This research presents a novel optimization method for the inverse design of flexible mechanical metamaterials, enhancing traditional strategies by evaluating deformation in subregions. Combining simulated annealing with a genetic algorithm, it improves design efficiency and enables customizable deformation paths. Experimental validation confirms its effectiveness, with potential applications in biomedical devices and adaptive structures.
Abstract
The flexible mechanical metamaterial leverages the reversible large deformation of its lattice unit cells to transfer force, motion, and energy. This characteristic imparts significant mechanical advantages for both morphological and functional regulation. However, most existing topology optimization methods primarily focus on initial stiffness and degrees of freedom under small deformations, which are incapable of the customization of large deformations in mechanical metamaterials. In this research, the concept of partitional semi-random optimization is proposed, which respectively records the structural evolution history in each subregion and evaluates the value. The new concept involves a statistical decision-making procedure, which avoids the repetitive iterations in traditional completely random heuristic optimizations. In addition, a two-step optimization scheme for local stiffness regulation is developed to expand the design space of one single material. This method facilitates the personalized customization of both the multi-objective deformation pattern and the deformation path of highly deformable mechanical metamaterial made from homogeneous rubbers. The efficiency of the design process is verified by experiments with 3D-printed metamaterial samples. This research offers a novel solution to the large deformation design of shape-morphing structures and soft robotics.
Advanced Science, EarlyView. Read More
