L'impact de la modélisation des circuits et des particules à faible énergie sur la prédiction du taux d'événements singuliers.

As microelectronic technology continues to scale down, electronic devices have become increasingly sensitive to radiation effects, particularly Single Event Effects (SEE). The study of SEE characterization relies heavily on the use of radiation testing. However, experimental activities are complex, time-consuming, and costly. In this context, modeling and prediction tools offer a complementary alternative to predict behavior before testing and to understand post-irradiation data. However, obtaining realistic parameters for modeling the simulations can be challenging, as they are typically confidential or not openly available from manufacturers. Furthermore, depending on the method used, the modeling may require a high computational effort, which impacts in very long simulations. Given the importance of SEE modeling and the obstacles associated with traditional methods, it is crucial to explore alternative modeling approaches that can overcome these limitations. This thesis investigates the impact of circuit modeling and low-energy particles on the prediction of SEE, focusing on alternative and simplified approaches. The study encompasses different methodologies, including analytical modeling, Monte Carlo simulations, electrical models, and Technology Computer-Aided Design (TCAD), to explore the advantages and limitations of these alternative modeling approaches, while providing significant insights into critical concerns related to radiation effects on electronic components.