Probabilistic modeling of microelectromechanical systems (MEMS)

Micro-Electro-Mechanical Systems (MEMS) are a fast-developing technology that have a potential to permeate most engineering and medical applications. For this technology to continue expanding, issues regarding the cost of manufacturing and reliability of the devices have to be addressed. To improve the reliability, probabilistic design methodologies are potent in both the modeling and testing of high-performance MEMS. The benefit of probabilistic design approaches is a more rational basis for making design decisions that balance component or system efficiency with reliability or safety. Probabilistic methods are used to assess uncertainties involved in the manufacturing of MEMS devices. Probabilistic methods guide the design of these devices to achieve reliable design in a most efficient way.

The objectives of the research work were to formulate and analyze probabilistic failure criteria on a simplified capacitive accelerometer mode. In this respect, comprehensive probabilistic and deterministic analysis was carried out for the selected model.

The scope of work is threefold. First, two probabilistic failure criteria will be investigated on the capacitive structure, namely probabilistic clearance failure criterion and probabilistic fracture toughness failure criterion. Second, four kinds of probabilistic analyses for characterization of MEMS will be used: probability of failure, sensitivity analysis, safety index, and probability-based design. Third, three kinds of finite element analyses, namely static, modal and spectral analysis, will be used to see the deterministic response and will be compared with probabilistic result