PSI - Issue 19

Yukitaka Murakami et al. / Procedia Structural Integrity 19 (2019) 113–122 Yukitaka Murakami et al. / Structural Integrity Procedia 00 (2019) 000–000

121

9

Figure 13 shows the fatigue crack emanating from a small defect which was present at notch root of a notched specimen. If the fatigue limit of for un-notched specimen  w0 with the same defect is estimated by the √area parameter model, we have  w0 = 477MPa. The values of applied stress  a = 450MPa and fatigue life N f = ~10 6 are quite higher in spite of stress concentration factor K t = ~1.6 than the cases with large defects shown in Fig. 2 and Fig. 11. For actual fatigue design, we need to consider the probabilistic analysis of actual components and defect size distribution of statistics of extremes. Thus, the statistics of extremes must be the basic concept for the quality control. 8. Effects of surface roughness Figure 14 shows an incomplete trial for eliminating the as-built surface roughness. The sample surface after polishing looked shiny but there is still hidden roughness. In this case, R a = ~10  m but the maximum R max = ~100  m which reduces fatigue limit drastically. The surface roughness with the morphology as Fig. 14(a) of the order of R max = ~100  m can be categorized by the small crack problem. If the fatigue limit in this case is estimated by the √area parameter model, it is ~360MPa. This value is far below the ideal fatigue strength, ~580MPa estimated from Fig. 1 and Fatigue-Grade is still 3, though improved from Fatigue-Grade 1.

Fig. 14 Effect of surface polish. (a) Surface morphology, (b) Surface profile, R a = 13μm, R max = 107μm.

9. Effect of defects on tensile properties Although tensile properties of AM materials have been reported in many literature, we need to carefully treat the data, because tensile properties are insensitive to small defects which absolutely cause detrimental influences on fatigue strength. Even if the tensile strength is almost the same level of rolled materials, the reduction of area is influenced by presence of small defects (see Table 1).

Fig. 15 Tensile specimen fractured by “incomplete cup-and-cone” fracture due to presence of defects. Tensile strength is the same level of rolled material. Large scatter in Reduction area ranging from 25% to 47%. (a) Overall view of fractured specimen in tensile test. Full cup-and–cone fracture is not made. (b) View of fractured section. Tensile strength: 1020MPa, Reduction of area: 26%.

It must be noted that elongation in tensile test has no physical meaning, because it varies depending on the gauge length of specimens. Figure 15 shows a specimen which fractured by incomplete cup-and-cone shape and reduced the reduction of area. This type of shear fracture occurs in tensile test on low cycle fatigue specimens after fatigue crack grows to a critical size (Murakami and Miller (2005)). Therefore, we cannot evaluate exactly the fatigue quality of AM materials by tensile test properties. The reduction of area of rolled materials has less scatter while AM specimens containing defects show a big scatter of the reduction area as shown in Table 1.