Characterization of the Properties and Effects of the As-Built Geometry on the Quasi-Static Behavior of Thin-Walled DED-Arc Components

The process of directed energy deposition with electric arc (DED-Arc) is an advanced additive manufacturing process for the production of complex and optimized steel components. A major challenge is the varying surface structure of the components, which is caused by the layer-by-layer production and stochastic environmental influences. This study investigates the relationship between the surface properties of tensile specimens produced with DED-Arc and their mechanical behavior in order to predict the mechanical properties of components in the non-post-processed state. Various manufacturing configurations that can occur in complex, thin-walled componentswere investigated, including regular areas, strong overhangs (75°), starting points of layers and printing pauses. A total of 107 tensile specimens with machined and asbuilt surfaces were examined. The surfaces of the as-built specimens were analyzed according to ISO 25178-2 standards, while the mechanical properties were determined by quasi-static tensile tests. The local deformation behavior was recorded using digital image correlation (DIC). The results showthat certain manufacturing configurations, especially starting points, resulted in significant dales on the surface characterized by reduced thickness and increased notch sharpness. In addition, wave-like structures were observed on overhanging walls in the direction of deposition, which led to a significant reduction in component thickness. The notch sharpness of the dales was determined based on the local curvature and depth of the valley lines. The notch factors, determined by analyzing the local strain fields in the elastic range using DIC, correlate strongly with the surface notch sharpness. Specimens with high notch sharpness also showed high notch factors and resulting increased local strains, which led to lower integral strengths and elongations after fracture. Machined specimens exhibited no significant differences in material behavior regardless of the manufacturing configuration. Based on these results, approximate formulas were developed to predict the mechanical properties of unmachined specimens from the mechanical properties of machined specimens and the geometric surface properties. These integral properties can be used in numerical simulations of components. In addition, proposals for meaningful surface and volume parameters were developed to improve the comparability of studies on surface properties and thus contribute to the optimization of the DED-Arc process.