PSI - Issue 28

S. Henschel et al. / Procedia Structural Integrity 28 (2020) 1369–1377

1374

6

S. Henschel et al. / Procedia Structural Integrity 00 (2020) 000–000

Fig. 9. Fracture surfaces: Both flat and slant fracture regions were observed.

(a)

(b)

K I K II

K I K II

100

100

calc. (~ F ) strain gauge

calc. (~ F ) strain gauge

0.5

0.5

= 0°

50

50

0 K I , K II / MPa m

0 K I , K II / MPa m

= 30°

0

50

100

0

50

100

Time / s

Time / s

Fig. 10. K I and K II calculated from the straing gauge signals and from the external force. (a) α = 0 ◦ , (b) α = 30 ◦ .

(a)

(b)

K IQ K IIQ

100

+

K I

K IQ

K II K IIQ

150

2

2

0.5

= 1

0.5

100

50 K IQ , K IIQ / MPa m

50

K IIQ / MPa m

0

0

0

50

100

150

0

15

30

45

0.5

K IQ / MPa m

/ °

Fig. 11. K IQ and K IIQ : (a) as a function of α , (b) approximate fracture criterion.

shortly before fracture, the evaluation of K I by the strain gauges showed a significant decrease. Only a low loads, the measurement of K II by the strain gauges led to plausible results. The main cause for these behaviors were identified: The strain gauges had a relatively large distance to the crack tip. Hence, the plastic zone was avoided. This led, however, to small measurable strains. Furthermore, the strain gauge signals exhibited some noise. For the calculation of K I , three pairwise sums of the strain gauge signal were used, cf. Figure 6. In contrast, three pairwise di ff erences are needed for calculation of K II . Hence, the uncertainty due to the noise was increased.

3.4. E ff ect of mode I / II ratio

Figure 11 shows all measured K IQ and K IIQ for di ff erent loading angles. It was observed that the mode I fracture