Crack Paths 2006

200+1.0/-0.1

52+0.1/-0.00 33.0±0.1

30.0+0.1/-0.00

2.50±0.01

8

8

2

0 0

. 0 / - 1 . 0

+ 0 . 1 2 7 / - 0 . 0

4 1 . 2 7 5

2.1 0 0

0 2

4

±.0

+ 0

ø0.25±0.01

45+0.0/-1.0

15.0±0.05

R75±1

Figure 2. Tubular crack growth specimen. All dimensions in mm.

116.84

45.72

45.72

22.86

22.86

4.45

.86

2 2

6.35

R0.50

R7.5

2 . 5 4

Figure 3. Specimen used for modeI crack growth testing. All dimensions in mm.

Specimens were precracked to a total crack length of 2.5mmusing a fully reversed con

stant amplitude nominal stress of 276MPa. Crack growth measurements made below ΔK=10MP√am were taken using a high R-ratio loading cycle such as found that in

Figure 4a. Above that level a periodic overload history with a fully reversed overload

cycle of 276MPawas used, as in Figure 4b. Only in those cases where the required stress

intensity range caused the small cycle minimumto approach that of the overload was the

size of the overall overload cycle increased. Crack length measurements were made using

a 900x long focal length microscope that allowed crack length to be determined to within

2μm. Each crack length measurement was made twice to reduce measurement error, and

each measurement was referenced to the root of the initial 0.50mmradius notch. Crack

opening stress level was verified optically as in reference [12]. Stress intensity factors

were calculated using geometry factors from reference [13].

In the case of the modeII crack closure free crack growth experiments a 220kNax

ial and 2250N-mtorsional servohydraulic frame with hydraulic collet grips was used.

The notched tubular specimens were precracked using a torsional periodic overload strain