Soutenance de thèse Raphaël Guffond

This work aims at developing a new model able to simulate the modifications over time of the micro-structure in insulation polymers submitted high to electric field and temperature as well as the related impacts on electrical properties and durability. The two main objectives of this study are i) to model electrical properties of polymer by taking into account all the influencing heterogeneities (physical and chemical) and ii) to highlight the impact of this heterogeneous micro-structure modification over time on these electrical properties. To do so, a new model approach is used based on Markov model and genetic evolution. Markov model is a stochastic model where system is composed of a states distribution changing with time. At a given time, the distribution evolves (mutates) into a new state distribution based only on the distribution of previous state. The main advantage of this model compared to others is that its simulation is based on local interaction calculation and not on global criteria.