Determining causality in Hall effect thrusters using extended convergent cross mapping: II*

Abstract

Extended convergent cross mapping was previously shown to be a powerful tool for analyzing the dynamics of Hall-effect thrusters and has applications in many diverse fields. Due to its limitations, data splitting and self-reconstruction are introduced for robustness, and the results are judiciously used to eliminate inconclusive and false links (sample scenarios are discussed). The analysis is then applied to Hall-effect thruster data, and the results are used to construct causality network graphs that map the flow of information inside the thruster. Two operating modes are studied: quiescent and oscillatory. For both cases, the causality links qualitatively agree with calculated plumes and demonstrate that charged particles carry information about the currents throughout the experiment, as expected. Causal links carried by neutral particles or propagating through waves are possible, but outside the scope of this work. Additionally, the extended convergent cross mapping results can be used to estimate the first two harmonic frequencies in the oscillatory mode to great precision

Document Details

Document Type
Pub Defense Publication
Publication Date
Mar 01, 2022
Source ID
10.1088/1361-6595/ac511e

Entities

People

  • Cesar E. Huerta
  • Daniel Eckhardt
  • J W Koo
  • Robert Martin

Organizations

  • Air Force Office of Scientific Research

Tags

Readers

  • Aerospace Propulsion Engineering.
  • Neural Network Machine Learning.
  • Plasma Physics / Magnetohydrodynamics

Technology Areas

  • Space
  • Space - Hall-Effect Thruster