SIMULATION AND MODELLING OF MEMS CANTILEVER BEAM FOR LOW PULL-IN VOLTAGES

Abstract

This thesis presents in-depth simulation and modelling study on electrostatically driven MEMS cantilever beam switches with a main goal of obtaining low pull-in voltage for efficient and dependable operation in low-power electronic systems. MEMS switches are increasingly important in modern technology because of their possible miniaturization, high speed switching, low power consumption, and integration into complex microsystems. This work aims to meet this challenge by investigating, using COMSOL Multiphysics 6.2, several structural and material parameters on the performance of a cantilever-based MEMS switch through finite element simulation. Ten independent configurations in all were modelled, varying key parameters including the cantilever beam material (gold and polysilicon), the electrode material (aluminium, copper, and gold), the beam dimensions (small and large), and the presence or absence of a dielectric layer between the beam and the electrode. Each configuration was painstakingly modelled with physics-driven meshing (set to a fine level), and the simulations included electromechanical coupling to faithfully record the interaction between the electrostatic force generated by the applied voltage and the resulting mechanical deformation of the cantilever beam. The main focus of the study was extensive performance measurements covering pull-in voltage, maximum displacement, and von Mises stress distribution across the structure. The results revealed clearly different performance depending on configurations. Particularly, a design combining a gold cantilever beam of larger dimensions with an aluminium electrode without dielectric layer showed the lowest pull-in voltage yet maintaining structural integrity. Rising as the most likely candidate for practical use, it struck a good mix between mechanical flexibility and electrostatic efficiency. To match normal voltage levels from 1.V to 10.V, the original voltage sweep—from 0.V to 9.15’V in fine steps was scaled and interpolated, so improving the interpretability and usability of the simulation results. This conversion enabled simpler integration into system-level circuit simulations or control strategies, helped to better analyse data, and present information.