Shape optimization of microlever for mode-localized accelerometer

A microlever is an effective approach to increase the sensitivity of MEMS sensors. However, traditional optimization methods for microlevers have limitations in improving the amplification factor. In this paper, a shape optimization method based on the BOBYQA algorithm is utilized in microlever optimization for the first time to improve the amplification factor. Compared with the traditional rectangular microlever, the optimized dragonfly-shaped microlever achieves superior performance under the same dimensions. Mode-localized accelerometers integrating the optimized dragonfly-shaped microlever are fabricated to verify the proposed optimization method. The experimental results show that following shape optimization, the accelerometer achieves a sensitivity of 630/g, which is 264 % higher than before optimization.