PSI - Issue 42

Aditya Pandey et al. / Procedia Structural Integrity 42 (2022) 1017–1024 Pandey et al. / Structural Integrity Procedia 00 (2019) 000–000

1021

5

3.2. Cyclic loading behavior

The curve for stress amplitude σ a vs the number of cycles to failure N f or the S-N curve at R = -1 is shown in Figure 5. The data with some scatter was fitted to a power law. The fatigue lives of as-built samples got improved sig nificantly post heat-treatment. The endurance limit at load-ratio of R = -1 were 200 MPa and 400 MPa for the as-built and heat-treated samples respectively. Rapid thermal cycles during the SLM process might result in the development of residual stress, which can significantly impact the fatigue life in terms of crack initiation sites. The heat-treatment of the as-built specimens could have certainly reduced the residual stresses (Yadav et al. (2022)) and thus can easily ex plain the observed trend. Additionally, the evolution of γ ’ and γ ” strengthening precipitates during the heat-treatment process can also improve the fatigue lives of the heat-treated coupons.

Fig. 5. Stress vs. cycles to failure (S-N) curve at ambient temperature

3.3. Fractography

Figure 6a and Figure 6b show a typical fracture surface for the as-built fatigue tested samples ( σ a = 500 MPa, R = -1, N f = 80534 and σ a = 300 MPa, R = -1, N f = 502189 respectively). Figure 6c shows the same for heat-treated sample ( σ a = 500 MPa, R = -1, N f = 367377). Fracture surface consists of crack initiation, crack propagation, and final fracture zone, clearly marked in Figure 6a. A fatigue crack mostly from the surface except few cases where defect (un-melted powder particles, Figure 6c) was observed. Figure 6b shows striations produced by underloads after every amplitude cyclic loading. The melt-pool measurements were carried out using optical microscopy. The dimensions of molten pool depth and width were calculated from side face of the cubical sample, which was parallel to material deposition direction while the molten pool tracks were measured from top surface, as shown in Figure 7. The penetration into the powder bed was taken as the depth while the length between the two edges of the molten pool as width. The center distance between two laser tracks defined the hatch spacing. Four tracks of laser scan were simulated by incorporating the SLM process parameters listed in Table 2 and the temperature dependent material properties reported in (Lee et al. (2016)): laser power 280 W, laser travel speed of 1000 mm / s, laser spot size of 0.13 mm, hatch spacing of 0.15 mm and the initial temperature 298 K. Solidus, liquidus and latent heat of fusion were assigned as 1533K, 1603K and 22.7 KJ / kg respectively. The convective heat transfer coe ffi cient was considered as temperature independent and was set to 50 W / m 2 / K , the Stefan Boltzmann constant ( σ ) was set to 5.67 × 10 − 8 W / m 2 / K 4 and the emissivity ( ϵ ) of the domain was taken as 0.8. An exponentially decaying Gaussian moving heat source was modeled to accurately simulate the AM process (Lee et al. (2016)). A three-dimensional view of the simulated melt pool contour is shown in Figure 8a-d, where the heat source travelled 3.4. Melt-pool characterization