AUGMENTED DIAGNOSTICS FOR ELECTRIC PROPULSION

Abstract

Hall effect thrusters (HETs) have become the leading electric propulsion technology for satellite constellations and deep-space missions, yet predicting their in-space performance from ground test data remains a central challenge. Extended Kalman filters and other data-driven modeling frameworks offer a possible path toward bridging the ground-to-orbit gap by fusing experimental measurements with physics-based models to estimate poorly understood model parameters, such as anomalous electron transport. However, these frameworks are fundamentally limited by the information content of the measurements supplied to them. This dissertation addresses the challenge of improving the specificity of traditional diagnostic measurements used to inform Hall thruster models.Three complementary techniques are developed, each targeting a distinct dimension of measurement specificity. The first is ion-species specificity, achieved with a combined energy-velocity analyzer (EVADER) probe that sequentially filters ions by energy per charge and velocity. By collecting ExB spectra at multiple pass energies set by an upstream electrostatic energy analyzer, the EVADER deconvolves overlapping ion energy distributions that can be inseparable with conventional diagnostics. This can be used to produce species-specific ion energy distribution functions for singly, doubly, and triply charged ions across multiple thruster operating conditions. The second contribution addresses temporal specificity through the application of shadow manifold interpolation (SMI), a nonlinear state-space reconstruction technique based on Takens' embedding theorem. The SMI technique is used to reconstruct time-resolved ion current density profiles from slow-sweep Faraday probe measurements, using synchronous discharge current waveforms as the reference signal, producing spatiotemporal current density distributions that capture breathing-mode oscillation dynamics inaccessible to conventional time-averaged measurements. The third contribution targets causal specificity through extended convergent cross mapping (eCCM), a causality determination technique applied to HET telemetry signals. The eCCM analysis evaluates causal relationships between cathode current, discharge voltage, and cathode-to-ground voltage across multiple operating conditions and background pressures. The results indicated that discharge voltage retains the most dynamically relevant information and that facility pressure degrades causal relationships among telemetry signals. Together, these improvements in species, temporal, and causal specificity form a measurement framework that can enhance the data supplied to next-generation data-driven thruster models, supporting both faster qualification of current flight hardware and the development of more capable Hall effect thrusters.