PSI - Issue 39

11

Hannes Panwitt et al. / Procedia Structural Integrity 39 (2022) 20–33 Author name / Struc ural Integrity Procedia 00 (2019) 000–0 0

30

a

b

mode I pre-crack

notch

back side

back side bottom half

5 mm

c

d

front side top half

front side

main crack

mode I pre-crack

notch

5 mm

Fig. 7: Fracture surface of a specimen after an in-phase mixed mode loading: a) Overview of the back side; b) Detail of the back side with arrest marks; c) Overview of the front side; d) Detail of the front side with arrest marks.

The resulting a-N- curves are shown in Fig. 8a and b. The curves show an increasing crack growth rate for the back side and a rather constant crack growth rate for the front side, resulting in a shorter crack length on the front. Furthermore, these figures show the a-N- curves determined with both methods as well. With the calibrated threshold value of th cw = 0.3 % a good agreement of the curves with the respective results from the marker load technique have been obtained (Fig. 8a). Furthermore, the data obtained with the new methods show a good resolution of the crack growth, because both methods can represent the well-known retardation effect after an overload.

15

15

30

30

a

b

Back side: Markerload th c w

12.5

12.5

25

25

= 0.3%

10

10

20

20

= 0.4-1.0%

th

ε 1

7.5

7.5

15

15

Front side: Markerload th c w

9

10

11

12

9

10

11

12

a [mm]

a [mm]

10

10

= 0.3%

= 0.4%

th

ε 1

5

5

= 0.4-1.0%

th

ε 1

0

0

0

2

4

6

8

10

12

14

16

0

2

4

6

8

10

12

14

16

N [-]

N [-]

10 5

10 5

Fig. 8: Validation of the a - N -measurement methods for mixed mode loading conditions with the marker load technique: a) Crack width method; b) ε1 -method.

But, i n comparison to the crack width method, the data obtained with the ε1 -method (Fig. 8b) are less noisy and the retardation effect caused by the overloads is more clearly visible. To achieve these results, the calibrated non-