Bayesian Inference of the Anomalous Electron Transport in a Multi-fluid Hall Thruster Model

The lack of a first-principles understanding of the anomalous electron transport in Hall thrusters has precluded the development of fully-predictive engineering models of thruster operation. The current standard method for fluid-based simulations of Hall thrusters is to represent the electron transport as an anomalous collision frequency with a static, spatially-varying profile along channel centerline. The shape of this profile is adjusted until key quantities of interest—the ion velocity flow field and performance—match experimental measurement. The shapes of these profiles are typically hand-tuned in a process informed by user-experience and intuition. In this work, we develop an algorithm based on Bayesian Inference to rigorously determine the shape of the anomalous transport profile. We also develop a method to quantify the impacts of experimental and model-based uncertainty on our confidence in the median values of this profile. This approach is demonstrated on a multi-fluid Hall thruster code with an experimental dataset from the H9, 9-kW class magnetically-shielded Hall thruster.