Soft network metamaterials are widely used in fields such as flexible electronics, tissue engineering, and biomedicine due to their superior properties including low density, high stretchability, and high breathability. However, the prediction and customization of the nonlinear mechanical behavior of soft network metamaterials remain a challenging problem. In this study, a family of hydrogel-based network metamaterials with biological tissue-like mechanical properties are developed based on a machine learning-driven optimization design method. Numerical and experimental results explain the relationship between the mechanical properties of the designed metamaterials and their microstructural features and stretching ratios. The results indicate that the hydrogel-based network metamaterials exhibit J-shaped stress-deformation (σ-λ) behavior similar to biological tissues. This phenomenon arises from the transition of the deformation mode of metamaterials from bending-dominated to stretching-dominated as the stretching ratio increases. Based on the proposed design scheme, the Poisson's ratio of metamaterials can be adjusted within a remarkably wide range of -1.06 to 1.34. Furthermore, through optimizing the design parameters of the metamaterial, the customization of network metamaterials with biological tissue-like zero Poisson's ratio behavior and stress response is achieved. The potential applications of hydrogel-based network metamaterials are demonstrated through artificial skin and LED integrated device. This research offers novel insights into predicting, designing, and fabricating the mechanical behavior of soft network metamaterials.
Read full abstract