Fatigue Crack Paths 2003

experimental points superpose onto the theoretical elastic asymptotic profile but are well

below the elastic-plastic

was thus less than 4 3 M P abumt it is

profile. ΔKIIeffective

difficult to specify its value (whence the arrow on Fig. 1b).

Figure 2. Sliding displacement profiles for ΔKII=7 (left) and 1 3 . 6 M P a(rmight), R=-1.

Figure 3. a) Sliding displacement profiles for ΔKII=37MPa m , R=0. b) Evolution of the

"plastic sliding displacement" ρII , for ΔKII=20MPa, 2m50 μ mbehind the crack tip.

For ΔKII=29 and 37 M P aanmd R=0 (for which KIImax were higher than for R=-1,

but maximumapplied shear stresses lower than 38%of the 0.2% shear yield stress), the

measured displacement profiles were above that predicted by L E F M(Fig. 3a) but

anyway still below the “elastic-plastic”

profile. The difference between measured and

computed elastic-plastic profiles, which can be attributed to friction on the crack flanks