Inverse Design of Highly Deformable Mechanical Metamaterial Based on Partitional Semi-Random Optimization

Area of Science:

• Materials Science
• Mechanical Engineering
• Robotics

Background:

• Flexible mechanical metamaterials utilize reversible large deformations for force, motion, and energy transfer.
• Existing topology optimization methods are limited to small deformations, hindering customization of large deformation behaviors.

Purpose of the Study:

• To develop a novel optimization method for customizing large deformations in mechanical metamaterials.
• To enable personalized design of multi-objective deformation patterns and paths for soft robotics and shape-morphing structures.

Main Methods:

• Proposed a partitional semi-random optimization concept that records and evaluates structural evolution in subregions.
• Implemented a statistical decision-making procedure to avoid repetitive iterations common in heuristic optimizations.
• Developed a two-step optimization scheme for local stiffness regulation to expand design possibilities with a single material.

Main Results:

• Successfully demonstrated personalized customization of deformation patterns and paths in highly deformable mechanical metamaterials.
• Verified the efficiency of the proposed design process through experiments with 3D-printed metamaterial samples.
• Expanded the design space for mechanical metamaterials using homogeneous rubbers.

Conclusions:

• The novel optimization method provides a solution for the large deformation design of mechanical metamaterials.
• This research facilitates advancements in shape-morphing structures and soft robotics.
• The partitional semi-random optimization approach offers a more efficient design process compared to traditional methods.