Beam parameters optimization of MEMS piezoresistive accelerometer by using response surface method

Abstract

This article presents the optimization of a MEMS-based piezoresistive accelerometer sensor using design of experiment (DOE) approach. Two structures of accelerometers, which consist of a proof mass suspended by eight beams, have been investigated. Response surface method (RSM) was employed to optimize the geometric beam parameters (thickness, width and length) in order to obtain high sensitivity with an appropriate resonant frequency that satisfies the design requirements for airbag application. The beam thickness, beam width, and beam length were optimized to achieve 0-400Hz bandwidth with high sensitivity. By implementing RSM, the best combination of the three parameters for structure 1 was thickness, = 23 μm, width, = 210 μm, and length, = 800 μm. On the other hand, the best combination of the three parameters for structure 2 was thickness, = 20 μm, width, = 300 μm, and length, = 800 μm. As a result, by using optimized beam parameters, the mechanical sensitivity of the MEMS accelerometer sensor was increased by almost 34.7% and 22.1% for structure 1 and structure 2, respectively. The optimization results showed that the predicted beam parameters of MEMS accelerometer managed to achieve the target specifications for airbag application. Hence, the approach can be successfully applied in improving performance of MEMS-based devices.