PSI - Issue 38

Michaela Zeißig et al. / Procedia Structural Integrity 38 (2022) 60–69 Zeißig, Jablonski / Structural Integrity Procedia 00 (2021) 000–000

64

5

The DSG was originally applied to casting materials which like AM materials include process-related defects. In Nadot et al. (2020) the applicability of the DSG is shown for AlSi10Mg manufactured via SLM. Extensions of the approach regarding residual stresses, see Nadot and Mendez (2008), or the microstructure have already been carried out, see Vincent, Nadot et al. (2014). The chosen approach regarding input data requires defect-free reference material data. As SLM material might always contain a certain number of defects, in the following, these values are approximated based on the ultimate tensile strength � . Here, a relation of 0.4 between the fatigue limit under tension and compression loading and � is assumed. The defect-free torsional fatigue limit is approximated as 70% of the fatigue limit under tension and compression. Both values are in the range of the data given in Forrest (1962). As stated before, the shape of defects up to a size of about 100 µm may be classified as spherical. A comparison of the effect of increasing defect sizes for this shape at a given fatigue limit is shown as a synthetic Kitagawa-Takashi like diagram in Fig. 1, left. The value of defect size which does not lower the overall specimen fatigue limit increases with the size of the reference defect size, marked for each line with a cross. Furthermore, the effect of the reference defect size on the values for bigger defect sizes becomes less prominent with increasing defect size, thus the asymptotical limit of all three curves is almost the same. Non-spherical defects might be included via FE calculations, however, as each defect varies in shape, using ellipsoidal-shaped defects serves as a first approximation. An ellipsoid with a ratio of approximately 2 between the semi-major and the semi-minor axis is used. Calculations are carried out for the horizontal and vertical building directions and thus adapted defect to load orientation. For reasons of comparison, the DSG is also applied to a spherical pore. The same fatigue limit and reference defect size is assumed. Furthermore, no distinction regarding material data based on building direction is included. The resulting diagram is shown in Fig. 1, right.

Figure 1. Synthetic Kitagawa diagram for 316L based on the DSG approach assessing different defect sizes, left, and geometries, right. Defect geometries and respective load directions are indicated.

As the trend of the curves does not significantly change for defect sizes bigger than 400 µm the curves are only plotted up to this value, although defect sizes well above this level may occur. It may be seen that the fatigue limits for the different shapes vary significantly for sizes apart from the reference value with the transversely orientated ellipsoid showing the lowest values. It is assumed that greater differences in the shape ratio of the ellipsoids will lead to a wider spread between the various curves. The different shape of the curve of the longitudinal ellipsoid and thus the prediction that basically every defect present lowers the fatigue limit, is attributed to the low value of ∇ . This in turn is a result of the combination of reference fatigue limit defect size and occurring defect stresses. Although the fatigue limit promptly reaches almost constant values for increasing defect sizes, in reality, a further reduction seems likely, as in the calculations, the defect size in relation to specimen size is not accounted for.