Fatigue Crack Paths 2003

Sp. Code a rn rn/a ββ

w Fa Nf Run θθA rA θθB rB ρρ θθLM θθρρ

[°] [mm] [kN] [Cycles] Out [°] [mm] [°] [mm]

[°] [°]

[mm] [mm]

45_05_1 5.00 0.073 0.01557° 13' 10.2 2.00 614898 -28.90.086-45.5 0.6 0.21

45_05_2 4.99 0.069 0.01457° 59' 10.3 2.50 164588 -5.3 0.136-31.2 0.6 0.20

45_05_3 5.01 0.054 0.01157°03' 10.3 1.801572192

-50.6 0.6 0.29

45_05_4 4.97 0.071 0.01457° 54'10.15 2.001627475 -57.60.358-55.1 0.6 0.37

45_05_5 5.1 0.079 0.01557° 06'10.15 2.50 291357 -55.50.260-57.9 0.6 0.34 6 5 08 68 3 12 0 1 801338746 -48.0 26

-22.1-14.5

45_05_7 4.98 0.072 0.01457° 21'10.25 3.00 56444

-33.00.412-33.5 0.6 0.09

45_05_8 5.11 0.067 0.01357° 18' 10.2 2.001067819

-50.8 0.6 0.29

45_05_9 5.02 0.069 0.01457° 07' 10.1 3.00 38131

-32.4 0.6 0.09

45_03_1 3.08 0.071 0.02357° 32' 3.1 1.75 32770

-55.00.199-36.4 0.6 0.40

45_03_2 2.98 0.068 0.02357° 43' 3.2 1.60 54433

0.0 0.466 -6.0 0.6 0.17

45_03_3 2.94 0.082 0.02857° 19' 2.9 1.20 349054

-47.8 0.6 0.31

45_03_4 2.98 0.091 0.03157° 22' 2.95 1.001212560 5 3 03 55 18 1

2 602101 -20.10.205-40.4

-44.9 0.6 0.29 03

-15.8 -9.9

45_03_6 2.94 0.076 0.02657° 13' 3.05 1.001535352 -41.60.160-49.9 0.6 0.25

45_03_7 3.05 0.077 0.02557° 19' 3.05 1.101115318 -40.30.331-47.9 0.6 0.24

45_03_8 3.08 0.073 0.02457° 38' 2.95 1.10 551675

-45.6 0.6 0.29

Table 1. Specimen dimensions, fatigue test results and measured crack paths.

DISCUSSION

As it can be observed from Table 1, crack paths were, in general, characterised by θA

and θB values less the zero. Only the 90_03 configuration (and the specimen named

90_05_08) showed crack paths having θA,B>0. Unfortunately, it was not possible to

explain this anomalous behaviour just by using the stress field distributions.

In order to classify the crack growth modes, we assumed that propagation could be

considered as modeI governed when the ratio ρ=τrθ/σθ was less than 0.2.

Appart form the 90_03 configuration, cracks always tended to propagate along

directions which were able to minimise the shear stress contribution. In fact, as showed

by Table 1, when r>L/2, the ρ ratio was, in general, less than about 0.2 and it decreased

as r increased (because normal stresses tended to be greater than zero, whereas shear

stresses approached a value very close to zero). Therefore, experimental evidences

suggest that, when r>L/2, the shear stress influence played a secondary role. At this

point, it can be formed the hypothesis that the material was capable of knowing a priori

the direction characterised, after a certain distance from the notch tip depending on the

material mecanichal properties, by a prevailing mode I fatigue damage. Aninteresting

consequence of this situation is that crack propagation directions seem to be more

influenced by the stress field distribution outside the Neuber’s structural volume than by

that happens very close to the notch tip. Moreover, comparing the ρ value and the

number of cycles to failure, it can be observed that, in general, the ρ ratio increased as