PSI - Issue 68

M. Mahatab et al. / Procedia Structural Integrity 68 (2025) 815–821 M. Mahatab and R. Ranjan / Structural Integrity Procedia 00 (2025) 000–000

819

5

available in the literature.

100 150 200 250

100 150 200 250

0 50

0 50

Nominal Stress (MPa)

Nominal Stress (MPa)

0 100 200 300 400 500

0 100 200 300 400 500

Peak Number

Peak Number

(a) (b) Figure 3: Loading history used in SBFM analysis, (a) VA1; (b) VA2

With the assumed semi-elliptical crack geometry and tensile residual stress distribution, the SBFM results show good predictions as compared to the fatigue test results from the literature for CA loading ( R = 0.1). Most of the fatigue test results are in good agreement with SBFM results on and before 1 million fatigue life cycles. Some test results from Lindqvist (2002) give lower fatigue life values than SBFM prediction due to the high thickness of the specimen (i.e. 12 mm) from the others. The prediction shows the potential of SBFM for the fatigue life evaluation of Indian high steel grades. In Figure 4(a), the fatigue class (FAT 63) curve for cruciform joints with fillet weld given in the International Institute of Welding (IIW) recommendation (Hobbacher (2016)) is also plotted with a black dotted line. The prediction curve for both bounds lies above the IIW FAT curve. The difference in the FAT curve and SBFM S-N curve is due to the deterministic nature of SBFM results. Figure 4(b) presents the SBFM S-N curves for VA1 and VA2 loading conditions along with the CA result to assess the effect of variable amplitude loading on the fatigue life of S550 steel in as-welded conditions. A significant reduction in fatigue life because of the VA1 and VA2 loading conditions is observed compared to fatigue life in CA loading using the SBFM model. The equivalent stress range at 2 million cycles (fatigue strength) was found to be 29% lower in the case of VA1 loading and 34% lower in the case of VA2 loading than the fatigue strength in the case of CA loading. The fatigue strength reduction kept increasing for the high cycle fatigue region due to the variable amplitude loading effect.

1000

1000

100

100

Lindqvist (2002) Gustafsson (2006) William Vincent Ellerbeck (2023)

CA VA1 VA2

SBFM -0.1 σy SBFM 1.5 σy FAT 63 Curve

10

10

2,E+04 Nominal Stress Range (MPa) 2,E+05

2,E+06

2,E+07

2,E+08

1,E+04 Nominal Stress Range (MPa) 1,E+05

1,E+06

1,E+07

1,E+08

Number of cycle (N)

Number of cycle (N)

(a) (b) Figure 4: 2D SBFM analysis result for as welded S550 steel under CA loading condition (a) and for different loading conditions

Figure 5 shows the crack shape evolution curve obtained using the 2D SBFM model for different loading conditions. The crack aspect ratio for assumed geometry at different maximum stress levels is plotted on the vertical axis, and the normalized crack depth is plotted on the horizontal axis. The applicability of weight functions utilized in 2D SBFM analysis lies within the range of crack aspect ratio of 0.2 to 0.8. It can be observed from Figure 5(a) that when the crack is propagated up to 60% of the depth of the specimen, there is a sudden drop in aspect ratio, representing a failure of the specimen or a change of semi-elliptical crack to a through-width crack. At the stress level of 180 MPa in CA loading condition as in Figure 5(a), initially, the aspect ratio increased from 0.5 to 0.65, i.e. propagation of the crack is dominant in the depth direction followed by a decrease in aspect ratio to a value of 0.47