Nos tutelles

CNRS UPS Enseeiht


Accueil > Actualités

Soutenance de thèse de Sarah SADOUNI

par Isabelle Clarysse - publié le

Sarah SADOUNI soutiendra publiquement ses travaux de thèse intitulés :

"Fluid modeling of transport and instabilities in magnetized low-temperature plasma sources".

Les travaux de thèse ont été réalisés sous la direction de Gerjan HAGELAAR et Andrei SMOLYAKOV.

La soutenance se déroulera jeudi 27 février 2020 à 14h30, salle des colloques au bât. 3R2 à l’Université Paul Sabatier, 118 route de Narbonne.

Jury :
M. Gerjan HAGELAAR - LAPLACE (UPS), CNRS - Directeur de thèse
M. Andrei SMOLYAKOV - Université de Saskatchewan - Co-directeur de thèse
M. Pierre FRETON - LAPLACE (UPS) - Examinateur
M. Jan VAN DIJK - Eindhoven University of Technology - Examinateur
Mme Anne BOURDON - Laboratoire de Physique des Plasmas (LPP), CNRS - Rapporteur
Mme Sedina TSIKATA - Institut de Combustion Aérothermique Réactivité et Environnement (ICARE), CNRS - Rapporteur
M. Jean-Pierre BOEUF - LAPLACE (UPS), CNRS - Invité

Mots-clés :
fluid modeling, plasma, instabilities, ExB drift, anomalous transport, Hall thruster

Abstract :
It is well known from experiments that magnetized low-temperature plasmas in devices such as Hall thrusters and ion sources often show the emergence of instabilities that can cause anomalous transport phenomena and strongly affect the device operation. In this thesis we investigate the possibilities to simulate these instabilities self-consistently by fluid modeling. This is of great potential interest for engineering. We used a quasineutral fluid code developed at the LAPLACE laboratory, called MAGNIS (MAGnetized Ion Source), solving a set of fluid equations for electrons and ions in a 2D domain perpendicular to the magnetic field lines. It was found that in many cases of practical interest, MAGNIS simulations show plasma instabilities and fluctuations.
A first goal of this thesis is to understand the origin of the instabilities observed in MAGNIS and make sure that they are a physical result and not numerical artifacts. For this purpose, we carried out a detailed linear stability analysis based on dispersion relations, from which analytical growth rates and frequencies were successfully compared with those measured in MAGNIS simulations for simple configurations forced to remain in a linear regime. We then identified these linear unstable modes and their responsible mechanisms (involving parameters such as the density gradient, electric and magnetic fields and inertia), known from the literature, that are likely to occur in these fluid simulations.
Subsequently, we simulated the nonlinear evolution and saturation of the instabilities and quantified the anomalous transport generated in different cases relevant to ion sources, depending on various key parameters of the system (electric and magnetic fields and electron temperature).
Finally, we highlighted several limitations of MAGNIS, and more generally of fluid models, due to the physical approximations made (quasineutrality, absence of kinetic effects). We showed that the fluid modes are sometimes most unstable at infinitely small scales for which the theory is no longer valid and which cannot be resolved numerically. We proposed, and tested in MAGNIS, ways to overcome this problem by introducing effective diffusion terms representing small scale processes (non-neutrality, Larmor radius).