Deep learning-driven ultra-stretchable kirigami metamaterials: towards surface texture modulation via buckling.

Area of Science:

• Mechanics of Materials
• Computational Materials Science
• Metamaterials Design

Background:

• Kirigami structures exhibit unique buckling instability under tensile stress, offering stretchability and design versatility.
• Conventional kirigami design often focuses on unidirectional cuts and symmetric geometric configurations.
• Exploring symmetry-disrupted kirigami is crucial for understanding buckling behavior and enhancing programmability.

Purpose of the Study:

• To analyze buckling instability mechanisms in tessellated kirigami structures with disrupted geometric symmetry.
• To develop an innovative, programmable kirigami design strategy using deep learning.
• To predict mechanical performance and optimize kirigami patterns for specific tensile strain requirements.

Main Methods:

• Analysis of buckling instability in tessellated kirigami structures.
• Application of deep learning techniques to predict nonlinear constitutive relationships.
• Development of a programmable design framework for kirigami pattern identification.

Main Results:

• Achieved 94.29% accuracy in predicting kirigami mechanical performance.
• Demonstrated that breaking geometric symmetry significantly expands the kirigami design space.
• Showcased the potential for information encoding and transmission using kirigami metasurfaces.

Conclusions:

• The proposed deep learning-based strategy enables accurate prediction of kirigami mechanical performance.
• Symmetry-breaking in kirigami design enhances adaptability and minimizes trial-and-error.
• Kirigami metasurfaces offer a novel platform for functional configuration and information transmission.