PSI - Issue 7

I. Milošević et al. / Procedia Structural Integrity 7 (2017) 327 –334

332

6

I. Milosˇevic´ / Structural Integrity Procedia 00 (2017) 000–000

0 . 7 0 . 75 0 . 8 0 . 85 0 . 9 0 . 95 1 1 . 05 1 . 1 1 . 15 1 . 2

Normalised stress, σ norm , - (log.)

Surface Subsurface None

10 4

10 5

10 6

10 7

10 8

Number of cycles, N (log.)

Fig. 8. S-N data for specimens D 7 at T = 350 ◦ C , R = − 1.

a b Fig. 9. Characteristic subsurface failure of D 7 . 5, T = 350 ◦ C specimens. Cracks mainly emanated from slip planes.

Table 3. Fatigue tests and fractography results for D 7 . 5 specimens at T = 350 ◦ C. Nr. σ norm Cycles

√ area

ECD

µ m

µ m

1 2 3

0.82 0.78 0.78

4.749.160 3.886.450 2.856.410

20.48 38.92 35.49

23.14 43.91 40.04

The inclusions, which were investigated in D 7 . 5 specimens and at 350 ◦ C, indicated a bigger size compared to D 7 . 5 at 20 ◦ C. It is clear that a comparison ( D 4 and D 7 . 5 ) is very di ffi cult due to the di ff erent defect sizes observed. No clear reason could be identified so far. The mechanism of subsurface cracking started at a similar area of 10 7 cycles. The average value of √ area from D 4 to D 7 . 5 decreased nearly threefold. Besides the assessment of defects ( √ area ) other influences, such as stress concentrations, surface layer e ff ects of the specimen, local properties within the defect area and the size of the specimen could have made a greater contribution to the fatigue behaviour. The presented results cannot be taken into account for a serious fatigue strength estimation up to 10 9 cycles. The fact that di ff erent inclusion sizes were found within di ff erent specimen geometries was another critical issue. Results of the √ area approach can be seen in Figure 4 and 7 in the previous section. For D 4 specimens, which showed bigger inclusion sizes, a moderate defect influence on the normalised stress S a , norm could be observed. Con ventional approaches (discussed in section 3) were displayed in Figure 10 and Figure 11. Using current findings for conventional lifetime assessments it is obvious that after the turn-o ff point a horizontal course should be avoided due to the material’s fatigue behaviour (no infinite life up to 2 × 10 7 ). A Miner elementary approach, where the fatigue strength is extended ( k 2 = ∞ ) straight to 10 9 cycles should not be considered. The approach used for actual tests, Miner’s approach modified by Haibach with a k 2 = 2 ∗ k 1 − 1, represented a preliminary study with space for further improvements.