Expansion of a collisionless hypersonic plasma plume into a vacuum

Ion wave formation during ultracold plasma expansion

We present the results of a direct simulation of the expansion of a two-component ultracold plasma for various numbers of particles, densities, and electron temperatures. A description of the expansion process common to all plasma parameters is given. After the escape of fast electrons from the plasma cloud, the excess positive charge is localized at the outer boundary, in a narrow layer. This layer has a characteristic front shape with a sharp drop in the charge concentration. The charged layer retains the remaining electrons during the entire expansion process. As the plasma expands, the speed of movement of the charged layer becomes constant and significantly exceeds the sonic speed of ions. In addition, the dependence of the radial velocity of ions on the radius acquires a self-similar character long before the final stage of expansion. Based on the calculation results, equations and self-similar solutions are obtained. General dependencies on plasma parameters are determined, which are compared with experimental data.

Application of the view factor model on the particle-in-cell and Monte Carlo collision code

Particle-in-cell and Monte Carlo collision (PIC-MCC) has been widely adopted as a simulation method for electric propulsion. However, neutral atoms move much more slowly than other species, which can cause a serious reduction in simulation speed. In this work, we investigate the view factor model in combination with the PIC-MCC method and propose a method for simulating three-dimensional neutral atoms. The accuracy of the PIC-MCC method can be significantly improved by updating the neutral distribution periodically. We compare the computational results with the fixed-neutral PIC-MCC model of the miniature ring-cusp discharge experiment at the University of California, Los Angeles (UCLA). The plasma distribution and potential distribution of the simulation match well with the UCLA experimental data. Compared with the fixed-neutral model, the view factor model increases the simulation time by only 33% while it improves the distribution accuracy of neutrals, plasma density, and electric potential, and reduces the simulation errors of discharge current and discharge power from 19.8% to 9.8%. The accuracy of PIC-MCC simulation has been improved at the expense of slightly increasing the computational time.

Discrete unified gas kinetic scheme for electrostatic plasma and its comparison with the particle-in-cell method

In this paper, we present a finite-volume direct kinetic method, the so-called discrete unified gas kinetic scheme (DUGKS), for electrostatic plasma. One key feature of this method is the semi-implicit unsplitting treatment of particle transport and collision, and thus the time step in current DUGKS is not limited by the particle collision time. In addition, a fourth-order compact MUSCL scheme with a positivity preserving limiter is implemented in the interface reconstruction, which enables present DUGKS to preserve the favorable conservative property and positivity of distribution function. Combined with this MUSCL method, the semi-Lagrangian scheme is used for the particle transport in velocity space to remove Courant-Friedricks-Lewy restriction induced by the large electric force. As a result, the proposed DUGKS becomes an efficient and stable multiscale scheme. Several numerical experiments, including plasma sheath, linear Landau damping, collisional nonlinear Landau damping, and plasma ion acceleration, are performed to validate current DUGKS. A comparative study of current DUGKS with a general particle in cell (PIC) method which could handle particle collision in a conservative way is also presented. Theory and numerical experiments demonstrate that DUGKS is preferred for plasma flows involving small electrostatic perturbation and high collision regimes, while the PIC method is desired for the field- dominated plasma flows involving a wide range of velocities.