Implicit low-noise particle simulation methods

Particle simulation of kinetic plasma phenomena is an important tool in numerous application areas: magnetic and inertial fusion, space and astrophysical plasmas, high power microwave devices, and plasma processing. Recent innovations in implicit techniques, e.g., quiet implicit PIC and implicit δf, involving the use of both fluid and particle equations to achieve greater efficiency in the control of unwanted high frequency modes and particle noise, are being exploited in two distinct examples reported here. We are extending an implicit gyrokinetic δf model to accommodate drift-kinetic electrons and electromagnetic effects for application to drift-wave turbulent transport simulation. Implicit equations for both the scalar electric potential and the parallel vector potential are coupled iteratively. We report our progress in implementing this new algorithm in a two-dimensional slab testbed code suitable for studies of ion-temperature-gradient and collisionless drift instabilities with electromagnetic modification. We have also developed a one-dimensional general-hybrid open-quotes quiet-δfclose quotes code. The distribution function is split into a local Maxwellian and a kinetic remainder. At each timestep, the resulting fluid-equation components of the system are evolved with a standard fluid push, while the kinetic δf remainder is evolved via a direct nonlinear δf-particle push. The code correctly reproduces the frequencies and growth and damping rates of plasma oscillations and beam-plasma instabilities. The code is useful, both for one-dimensional applications and as a test-bed code for multiple-timescale schemes that are expected to be useful in magnetic-fusion-plasma edge-transport and core-turbulence simulations, as well as in space-plasma and ICF simulation. The results of the above test cases and from comparisons of the quiet-δf algorithm with a standard nonlinear-δf algorithm and with standard PIC will be shown