Nos tutelles

CNRS UPS Enseeiht

Rechercher




Accueil > Actualités

Khaled HALLAK’s Thesis Defense

par Laurence Laffont - publié le

Khaled HALLAK’s thesis defense, intitled "Optimisation des paramètres d’un modèle de transport de charge pour la caractérisation électrique des matériaux diélectriques" (Parameters’s Optimization of a Charge Transport Model for the Electrical Characterization of Dielectric Materials), will be taking place on Wednesday, April 13th 2022 at 10:30 am.

A visioconferencing link will be given ulteriorily.

Jury :

Mrs Virginie GRISERI - Thesis director - Senior Lecturer - Université Toulouse III - Paul Sabatier - LAPLACE
Mr Samir ADLY - Rapporteur- Professor - Laboratory XLIM, Université de Limoges
Mr Serge AGNEL - Rapporteur - Professor - Institut d’Électronique et des Systemes (IES) UMR 5214, Université de Montpellier / CNRS
Mr Fulbert BAUDOIN - Thesis co-director - Senior lecturer - Université Toulouse III - Paul Sabatier - LAPLACE
Mr Florian BUGARIN - Thesis co-superviszor - Senior Lecturer - Université Toulouse III - Paul Sabatier
Mr Stéphane HOLÉ - Reviewer - Professor - Sorbonne Université / Faculté des Sciences

Abstract

A vast majority of the current physical concepts used to describe charge transport and breakdown in solid dielectrics has been studied for more than 30 years. A mathematical model based on the physics of insulating materials has been developed in our laboratory to describe the bipolar charge transport (BCT) in low-density polyethylene (LDPE) under DC stress. The phenomena of trapping and detrapping, the barrier height for injection, the mobility, and the recombination process of positive and negative charges are considered. The charges present are generated by a modified Schottky injection at the metal-dielectric interface and are extracted without a barrier. The model is based on the Poisson equation and the law of conservation of charges to which source terms are added to take into account the microscopic phenomena occurring in this type of material. This model requires inputs that are related to the experimental conditions such as temperature, applied voltage, dielectric thickness, etc., as well as a set of parameters such as the injection barrier, mobility, trapping, and detrapping coefficients. Most of these parameters cannot be predicted, observed, or estimated by independent experiments. In other words, it is difficult to estimate these parameters in such a way as to have a model capable of predicting the experimentally observed behavior of the charges. For this reason, an optimization algorithm is used to optimize the BCT model to fit the experimental measurements, whatever the experimental conditions, by minimizing the sum of squares of the deviations between the experimental data and the model data. The experimental data used are the net charge density measured by the pulsed electro-acoustic method (PEA) as well as the external charge current measurements. All experiments and sample preparation were performed in the LAPLACE laboratory. After testing five optimization algorithms we selected the following algorithm Trust Region Reflective which best meets our criteria. This algorithm has allowed us to find a set of parameters allowing a good correlation between the simulated current and charge densities and those obtained experimentally. This optimization was performed by considering different electric fields applied to the material in order to have a unique set of parameters that best characterizes the study material. In addition, the optimization algorithm allowed to analyze the charge transport model used as the Schottky law and its injection barrier when the interfaces are of different natures or to show the weaknesses of the model when considering the phenomena of trapping and detrapping.