English
Inspired by biological systems, trainable responsive materials have attracted emerging research interest in future adaptive and intelligent material systems. However, the trainable materials available so far often cannot perform proactive work, and training only allows for one direction of functional change. Here, we show a thermally trainable hydrogel system consisting of two thermal responsive polymers, in which the volume response of the system during phase transition is enhanced or reduced by the training process above a specific threshold temperature. According to network design, positive or negative training of thermal deformation can be achieved. Importantly, the softening, hardening or toughening of hydrogels can be achieved through the training process. We showed that the trainable hydrogel actuator can perform more active work or perform tasks that are impossible to complete initially. The reported dual network hydrogel provides a new training strategy for bionic soft systems, such as adaptive artificial muscles or soft robots.
