Plasma Theory and Simulation 

John Verboncoeur | |


The Plasma Theory and Simulation Group engages in theoretical and computational plasma physics research, including algorithm, model, and code development, with a broad range of applications. Our most recent, popular, and well-kept-up codes are on bounded plasma, plasma device codes XPDP1, XPDC1, XPDS1, and XPDP2. The P, C, and S mean planar, cylindrical, or spherical bounding electrodes; the 1 means 1d 3v and the 2 means 2d 3v. These are electrostatic, may have an applied magnetic fi eld, use many particles (hundreds to millions), particle-in-cell (PIC), and allow for collisions between the charged particles (electrons and ions, + or –) and the background neutrals (PCCMCC). The electrodes are connected by an external series R, L, C, source circuit, solved by Kirchhoff ’s laws simultaneously with the internal plasma solution (Poisson’s equation). The source may be V(t) or I(t), may include a ramp-up (in time). XPDP2 is planar in x, periodic in y, or fully bounded in (x, y), driven by one or two sources.

The Plasma Theory and Simulation Group (PTSG) code suite is distributed for free on the internet, including full source code, and has provided the basis for models incorporated in codes around the world. The code suite has been used by more than 1,000 researchers worldwide, and played a key role in over 350 research publications in the past decade. PTSG pioneered models for bounded plasmas, including self-consistent circuit boundary conditions. It also pioneered models for collisional plasmas, including multiple background gases and multi-species chemistry. PTSG also pioneered the development of object-oriented models for plasma codes, extending lifetime and enabling dynamic model exchange. 

Kinetic global model

PTSG devised a new model for the rapid modeling of large numbers of species and reactions in high pressure plasmas, called the kinetic global model. This is a spatially independent model in which the reaction rates are determined by convolution of energy dependent cross sections with a parameterized distribution function.

High Performance Computing (HPC)

Models for massively parallel systems, including multi-core/multi-CPU clusters, GPU clusters, and specialized supercomputers with advanced programmable cache and memory topology capabilities. An example of HPC research is the study of partial sorting on the null collision model on GPU systems, where the use of bandwidth within GPU processors is crucial in performance. 

High voltage multipactor and breakdown

These study range from high vacuum multipactor in superconducting RF (SRF) cavities, to high power microwave breakdown in air. In the resonant multipactor process, electron impact in the range ~100 eV to a few keV can result in ejection of more than one secondary electron per impact, so that trajectories that result in resonant impacts in this range can lead to a rapidly growing electron current, and ultimately surface heating, damage, and desorption of atoms which can become ionized and do damage (Fig. 1). A similar sequence occurs with background gases, but quickly transitions into a plasma discharge and potentially an arc. These breakdown eff ects are oft en the limiting factor in high voltage devices ranging from accelerators to microwave sources for radar, as well as insulators on power transmission lines, high voltage pulsed systems, and many more. 

Plasma assisted combustion

PTSG is developing techniques for infl uence the electron energy distribution function by preferentially heating one part of the distribution, leading to a bias for one set of reactions over another (Fig. 2). This can be used to select reaction chains for more effi  cient fuel cracking and combustion, as well as selecting chains which lead to lower pollutant production. Current eff orts focus on the relatively simple atomic reactions of methane, and aft er demonstration of the technique, will move on to more complex fuels including gasoline, diesel, and biofuels. Biofuels, in particular, can be diffi  cult to combust eff ectively in standard conditions, but plasmas have proven an eff ective way to improve the process.

Microdischarges and dielectric barrier discharges

A variant of low-temperature plasmas includes high-pressure nonequilibrium discharges. This family of discharges includes atmospheric pressure non-equilibrium microdischarges, as well as dielectric barrier discharges (Fig. 3). These discharges are oft en ~100 microns in spatial extent, but have very high peak densities, and are characterized by sharp density and eff ective temperature gradients that have non-local properties due to intense driving fi elds localized to sheath regions.

Strongly coupled plasmas

Strongly coupled plasmas, including ultra-cold plasmas, are an interesting new class of plasmas in which the potential energy of the system exceeds the kinetic energy. These plasmas exhibit properties similar to liquids and solids. Interesting applications of this class of plasmas include formation of dynamic micro-structures, such as rapidly directed antennas, multi-directional optical switches.