Creation of scalable macromodels for micro-electromechanical systems

During the design of micro-electromechanical systems (MEMS) there is a significant demand for efficient and accurate simulation using macromodels where direct simulation is prohibitively time consuming. Recently, there have been. numerous publications for developing reduced-order macromodels for linear systems of equations. In this paper, these algorithms are applied to transient and steady state problems of large linear mass-stiffness finite element models of MEMS. The innovation reported in this paper is the ability to incorporate multiple design parameters such as the length, width and height of the MEMS structure. The algorithm is based on Krylov-subspace techniques and congruent transformation.