On-demand Reprogrammable Mechanical Metamaterial Driven by Structure Performance Relations.

The physical reprogrammability of metamaterials provides unprecedented opportunities for tailoring changeable mechanical behaviors. It is envisioned that metamaterials can actively, precisely, and rapidly reprogram their performances through digital interfaces toward varying demands. However, on-demand reprogramming by integration of physical and digital merits still remains less explored. Here, a real-time reprogrammable mechanical metamaterial is reported that is guided by its own structure-performance relations. The metamaterial consists of periodically tessellated bistable building blocks with built-in soft actuators for state switching, exhibiting rich spatial heterogeneity. Guided by the pre-established relations between state sequences and stress-strain curves, the metamaterial can accurately match a target curve by digitally tuning its state within 4 s. The metamaterial can be elastically tensioned and compressed under a strain of 4%, and its modulus tuning ratio reaches >30. Moreover, it also shows highly tunable shearing and bending performances. This work provides a new thought for the physical performance reprogrammability of artificial intelligent systems.