator of line radiation near 130Å is created.

A general approach for including lumped circuit elements in a finite difference time domain (FDTD) solution of Maxwell's equations is presented. The methodology allows the direct access to SPICE to model the lumped circuits, while the full 3-dimensional solution to Maxwell's equations provides the crosstalk and dispersive properties of the microstrips and striplines in the circuit.<>

Los Alamos National Laboratory has assembled an array of experimental and theoretical tools to optimize amplifier design for future single-pulse KrF lasers. The next opportunity to exercise these tools is with the design of the second-generation NIKE system under construction at the Naval Research Laboratory with the collaboration of Los Alamos National Laboratory. Major issues include laser physics (energy extraction in large modules with amplified spontaneous emission) and diode performance and efficiency. Low cost is increasingly important for larger future KrF single-pulse systems (low cost and high efficiency is important for larger repetitively pulsed applications such as electric power production). In this article, we present our approach to amplifier scaling and discuss the more important design considerations for large single-pulse KrF amplifiers. We point out where improvements in the fundamental database for KrF amplifiers could lead to increased confidence in performance predictions for large amplifiers and address the currently unresolved issues of anomalous absorption near line center and the possibility of diode instabilities for low-impedance designs. Los Alamos has applied these amplifier design tools to the conceptual design of a 100-kJ Laser Target Test Facility and a 3-MJ Laboratory Microfusion Facility.

Particulate contaminants in rf sustained plasma processing reactors remain a significant problem. Laser scanning measurements have demonstrated that these dust particles migrate to well-defined regions of the discharge; the traps have characteristic, long-lived boundaries. The authors analyze in detail the various forces on the dust: electric, gravitational, neutral-dust drag, and ion-dust drag, including the effects of the size of the dust grains compared to their relative spacing and the Debye length. The dynamics of dust particles subject to these forces are investigated using one- and two-dimensional fluid discharges models to determine the steady state macroscopic parameters. The dust tends to reside at the edge of the sheath where the dominant electric and ion-dust drag forces balance, producing structures similar to those observed.

The transverse wake field in a dielectric wake-field accelerator for drive beams with a small nonaxisymmetric component (l not-equal 0) has been calculated in the ultrarelativistic limit. It is found that for the lth azimuthal mode, where l > 0, the transverse component is comparable to the longitudinal component. This implies that these devices will be subject to a beam-breakup instability as the drive current becomes large.

A new moving coordinate particle-in-cell model has been developed as an option in the code ISIS, in an attempt to optimize accelerator transport modeling. Each particle is assigned a charge weight and is tracked in three-dimensional momentum and real space. Three-dimensional fields are represented on a series of cylindrical r−z grids, one for each azimuthal Fourier component (l-value). The meshes are redefined at each time-step, based on the location of the particles. This allows optimum use of the gridding to represent fields only in the vicinity of the particles. For axisymmetric external fields, such as solenoids and rf cavity fields, a rational approximation is used to represent the axial fields. Expansions based on requiring the fields obey Maxwell's equations are used to determine the off axis fields. Space charge effects are approximated by solving Poisson's equation in the rest frame of the beam on a nonuniform mesh which extends beyond the particle mesh to include boundary conditions. The particle push is the usual fully-centered second-order accurate Boris method using both external and self-fields simultaneously in the lab frame. The model has been applied to the Boeing APLE experiment.