Physics-based and data-driven models of partially ionized gases

Ionized gases (plasmas) play an important role in various engineering applications, including spacecraft electric propulsion, hypersonic flow, space weather, and microelectronic fabrication. The goal for computational plasma models is to predict dynamic behavior of complex systems, e.g., transport, reactions, energy heating and loss mechanisms, plasma-material interaction, etc. In this talk, I will present an overview of the theoretical and computational models for weakly ionized plasmas, with a particular focus on partially magnetized plasmas that are used in cross-field discharges such as Hall effect thrusters. First, I will discuss various kinetic models, such as particle-in-cell (PIC), Monte Carlo collision (MCC), and grid-based direct kinetic (DK) models, to investigate the effects of kinetic instabilities on anomalous plasma transport. Next, I will introduce the development of the full fluid moment (FFM) model, which is a new type of hydrodynamic model for low-temperature plasmas that capture electron inertia effects. Finally, I will present a data assimilation technique using extended Kalman filter (EKF) for plasma dynamics, which allows one to infer hidden states and parameters that are difficult to be measured.