Inverse design of multistable kirigami metamaterial via geometry-enabled shape programming and transforming

The inverse design of metamaterials with desired properties represents a significant challenge in mechanical science. Despite the potential demonstrated by recent algorithm models, their adoption has been limited by constraints such as the geometric limitations of elementary building cells. The use of the kirigami principle, which offers large deformation and nonlinear stiffness, has been explored. However, existing kirigami geometries, which remain isotropic, may restrict the design space. Our objective is to leverage the capabilities of geometry in shape programming and transformation to provide a framework for inverse design. This framework utilizes a unified geometry in kirigami cutting that is easily parameterized to generate independent anisotropic deformation and bistability. By integrating machine learning with a genetic algorithm, we achieve an inverse design process. The resulting kirigami architectures can be preprogrammed into target shapes and transformed between multiple stable states. This work underscores the significance of cell geometry topology, offering a powerful tool for the inverse design of metamaterials with reconfigurable and tailored mechanical properties, applicable in various fields such as robotics, electronics, and beyond.