PSI - Issue 17

Radomila Konečná et al. / Procedia Structural Integrity 17 (2019) 138 – 145 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

143

6

The notch root conditions are different for Type A+ specimens because the local surface roughness is strongly affected by stair stepping associated to the finite layer thickness, Gebhard et al. (2014). The magnified view of the notch root in Fig. 6b demonstrates the stair stepping introduced by the 50  m thick layers (broken lines).

a)

b)

Fig. 6. Effective notch root profile of a) Type C specimen and b) Type A- specimen. Dashed lines are 50  m-thick-layers.

Finally, the notch of Type A- specimens in Fig. 5 shows the worst quality of all. Specifically, the notch root combines the negative impact of the stair stepping effect of Type A+ notch and the un-supported down-skin surface orientation. The latter generates dross formation, high roughness and lack of geometrical accuracy. The optical inspection of the respective as-built notch roots of the SLM Inconel 718 specimens demonstrated the significant difference in effective quality. The ranking of the quality (from best i.e. Type B to worst i.e. Type A-) matches the experimental ranking obtained by fatigue testing of the four sets of fatigue specimens with as-built notch surfaces. Therefore, the S/N curves shown in Fig. 3 quantify the impact of the SLM technology on notch fatigue behavior. The notch fatigue behavior of SLM Inconel 718 shown in Fig. 3 has been obtained with a non-standard specimen configuration and test methodology, Nicoletto (2017). An assessment of the experimental data is important to support further use of the innovative method. Unfortunately, notch fatigue data of SLM Inconel 718 are scarce in the literature, an important and recent exception being the work of Witkin et al. (2018) that published fatigue data obtained with notched specimens fabri cated on a Concept Laser M2 Cusing (layer thickness 30 μm) , subsequently HIPped, solution treated and aged per industry standards. Axial tensile pulsating loading was performed in compliance with ASTM standard E466. Flat fatigue specimens printed in the horizontal direction and with a mild notch effect and as-built surfaces are considered here. The notched specimens of Witkin et al. (2018) and the present miniature specimens are compared in Fig. 7. The respective thicknesses were 2.5 mm for Witkin’s specimens and 5 mm for the present miniature specimens. Both specimen geometries and loading are characterized by a mild notch severity (i.e. K t = 1.93 vs. K t = 1.63). Fig. 7 demonstrates the considerable difference in overall size and the same 2-mm radius of the notches. Since two notches were present in Witkin’s specimens , it is the worst-quality notch (i.e. down-skin) that controls fatigue crack initiation. On the other hand, the method of Nicoletto (2017) tested one notch at the time. So the two Type A miniature specimens not only characterized in fatigue the horizontal fabrication direction but also, selectively, the notch surface orientation (i.e. up-skin notch of Type A+ vs. down-skin notch of Type A-). 3.3. Assessment of the notch fatigue behavior