Inverse designing metamaterials with programmable nonlinear functional responses in graph space
This work addresses the challenge of engineering materials with tailored functionalities for applications such as protective equipment and electric vehicles, representing a novel method for a known bottleneck in metamaterial design.
The authors tackled the problem of designing 3D metamaterials with complex nonlinear functional responses, such as stress-strain curves and viscoelastic behaviors, by developing GraphMetaMat, a graph-based framework that integrates graph networks, physics biases, reinforcement learning, and tree search to achieve programmable responses and outperform commercial materials in applications like cushioning and vibration damping.
Material responses to static and dynamic stimuli, represented as nonlinear curves, are design targets for engineering functionalities like structural support, impact protection, and acoustic and photonic bandgaps. Three-dimensional metamaterials offer significant tunability due to their internal structure, yet existing methods struggle to capture their complex behavior-to-structure relationships. We present GraphMetaMat, a graph-based framework capable of designing three-dimensional metamaterials with programmable responses and arbitrary manufacturing constraints. Integrating graph networks, physics biases, reinforcement learning, and tree search, GraphMetaMat can target stress-strain curves spanning four orders of magnitude and complex behaviors, as well as viscoelastic transmission responses with varying attenuation gaps. GraphMetaMat can create cushioning materials for protective equipment and vibration-damping panels for electric vehicles, outperforming commercial materials, and enabling the automatic design of materials with on-demand functionalities.