A MEMS electro-mechanical co-optimization platform featuring freeform geometry optimization based on a genetic...

Area of Science:

• Engineering
• Materials Science
• Computer Science

Background:

• Traditional microelectromechanical systems (MEMS) design often faces limitations in optimizing complex geometries and concurrently integrating electronic and mechanical components.
• Achieving high sensitivity, stability, and fabrication tolerance in MEMS devices remains a challenge, impacting overall system performance and reliability.

Purpose of the Study:

• To develop and demonstrate a novel system-level design methodology for MEMS devices using a genetic algorithm (GA) and freeform geometries.
• To concurrently design and co-optimize electronic and mechanical parts of MEMS devices for improved performance metrics.
• To experimentally validate the effectiveness of the proposed design approach on a MEMS accelerometer with a closed-loop control system.

Main Methods:

• A system-level design methodology employing a genetic algorithm (GA) was developed for microelectromechanical systems (MEMS).
• Freeform geometries were utilized to increase design freedom and explore a wider range of MEMS device configurations.
• A MEMS accelerometer with a freeform mechanical motion preamplifier in a closed-loop control system was designed and experimentally tested.

Main Results:

• The overall figure-of-merit (FOM) of the MEMS device was improved by 195% through the GA optimization.
• The product of sensitivity and bandwidth in the open-loop system improved by 151%, with a 276% increase in sensitivity.
• The closed-loop system demonstrated significant improvements, including an 86% decrease in displacement, 64% improvement in static nonlinearity, and 18.43% reduction in cross-axis sensitivity.

Conclusions:

• The proposed GA-based design methodology effectively co-optimizes electronic and mechanical components of MEMS devices with freeform geometries.
• The implemented closed-loop MEMS accelerometer shows substantial enhancements in performance metrics, including sensitivity, stability, and robustness to fabrication errors.
• This work represents the first experimental implementation of a closed-loop system for a MEMS accelerometer incorporating a mechanical motion preamplifier, validating the novel design approach.