PSI - Issue 38

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

62

3

strength behaviour, meaning different fatigue behaviour for the horizontally (relative to the building platform) and vertically built specimens, see Blinn et al. (2018) for 316L. In general, the fatigue lives of SLM specimens are significantly lower compared to conventionally manufactured specimens, as e.g. shown by Zhang, Li et al. (2017) for Ti-6-Al-4V, yet might be improved by post processing such as hot isostatic pressing as stated by Leuders et al. (2014). Furthermore, the dominating fatigue life determining characteristics and thus the fatigue behaviour seem to differ depending on the material. Thus, it may not be possible to make general statements on fatigue behaviour of AM materials. For example, Leuders et al. (2014) report process induced pores as dominating for Ti-6-Al-4V, yet report the monotonic strength as key parameter for 316L. 3. Material model data - 316L As stated in the previous section, the material properties of SLM specimens may be non-isotropic. Some authors like Lindström et al. (2018) and Bräunig et al. (2015) assume the material to be transversely isotropic because of almost uniform material properties in each material layer. In combination with the rotation of the scanning direction of each layer, material properties are only changing in the vertical (i.e. building) direction, thus a transversely isotropic material model seems justified. This aspect should be included in fatigue calculations, however, due to the lack of available material data, instead of a transversely isotropic model, the smaller directional material values are often used for an isotropic material model. In the following, the fatigue models are applied for one of the most commonly used materials for AM, the austenitic steel 316L, also referred to as 1.4404. The material data for 316L manufactured via SLM in different directions is given in Tab 1. Poisson’s ratio is assumed to have a constant value of 0.3.

Table 1. Direction-dependent material data for 316L taken from Gläßner et al. (2017).

Manufacturing direction Horizontal Vertical

Young’s modulus [GPa]

167 609 681 218

152 490 612 213

��,� [MPa] � [MPa]

Hardness [HV30/10]

4. Fatigue evaluation approaches The fatigue life of components may be divided into the stages of crack incubation and crack propagation until final failure, see Radaj and Vormwald (2007). Here, we use crack incubation to determine the fatigue limit. For the following calculations, unnotched fatigue specimens under uniaxial tension-compression loading with load ratio R = - 1 are assumed. Based on the aforementioned characteristics, two different approaches regarding the fatigue evaluation are applied, the defect stress gradient (DSG) approach and a Weibull distribution. For both concepts consideration of the features mentioned above is possible as will be shown and discussed in the following. Regarding the feature of porosity, however, although they share the same assumption regarding non-interacting defects, one of the approaches uses an explicit, the other an implicit way of incorporation. 4.1 Defect stress gradient approach As stated before, one of the characteristic features of AM materials are pores and other defects. Their shapes may be almost spherical in the case of gas porosity or highly irregularly shaped in the case of lack of fusion defects. The