Two-Phase Approach for Fast Topology Optimization of Multi-Resonant MEMS Involving Model Order Reduction.

Area of Science:

• Engineering
• Materials Science
• Computational Mechanics

Background:

• Topology optimization is crucial for designing micro-electro-mechanical systems (MEMS).
• Classical methods like density-based optimization can be computationally intensive.
• Bi-directional evolutionary structural optimization (BESO) offers faster convergence but faces challenges.

Purpose of the Study:

• To develop a fast topology optimization approach for multi-resonant MEMS.
• To minimize computational effort in achieving optimal MEMS designs.
• To address convergence issues encountered in BESO.

Main Methods:

• A two-phase optimization strategy combining BESO and density-based methods.
• Implementation of model order reduction (MOR) to decrease computational time.
• Benchmarking on linear gyroscope and micromirror designs.

Main Results:

• The two-phase approach achieved optimal MEMS designs within 200 iterations.
• BESO was used for initial optimization, followed by a density-based phase to ensure convergence.
• Model order reduction reduced goal function computation by 50% in tested examples.

Conclusions:

• The proposed two-phase method significantly accelerates MEMS topology optimization.
• Combining BESO with a density-based phase effectively resolves convergence issues.
• Model order reduction further enhances computational efficiency for MEMS design.