Thèse soutenue

Modelisation et optimisation des actionneurs électrostatiques basés sur la membrane
FR  |  
EN
Accès à la thèse
Auteur / Autrice : Cezary Maj
Direction : Patrick PonsAndrzej Napieralski
Type : Thèse de doctorat
Discipline(s) : Conception des circuits micrélectroniques et microsystèmes
Date : Soutenance en 2009
Etablissement(s) : Toulouse, INSA en cotutelle avec Technical university of Lodz (Pologne)
Partenaire(s) de recherche : Laboratoire : Laboratoire d'Analyse et d'Architecture des Systèmes (Toulouse ; 1968-....)

Résumé

FR  |  
EN  |  
PL

This dissertation is concerned with an analytical modelling and optimization of electrostatic membrane-based actuators. An analytical model of a thin, perfectly clamped, silicon membrane bending described with Lagrange/Newton equation for small deflections is presented taking into account parasitic phenomena like residual stress and initial deflection of a membrane. Responses for uniform pressure are obtained with Galerkin method and are studied in order to describe the membrane’s behaviour with a reduced model. Both models are analyzed in the case of electrostatic actuation and are compared in terms of precision, range of use, calculation time and easiness of use in the optimization process. The reduced model is then corrected to be fully applicable for hydrostatic and electrostatic actuation without significant precision loss and is compared with FEM simulations performed with ANSYS®. Obtained results allow defining the model advantages and disadvantages. Next, the analytical model is used to determine the influence of input parameters on actuator behaviour and to perform an optimization phase for various applications. Furthermore, statistical simulations are presented which are useful in the estimation of the influence of parameters dispersion, resulting for example from the technological process, on the device behaviour. All simulations are performed using dedicated tools written in MATLAB®. Finally, the fabrication and the characterization of test structures performed in LAAS-CNRS are described which allows the validation of the developed analytical model