*3.4. Porosity*

In order to distinguish between pores and trapped particles that fell out after the tensile test, a micro-CT scan of both samples was performed. Figure 7a shows the scan image of one half of the solid sample L10D07 through the center of the specimen. Figure 7b displays the right two struts close to the rupture. Several small pores with diameters between 10 μm and 40 μm can be observed. Scans of the support sample showed, as expected, no signs of pores, and they were therefore not included in this article.

**Figure 7.** Micro-CT scan of the solid sample L10D07 with porosity. (**a**) Overview of all four struts. (**b**) Magnification with visible pores

Table 4 lists the mean properties of the porosity of the samples with 0.6 mm and 0.7 mm diameter struts for both parameter sets. It is clearly visible that samples produced with the solid parameter set contain more pores and thereby have larger defect surfaces. In total there is around 2 % of the fracture surface subject to porosity. There is no obvious reason for the solid parameter combination to show porosity, as the combination is located in the fully dense region of Figure 1. Still, it might result from an already too high energy input or the fact that small amounts of porosity were not considered in the definition of the fully dense zone. Occasional pores in the support samples seem to have no significant differences in diameter. The sample L10D07S showed a single pore in one of the four struts. As a mean value, this leads to an amount of zero pores, but a remaining defect surface of 61 μm<sup>2</sup> still.

**Table 4.** Mean properties of pores, visible on the fracture surfaces of the test samples.


#### **4. Discussion and Conclusions**

In this article, the characteristics of thin struts, which act as a support structure, were analyzed for two sets of build parameters. Tensile tests and subsequent fracture surface analysis were performed. It was shown that semi-molten and sintered particles on the outside surfaces of the samples affect the load bearing of the parts as the designed dimensions include these regions with particularly lower bonding strength. Small amounts of porosity in the parts showed no effect on the overall tensile strength.

Compared to round tensile test specimens produced according to the ASTM standards, the ultimate tensile strength evaluated in this article, is significantly lower. For cross-sectional areas calculated with the outside diameters of the specimens, the UTS was below 900 MPa and thereby strongly deviated from the range reported by Kasperovich and Hausmann given as 1040 MPa to 1062 MPa [20]. On the other hand, if only the core area is considered, values above 1300 MPa are even higher than the strength of as-built and machined samples, for which Vilaro et al. evaluated 1166 MPa [21] and Rafi et al. 1219 MPa [12]. This confirms that the outside, non-fully-molten region carries significantly less load than the core region, while still contributing to the overall load bearing. The values for L10D06S are considered statistical outliers, as the failure at low stress levels might result from a slightly tilted clamping or minor defects in the sample. The high standard deviation indicates that it was an unusual failure and if added to the UTS value it results in the same range as the other values.

The fracture surface analysis revealed a mixed fracture behavior showing micro-dimpled regions for ductile and quasi-cleavage sections for brittle failure. This is in accordance with results from Krakhmalev et al. [13] and Bassoli and Denti [19]. The cup-and-cone shape of the necking region reported by Krakhmalev et al. [13], however, was only present to some extent for the struts with small diameters. Additionally no distinct necking could be observed, as the diameter reduction was only in the range of the standard deviation of the measurements.

The results discussed in this article helped us to further understand the behavior of a support structure built with specific support parameters compared to solid parameters. Using these results, support structures can be investigated in process simulations with adapted material characteristics or be designed to meet load bearing needs. In addition, it was shown that the small diameters of the struts affected the fracture behaviors and tensile strengths of the samples. The authors sugges<sup>t</sup> for future research to investigate the effect of the diameter on the micro structure, especially the differences between outside regions and the cores of the samples.

**Author Contributions:** Conceptualization, S.W. and J.M.; data curation, S.W. and T.S.; investigation, S.W., C.P. and T.S.; resources, C.P.; supervision, K.P.; writing—original draft, S.W. and J.M.; writing—review and editing, M.B. and K.P. All authors have read and agreed to the published version of the manuscript.

**Funding:** This research received no external funding.

**Acknowledgments:** The authors like to thank the 3D Printing Center at the Bundeswehr Research Institute for Materials, Fuels and Lubricants (WIWeB) for their assistance with the L-PBF print of the test samples and for providing the CT and SEM images.

**Conflicts of Interest:** The authors declare no conflict of interest.
