Tailoring Stress-Strain Curves of Flexible Snapping Mechanical Metamaterial for On-Demand Mechanical Responses via Data-Driven Inverse Design

By incorporating soft materials into the architecture, flexible mechanical metamaterials enable promising applications, e.g., energy modulation, and programmable shape morphing, with a well-controllable mechanical response. However, the lack of achievable spatial and temporal programmability hinders the way of flexible mechanical metamaterials towards higher-level mechanical intelligence, such as autonomous sequential behaviors. One feasible solution is to introduce snapping structures that can modulate the absorption and release of mechanical energy and then tune their responses by accurately tailoring the stress-strain curves. However, owing to the strong coupling of non-ideal architecture (the structures that are hard to be simplified as hinges, springs, and beams, etc.), nonlinear structural deformation, and nonlinear material constitutive model, it is difficult to deduce the stress-strain curve of snapping metamaterials using conventional modeling ways. Here, a machine learning pipeline is trained with the finite element analysis data that considers those strongly coupled nonlinearities mentioned above to accurately tailor the stress-strain curves of beam-based snapping mechanical metamaterial sheets for on-demand mechanical response. Such a machine-learning-based data-driven inverse design method shows a good accuracy of 97.41% in the testing dataset, and our prediction, as well as the simulation results, conform to experimental data well. Utilizing the established approach, the energy absorption efficiency of the snapping metamaterial-based device can be tuned within the accessible range to realize different rebound heights of a falling ball. And reconfigurable soft actuators that actuated by a single pneumatic energy input can be spatially and temporally programmed to achieve synchronous and sequential actuation with the flexible snapping metamaterial skins designed through our method. Purely relying on structure designs, the accurately tailored snapping metamaterials increase the tunability of devices. Such an inverse design approach can potentially extend to similar nonlinear scenarios towards more responsive and interactive systems with embodied mechanical intelligence.