Nonlinear hybrid Boltzmann-particle-in-cell acceleration algorithm

Kinetic simulation of plasmas in which equilibrium occurs over ion time sca les poses a computational challenge due to disparity with electron time sca les. Hybrid electrostatic particle-in-cell (PIC) algorithms are presented i n which most of the electrons reach thermodynamic equilibrium [Maxwell-Bolt zmann (MB) distribution function] each time step. Conservation of charge en ables convergence of the nonlinear Poisson equation. Energy conservation is used to determine the temperature of the Boltzmann species. This article f irst develops an algorithm where all the electrons have a MB energy distrib ution, either with a full MB distribution or with a truncation of the high energy tail. Second, high energy PIC electrons are added to the truncated d istribution so that high energy electrons are modeled kinetically by PIC an d low energy electrons (the majority) are modeled by the MB distribution. C ollisions for PIC electrons are included via a Monte Carlo model, while for the MB electrons, the distributions are integrated with energy dependent c ross sections. The MB model is not constrained by the electron time scale w hich decreases the required computer time by about the square root of the m ass ratio of ion to electron. However, the hybrid boundary conditions are m ore complex and the simulation is not quite self-consistent. Comparison bet ween full PIC and the PIC-MB hybrid is made for simulations of photo-ionize d sustained discharges and current-driven dc discharges. (C) 2000 American Institute of Physics. [S1070-664X(00)03408-X].