Evolutionary computation for design and characterization of nanoscale metastructures

Abstract Designing architected metastructures with desirable characteristics is typically associated with complex fabrication and testing procedures. The challenges ahead for nanoscale fabrication of these engineered structures lead to severe obstacles to investigate their complex design patterns and corresponding mechanical properties. Here, we introduce a striking artificial intelligence concept based on evolutionary computation for the characterization of the mechanical properties of nanoscale metastructures with corrugation. The corrugated metastructures are fabricated and experimentally tested at the nanoscale. New evolutionary computation-based models are presented to characterize and predict the tensile stiffness and postbuckling compressive stiffness of the metastructures. The proposed models are the outcomes of an extensive simulation process which involved evolving millions of design models with different combination of the predictor variables. Thousands of data obtained from the experimentally-calibrated numerical models are used to develop the models. Further experiments are carried out to verify the efficiency of the evolutionary computation models for the prediction of the tensile and postbuckling compressive stiffnesses. In the end, we envision that the proposed evolutionary computation approach can be integrated with topology optimization to evolve optimized metastructures with programmable mechanical characteristics.