Incorporating defects into model predictions of metal lattice-structured materials

Abstract

Metal lattice structures are optimized for high specific strength properties and are now customizable using additive manufacturing methods. However, many of these methods, like selective laser melting, can introduce defects such as porosity, parasitic material, and disconnected struts into the structure, which can negatively affect mechanical behavior. While there are many computational models used to predict the mechanical response of lattice-structured materials, most use an idealized structure and are often not robust enough to account for defects. In this study, metal lattice structures are characterized for defects and mechanically tested in compression. The defect types and distributions are characterized using x-ray micro-tomography and the tomography analyses is used in two different model predictions. First, in an equivalent continuum model, where the defects are used to predict the variability in mechanical properties, and second, in a finite element analysis, where the predicted stress-strain response for both realistic and idealized structures are compared. Investigating this further, a finite element analysis of an octet lattice quantifies the reduction in strength associated with disconnected struts and captures a dependence on the strut's orientation relative to the loading direction. Overall, incorporating defect information gleaned from tomography data improves predictions of mechanical properties by capturing a more realistic deformation response for lattice-structured material.