Additive Manufacturing is one of the most important enablers of metamaterials as it provides unprecedented design flexibility. However, additively manufactured structures are typically riddled with randomly distributed defects, e.g., surface roughness, microcracks, or embedded brittle phases that reduce mechanical performance and introduce a high degree of uncertainty. Due to their rather brittle behavior, many materials that are good performers in extreme environments (e.g. ceramics and refractory metals), are particularly susceptible to such manufacturing induced flaws, limiting their applicability as constituents for metamaterials. This raises the question of how metamaterials should be designed to mitigate these drawbacks, and what acceptable levels of variability are. We have developed a high-throughput finite-element tool that we use in a Monte Carlo approach to investigate how local property variations in 3D printed strut-based lattice materials influence effective properties. The study provides guidance for design engineers in choosing lattice topologies based on robust stochastic failure criteria and for process engineers in choosing processing targets.

For structures made from perfectly elastic-brittle parent materials, we show how variability of local properties affects localization of damage in three different lattice topologies. Localization of damage – or the lack thereof – drives effective properties such as scatter of effective tensile strength and gives rise to surprising effects like “apparent” ductility [1]. For parent materials that can accommodate plastic deformation, we show that at a critical threshold ductility, the entire strut-based structure yields before any strut ruptures. The widespread plastic yielding leads to a loss of constraint for many struts, leading to large rotations and a localization of strains, followed by fracture. Past this point, the effective strength of the lattice material is no longer dominated by the ductility of the constituent material, but by the plastic collapse load of the structure. We determine the threshold where the dominating failure mechanism transitions from strut rupture to plastic collapse as a function of topology, relative density and the degree of variability in local material properties. 

[1] P. Ziemke et al., Materials & Design, 2024, 239, 112776.