A configuration-driven nonlocal model for functionally graded lattices

Abstract

Existing nonlocal models cannot accurately capture the size-dependent mechanical behavior of functionally graded lattices because they assume constant intrinsic length, which oversimplifies the nonlocal effects of varying lattice topology microstructures. In this paper, we unveil that the intrinsic length obeys a gradient law determined by the configuration of the functionally graded lattices. Based on the unveiled gradient law, a configuration-driven nonlocal model is developed to predict the size-dependent mechanical behavior of axially graded lattices. An offline dataset of the intrinsic length is constructed based on the gradient law and the high-throughput simulations. With the help of the offline dataset, the configuration-driven nonlocal model can be used to accurately and efficiently analyze the mechanical behaviors of the functionally graded lattices online. The configuration-driven nonlocal model improves the accuracy of the classic micromechanics homogenization method and reduces the computational cost of the high-resolution finite element method. The developed model not only guides the design of functionally graded lattices but also offers an effective multiscale approach for their performance prediction.