PSI - Issue 39

Hannes Panwitt et al. / Procedia Structural Integrity 39 (2022) 20–33 Author name / Structural Integrity Procedia 00 (2019) 000–000

27

8

Element

C

Mn

Si

P

S

Cr

Mo

Ni

Al

Composition

0.36

0.73

0.26

0.006 0.001

1.57

0.22

1.57

0.022

Table 2. Mechanical properties of the tested material 34CrNiMo6. (Pilsen Toll s.r.o., 2016) R p0.2 in MPa R m in MPa A 5 in % Hardness in HB E in MPa ν 819 963 15 326 210.000 0.3

For crack initiation, a wire eroded V-shaped sharp starter notch (detail in Fig. 4a) is used. A crack is initiated under pure mode I. Therefore, a cyclic tension-compression load ( F max = 30 kN and R = -1) is applied for 10 5 load cycles. This results in an initial crack length of approx. ∆ a = 0.3 mm, which is measured with the direct current potential drop technique and corresponds to a change in the electrical potential of 2%. Both, the mode I crack initiation and the crack propagation experiments were performed on a servo-hydraulic tension/torsion testing machine (Fig. 4b). In the experimental setup, a digital camera was pointed at each specimen side to investigate the crack propagation behavior at the respective specimen surface. In this context, every 4,000 cycles in each experiment an image at F max has been recorded. The resulting image sequence over the course of the test was then analyzed in the DIC post processing program VIC-2D v6 (Correlated Solutions, Inc., 2016). 4.2. Calibration under mode I loading conditions For both new methods the respective threshold values need to be calibrated first. Therefore, two different tests on mode I fatigue cracks were performed using the direct current potential drop technique. The resultant a-N- curves are used as the reference for the calibration. The first test was conducted K I -controlled in terms of a block loading in the lower Paris regime with a constant block loading ratio. As a result of this, the typical retardation effect of the crack growth occurs (Fig. 5a and b). The calibration results show the best agreement with the reference curve measured with the potential drop technique for th cw = 0.3 % (Fig. 5a) and th ε1 = 0.4 % (Fig. 5b), respectively. It should be mentioned that th cw = 0.3 % corresponds to a crack width of 3 µm, because the distance between the reference points is set to 1 mm (chapter 3.1.1). However, for both methods, an increasing overestimation (up to 4 %) of the crack length is visible with an increasing crack length (Fig. 5c). This is caused by the crack path detection from discrete pixel locations with only horizontal, vertical or 45° directions possible (see the scheme in Fig. 5c). Smoothing the crack path shortcuts the sharp kinks to resemble the true crack path with better accuracy. The longer the crack, the more the effect becomes apparent. Thus, the effect of smoothing is the greatest at larger crack lengths and results into more accurate a - N -Curves. For all further investigations, only smoothed crack paths are evaluated.

a

b

c

35

35

Potential path not smoothed path smoothed

Potential th ε 1

Potential th c w

30

30

= 0.2%

= 0.1%

25

25

= 0.4%

th

= 0.3%

th

c w

ε 1

= 1.0%

th

= 1.0%

th

20

20

c w

ε 1

30

15

15

pixel

a [mm]

a [mm]

10

10

25

a [mm]

smoothing

5

5

0

20

0

0

0.5

1

1.5

2

2.5

3

2

2.2

2.4

2.6

2.8

0

0.5

1

1.5

2

2.5

3

N [cycles] N [-]

10 6

N [cycles] N [-]

N [cycles] N [-]

10 6

10 6