CMU MEMS Laboratory Publication Abstract

 

in Ph.D. Thesis, May 2003, Carnegie Mellon University, Pittsburgh, PA.
Modeling and Simulation for Design of Suspended MEMS
Q. Jing
ABSTRACT:
This thesis presents a modeling and simulation methodology for the design of suspended MicroElectroMechanical Systems (MEMS), called "NODAS" (NOdal Design of Actuators and Sensors).NODAS simulations are based on schematics composed of a small number of low-level "atomic" elements: anchors, beams, plates and electrostatic gaps. The lumped parameterized behavioral models are implemented in Verilog-A, which is an analog hardware description language. Key issues addressed include schematic representation, modeling physics, modeling accuracy, validation of the composibility and extensibility of the methodology.

Prior work on the NODAS model library is improved by adding in new features and new models. Model physics are expanded from 2D (in-plane) motion to 3D (in-space) motion. Beam and plate models include parameters for both single-conducting-layer processes and multipleconducting- layer processes. Mechanical physics are enhanced to include beam geometric nonlinearity and shear effects as well as plate elasticity. The electrostatic gap element models allow distributed electrostatic and damping forces acting on electrodes formed from displaced beams. Detailed model derivations are given, followed by verification simulations and discussion on advantages and limitations. The atomic element models have been verified to give simulation accuracy to within 5% of finite element analysis.

This thesis also validates the composibility of suspended MEMS by simulating a set of validation cases using the same small set of atomic elements. The validation cases covers a variety of suspended MEMS devices. Strengths and weaknesses of the current model library are discussed, suggesting directions for future work.

The good simulation accuracy and speed of structured modeling with the library supports iterative design and evaluation. The methodology supports multi-physics analysis and cosimulation with transistor-level electronics, handles hierarchical design for large systems, and therefore acts as a foundation for future system-level MEMS design.
© 2003 Carnegie Mellon University, Department of Electrical and Computer Engineering.
Full paper (PDF) (opens in new window).


This page was generated in 0.015657 seconds at 09:11:29 am UTC on 29 Mar 2024.

overview | projects | people | publications | intranet | resources         © 1998-2009  Carnegie Mellon