Modeling and simulation of the plasma discharge in a radiofrequency thruster
- Autores: Jiewei Zhou
- Directores de la Tesis: Pablo Fajardo Peña (dir. tes.), Eduardo Ahedo Galilea (dir. tes.)
- Lectura: En la Universidad Carlos III de Madrid ( España ) en 2021
- Idioma: español
- Tribunal Calificador de la Tesis: Ricardo Albertini (presid.), José Miguel Reynolds Barredo (secret.), Justin Little (voc.)
- Programa de doctorado: Programa de Doctorado en Mecánica de Fluidos por la Universidad Carlos III de Madrid; la Universidad de Jaén; la Universidad de Zaragoza; la Universidad Nacional de Educación a Distancia; la Universidad Politécnica de Madrid y la Universidad Rovira i Virgili
- Materias:
- Enlaces
- Tesis en acceso abierto en: e-Archivo
- Resumen
In the current electric propulsion industry for space applications, two of the main issues are: (i) the lifetime limitation of the mature technologies, Hall effect thrusters and gridded ion thrusters, due to the erosion of their electrodes; (ii) the search for alternative propellants due to the scarcity of xenon. Electrodeless thrusters with magnetic nozzles, in particular the helicon plasma thruster and the electron cyclotron resonance thruster, are disruptive electric propulsion concepts that offer prolonged lifetime and tolerance for a wide variety of propellants. These thrusters are potential substitutes for the mature ones but are still under development, and further research is necessary to understand the physics behind their operation and make them competitive in terms of propulsive performance.
This thesis is focused on the modeling and simulation of the plasma discharge in electrodeless thrusters with two codes. HYPHEN, a two-dimensional axisymmetric hybrid code, is used for full simulations. This code is extended from Hall effect thrusters to electrodeless thrusters, within the objective of developing a multi-thruster simulation platform valid for many types of electromagnetic thrusters. VLASMAN, a paraxial (one-dimensional) kinetic code, is used for simulations of the plasma expansion along magnetic nozzles.
The hybrid formulation of HYPHEN offers a good trade-off between computational cost and reliability of the results for full simulations. A particle-in-cell model complemented with Monte Carlo collisions methods is used for heavy species, and a fluid model is used for electrons. From previous works, the particle model was ready for use, while the fluid model, with the basis established, was incomplete from the numerical point of view. The strongly magnetized electron fluid requires the use of a magnetic field aligned mesh given the anisotropic character of the transport properties. However, the mesh, for realistic magnetic field topologies, can be highly irregular and the numerical algorithms were leading to inaccurate results. Thus, in this thesis, the numerical treatment of the fluid model is investigated, and robust numerical algorithms are found allowing to solve even complex magnetic topologies with singular points. Once the electron fluid model is complete, simulations coupled with the particle model are run for the helicon plasma thruster prototype HPT05M. The simulations are focused on the plasma transport assuming a known power deposition map from the helicon antenna. The thruster performances and profiles of plasma magnitudes are studied. HPT05M, such as many other prototypes, reports a very low thrust efficiency (about 1.3%), and simulations have helped to identify the mechanisms of inefficiencies and propose improvements. In this thruster, the applied magnetic field generated by a coil decays fast along its slender vessel, and the magnetic shielding of the walls is found poor with a huge amount of plasma recombined there. Alternative configurations of the thrusters, placing more coils along the vessel or shortening it, are tested. The magnetic shielding is improved and the plasma recombination is reduced. Thrust efficiencies of 9.3-10.4% are achieved with the alternative configurations. However, this optimization is partial, since the efficiencies are still within the state-of-art. The main limitations for a full optimization beyond the state-of-art are identified and possible solutions are proposed.
Furthermore, HYPHEN was initially developed to simulate xenon and other atomic propellants. In this thesis, as many candidates for alternative propellants usually have more complex chemistry, the code is implemented with the main collisions for diatomic substances. Simulations are run with air substances as propellant for HPT05M testing successfully the implementation. The results also allowed to evaluate the air-breathing concept, with potential application in low orbits for drag compensation, in helicon plasma thrusters. The air is found much worse than xenon, from a propulsive point of view, if operating at low powers (about 0.1kW), while it can be competitive if operating at high powers (about 1kW).
The kinetic formulation of VLASMAN is used for deeper studies of the plasma expansion along magnetic nozzles. In the expansion, the plasma becomes very rarefied, and more accurate simulations than those from HYPHEN are required. One-dimensional steady state models were used in previous works, however they were not able to solve self-consistently a subpopulation of electrons trapped along the expansion. VLASMAN implements a time-dependent, weakly-collisional, Boltzmann-Poisson system and models electron-electron intraspecies collisions with a Bhatnagar-Gross-Krook operator. Thus, the code can characterize the mechanisms responsible for the trapping of electrons, the transient and collisional processes. Simulations with VLASMAN are run to study the trapped electrons in terms of the transient history and collisionality. The solution of the subpopulation, and that of the whole plasma, reached in the steady state is found dependent on the transient history. Once the collisions are added, even if rare, the transient history is erased and the steady state solution becomes unique. The amount of trapped electrons is found important on the electron cooling and on the balances of electron momentum and energy. Furthermore, some studies focused on the extraction of results for implementation in macroscopic models are conducted and, in particular, closures for the electron heat flux are investigated.
