Modeling Anomalous Electron Transport in a Fluid Hall Thruster Code

Abstract

Hall thrusters are the most widely-used type of in-space electric propulsion device. However, aspects of the physics of their operation remain poorly-understood. Notably, the problem of enhanced cross-field electron transport prevents predictive modeling and simulation of these devices. This "anomalous" electron transport, which is often represented as an effective collisional scattering process likely stems from kinetic instabilities and plasma turbulence. This poses a challenge for incorporating this phenomenon into fluid models suitable for engineering applications.

In this dissertation, several models for the anomalous electron transport in Hall thrusters are reviewed and evaluated, and methods for model validation are assessed. Electron transport models from the scientific literature are reviewed, and several new models are derived. It is first shown that the common practice of evaluating models on calibrated simulation outputs and comparing them to hand-tuned, ad-hoc anomalous collision frequencies does not yield predictive results.

Next, the behavior of four electron transport models from the literature is investigated using a fluid Hall thruster code. The models are calibrated against a baseline experimental condition of a 9-kW-class magnetically-shielded Hall thruster operating at 300 V and 15 A on xenon propellant. The extensibility of the models is then assessed by using this calibrated model to simulate three additional operating conditions---300 V and 30 A, 600 V and 15 A, and 300 V and 15 A operating on krypton propellant. The quality of the model prediction is quantified by comparing the model outputs to experimental measurements of discharge current, thrust, and ion velocity. It is found that while none of the models can predict the ion acceleration characteristics accurately, some compare favorably in terms of the scaling of thrust and discharge current across operating conditions. The limitations of the models are attributed to the coupling between the functional scaling of the closure models with respect to the local plasma properties and the fluid model. The role of the electron energy balance in this coupling is also highlighted.

Finally, several novel models are assessed. These models are derived by altering the underlying assumptions of a previously-discussed first-principles transport model. It is found that some of these models are able to predict thruster performance characteristics---such as thrust and efficiency---better than the baseline model, while inaccurately reproducing the experimentally-observed spatial variation in ion velocity. Another model is then evaluated which improves upon these results, yielding improved accuracy with respect to these velocity measurements in all but one case. The results of this dissertation are finally discussed in the context of motivating improved closure models of the anomalous electron transport in Hall thrusters.