PSI - Issue 57

Okan Yılmaz et al. / Procedia Structural Integrity 57 (2024) 420 – 427 Yılmaz and van Hoecke / Structural Integrity Procedia 00 (2023) 000–000

426

7

(a)

Crack face

Hole face

Milled-cut edge

Crack face

Hole face

(c)

(b)

Fig. 6. Fracture surfaces of (a) sample with a punched hole, (b) sample with a laser-cut hole, and (c) a four-point bending sample with a milled cut-edge for tensile face.

1000

1000

S960MC, combined

S700MC, combined

Run-outs

Run-outs

100

100

Laser Punched (low clearance) Waterjet (4PB)

Laser Punched (low clearance) Waterjet (4PB)

Laser (4PB) Milled (4PB)

Laser (4PB) Milled (4PB)

FAT90 (m=3, corrected for R=-1) FAT80 (m=5, corrected for R=-1) FAT160 (m=5, corrected for R=-1)

Stress amplitude (MPa)

Stress amplitude (MPa)

FAT90 (m=3, corrected for R=-1) FAT80 (m=5, corrected for R=-1) FAT160 (m=5, corrected for R=-1)

10

10

1.E+04

1.E+05

1.E+06

1.E+07

1.E+04

1.E+05

1.E+06

1.E+07

N (cycles)

N (cycles)

Fig. 7. Combined fatigue results of axially loaded and four-point bending fatigue tests of S700MC (left) and S960MC (right). Four-point bending results are converted to account the di ff erence in load ratio and stress concentration factor of the hole. All results are then compared with relevant detail categories. Arrows denote run-outs.

Once divided by the stress concentration factor of the hole, four-point bending test results were comparable to the axial tests and in general lie in the safe side of the design curves. Only for S960MC, at low to mid-cycle fatigue life region, some points fall below the FAT160 line, which could be a result of the inaccuracy in the stress conversion. For better comparison with the design curves in practice, it is suggested that these cut edges are tested using classical-shaped fatigue samples in axial loading mode. Profile measurements showed that laser-cut hole surfaces were rougher than punched holes with a distinctively wavy shape. Punching with a small clearance value (c = 0 . 05 mm) made it possible to create smooth punched holes and this was demonstrated with achieved hole profiles. These specimens performed much better than the ones in Debruyne et al. (2019). With high punching clearance resulting in rough and damaged holes, FAT80 design line remained the norm. Using post-treatment methods such as deburring, fatigue results are likely to improve particularly when the fatigue performance is less tolerant of defects. Four-point bending tests of S700MC demonstrated similar fatigue performance of laser-cut and milled faces. Water-jet faces performed the worst, albeit there were minimal data