PSI - Issue 19
Shota Hasunuma et al. / Procedia Structural Integrity 19 (2019) 194–203 Shota Hasunuma, Ogawa Takeshi/ Structural Integrity Procedia 00 (2019) 000 – 000
201
8
4. Investigation of a method of predicting the fatigue life Small crack growth characteristic was evaluated. Fig. 12 shows the relationship between crack growth rate d a /d N and J integral range J . Crack depth a was determined from surface crack length 2 c under the assumption of a semicircular crack shape. Dowling ’ s simple formulation was used for the calculation of J ; Dowling(1977).
2
3.85(1
π )
π
n
Δ
a
2
2
f
(1)
F ΔJ
F
a Δ Δ n
p
f
E
n
f
where E , n f , p , F are young ’s modulus, cyclic strain hardening exponent, plastic strain range and geometric factor, respectively. We used the values n f = 0.404 and F = 0.714. Solid line shown in Fig. 12 indicates crack growth characteristics of stainless steel described in JSME S NA1-2004: in Kozokenzensei hyouka handbook henshuiinkai, (2005). When J is small, i.e., crack length is small, the experimental d a /d N data is scattered and larger than that of JSME S NA1-2004 because aspect ratio of crack is not 1. However, the data is similar to that of JSME S NA1-2004 when J was larger than 5×10 -4 MPam. The effect of the machined surface layer on the low cycle fatigue life is summarized as follows. We first discuss the residual stress. As shown in Fig. 7, plastic strain was observed in the present study, and it is thus suggested that residual stress was released. Therefore, residual stress does not affect the low cycle fatigue life. The effect of the variation in the material property on the fatigue life is next discussed. The fatigue life of the GR specimen, which had a plastic deformed layer, was similar to that of the EP specimen. Then, crack growth behavior and crack growth characteristics were similar to that of EP specimen. The effect of variation in the material property on the fatigue life was therefore weak. Finally, the effect of the variation in the surface shape on the fatigue life is discussed. Test results show that the effect of the variation in the surface shape on the fatigue life is strong. Cracks initiated in the scratch on the AM specimen. Cracks grew almost as large as scratch in early stage of fatigue life because of crack coalescence. Therefore, the scratches changed the initial size of the fatigue crack. A method for predicting the low cycle fatigue life of a component with machined surface layer was investigated. In this work, the fatigue life was predicted according to elastoplastic fracture mechanics. Cracks in the surface machined layer were modeled using a semi-elliptical crack in a rectangle plate. The proposed method is described as follows. First, the initial crack length and depth are determined. Second, J is calculated. Third, d a /d N is calculated using crack growth characteristics, and the increment in crack length is calculated. The crack length is then updated and the procedure repeated. The effect of the machined surface layer is modeled by changing the initial surface crack length and the depth.
10 -6
JSME S NA1
10 -4 Crack growth rate d a /d N , m/cycle 10 -9 10 -8 10 -7
No.1-EP No.1-GR No.1-AM
No.4-AM Fig. 10 No.4-AM Fig. 11
10 -3
10 -2
J integral range J , MPam
Fig.12 Relationship between crack growth rate and J integral range.
Made with FlippingBook - Online magazine maker