The low earth orbit plasma experienced by exposed interconnect-dielectric junctions commonly found on spacecraft solar panel surfaces was modeled using a fully kinetic particle-in-cell (PIC) simulation of both ambient ions and electrons. From time-accurate simulations, we observed that the plasma sheath (what is plasma sheath?) had a formation time somewhere between the ion and electron time scales of 17 μs and 30 ps, respectively, and electron and ion velocity distribution functions were observed to be highly non-Maxwellian. Comparison of the electron plasma sheath thickness with analytic cylindrical sheath models gave reasonable agreement if the embedded biased interconnect voltage was sufficiently high to cause the dielectric surface to act as a free electron flowing medium. Finally, it was found from the fully kinetic PIC simulations that the fundamental mechanism behind parasitic current is closely related to electron avalanche (LHS figure) and the fraction of SEE emitted from different avalanche levels as well as the current collected at the interconnect could be modeled by a power law series for avalanche levels greater than two.

We present a 3D-particle-in-cell (PIC) approach to modeling electrospray plumes typical of those formed by externally wetted emitter devices. Numerical grid-resolution techniques suitable for capturing strong electric fields in the emitter region were explored, and grid refinement criteria were quantified. The molecular dynamics simulations of the EMIM−BF4 ionic liquid system were modeled to determine the fragmentation mechanism in the presence of an electric field and dimer temperature as well as to provide fragmentation rates for the PIC simulations. An energy analysis of the molecular dynamics (MD) fragmentation demonstrated that the key mechanism for dimer fragmentation corresponds to a decrease in the Coulomb energy between the cation and anion in the system and that dimers of temperatures 300 and 600 K are extremely stable for electric fields smaller than 1.5 V/nm. Using probabilities of fragmentation consistent with the MD simulations, we implemented a dimer fragmentation model in our PIC simulations (shown in the LHS figure). The ion energy distribution functions obtained from the PIC simulations were used to predict retarding potential analysis (RPA) curves that were compared directly to measurements. The sensitivity of the RPA shape to the fragmentation probability was found to be significant. By comparing predicted and measured RPA curves for both negative and positive operation modes, and the fact that dimers do not fragment for electric fields less than 0.6 V/nm, we concluded that fragmentation of dimers occurs spontaneously due to their high thermal internal energies.

Plasma–surface interactions caused by electric propulsion devices are an important spacecraft aspect of the design that is difficult to measure in ground-based facilities. The negatively biased solar panel surfaces attract the slow-moving charge exchange (CEX) ions generated inside an ion core plume, which can cause surface sputtering on the protective coatings of the solar panels. We use a fully kinetic particle-in-cell direct simulation Monte Carlo (PIC-DSMC) approach that models both electron and ion trajectories to allow us to fully characterize the plasma sheath formed near these surfaces and to understand how the plasma sheath (shown in the LHS figure) affects the trajectories of CEX ions, their incident energies and angles, and surface sputtering rates. We find that, outside the plasma core, the ion and electron distribution functions are highly non-Maxwellian, and the assumption of electron temperatures is questionable. We introduce a novel floating potential ground boundary condition that enables us to emulate the spacecraft ground for a high range of plasma number densities and surface charging conditions. Finally, we estimate the erosion of the surface using the kinetic results and surface yield empirical relations.

Ion beam space-charge neutralization can be most economically achieved by an external emission filament placed near the ion beam. However, this and other Particle-in-Cell (PIC) studies have determined that a variety of waves, such as electrostatic solitary waves (ESWs), are excited in the process that limits the space charge neutralization. The generated ESWs (shown in the LHS figure) in this process are unusually wide, much wider than typical ESWs produced in a plasma with a Maxwellian electron velocity distribution function (EVDF) widely studied in many past theoretical studies. We show in this work, these wide modes appear because the densities of the trapped and un-trapped electrons in these ESWs nearly compensate each other due to the non-Maxwellian (EVDF) of the electrons produced in the neutralization process of an ion beam. In our (3D) PIC simulations, we also discovered that neutralizing electrons can excite Trivelpiece-Gould surface waves. The modes were observed in the three-dimensional but not in two-dimensional simulations. This points to the importance of realistic 3D simulations versus simplified 2D simulations.

An electron timescale pulse applied to a homogeneous plasma in a multipole plasma chamber (MPC) is modeled using a fully-kinetic particle-in-cell (PIC). From the time-accurate simulations, we observed that the ion-sheath expansion is accompanied with the Langmuir (LM) wave oscillations at the sheath edge, which decay after the application of the pulse. We validate our PIC results by comparing them with the previous analytical and experimental sheath expansion studies (LHS figure). Finally, we verified our PIC results by comparing them with the experiments conducted in an MPC where LM waves were excited when an electron timescale pulse is applied to a flat-conductor plate. In both PIC simulations and experiments, we found that the shape of the applied pulse dictated the wavelength of the sheath edge oscillations and the Landau damping had no discernable effect on them. The wavelength of the LM waves reduced with the applied pulse-width, where, in PIC simulations, these waves disappeared for a linear pulse of a longer time scale of tpulse = 1 μs (ωpetpulse = 178).