Designing additively manufactured lattice structures based on deformation mechanisms

Abstract

The post-yield mechanical behavior of additively manufactured lattice structures (AMLS) is governed by the interplay between intrinsic (microstructural) and extrinsic (structural topology) properties at different length scales. Herein, we introduce a novel design optimization approach that accounts for scale separation and size effects, which control deformation mechanisms, to achieve a certain targeted macroscopic mechanical response. The new topological designs are guided by finding a direct correlation between the distribution of local stresses within struts and the underlying microstructures. The local stresses are computed using a strut-level yield criterion that has been calibrated to strut-level tensile, compressive, and shear loading experiments. Therefore, the local response of the struts, including tension-compression asymmetry, build direction dependence, and size effects, are accounted for in the yield surface, enabling a more accurate representation of the local stress state. Accurate calculation of the stress state for a given microstructure and topology combination allows for optimizing the topology for the given strut-level microstructure. The interplay between the topology and microstructure is assessed by investigating the unit cell-level deformation mechanisms and quantifying their influence on the global stress-strain relationship via finite element simulations. Using these relationships, a new set of topologies is designed, built, and validated with experiments. On average, the new topologies demonstrate 40% and 72% improvement in energy absorption capacity and flow stress, respectively, compared to topologies that had been previously optimized using constitutive models, which are homogeneous throughout the unit cell. The goal of the presented article is to demonstrate that simultaneously considering the effects of topology and microstructure on the mechanical behavior of AMLS has the potential to substantially improve key performance metrics, including ultimate strength and energy dissipation. The distinguishing and novel feature of our approach is that the topological optimization is performed while accounting for the heterogeneous distribution of strut-level microstructural features and concomitant mechanical behavior, which leads to new insights relative to peak AMLS structural performance.