Image-based numerical characterization and experimental validation of tensile behavior of octet-truss lattice structures

The production of lightweight metal lattice structures has received much attention due to the recent developments in additive manufacturing (AM). The design flexibility comes, however, with the complexity of the underlying physics. In fact, metal additive manufacturing introduces process-induced geometrical defects that mainly result in deviations of the effective geometry from the nominal one. This change in the final printed shape is the primary cause of the gap between the as-designed and as-manufactured mechanical behavior of AM products. Thus, the possibility to incorporate the precise manufactured geometries into the computational analysis is crucial for the quality and performance assessment of the final parts. Computed tomography (CT) is an accurate method for the acquisition of the manufactured shape. However, it is often not feasible to integrate the CT-based geometrical information into the traditional computational analysis due to the complexity of the meshing procedure for such high-resolution geometrical models and the prohibitive numerical costs. In this work, an embedded numerical framework is applied to efficiently simulate and compare the mechanical behavior of as-designed to as-manufactured octet-truss lattice structures. The parts are produced using laser powder bed fusion (LPBF). Employing an immersed boundary method, namely the Finite Cell Method (FCM), we perform direct numerical simulations (DNS) and first-order numerical homogenization analysis of a tensile test for a 3D printed octet-truss structure. Numerical results based on CT scan (as-manufactured geometry) show an excellent agreement with experimental measurements, whereas both DNS and first-order numerical homogenization performed directly on the 3D virtual model (as-designed geometry) of the structure show a significant deviation from experimental data.

Automated summary

The production of lightweight metal lattice structures, thanks to additive manufacturing, has led to new design flexibility and complexity in terms of physics. However, process-induced geometrical defects can result in deviations between the designed and manufactured geometry, impacting mechanical behavior. Utilizing computed tomography (CT) for accurate shape acquisition is important, but integrating this information into computational analysis is often challenging due to meshing complexity and costs. Effective methods for simulating and comparing mechanical behavior of as-designed and as-manufactured lattice structures, such as the Finite Cell Method (FCM), are crucial for quality assessment.