Issue 35

A. Nikitin et alii, Frattura ed Integrità Strutturale, 35 (2016) 213-222; DOI: 10.3221/IGF-ESIS.35.25 221 growth in Mode II at the very first stage, but a crack growth bifurcation mechanism seems different for HCF and VHCF. In the case of VHCF branching can be observed in the bulk of material, while in HCF it appears at the surface [18] Such difference in the branching mechanism may have the same nature with one more interesting feature of VHCF torsion behavior. That is an internal crack initiation under VHCF torsion loading in spite of maximum shear stress located at the surface. An analysis of internal crack initiation site in VT3-1 titanium alloy did not show any structural flaw in the microstructure so that inclusions or clusters of alpha-platelets, Fig. 10. The crack initiation site is significantly destroyed by friction that is clearly seen in Fig.10b. A ‘step’ at the fracture surface and followed crack growth in two parallel planes means that internal crack is also initiated in the plane of maximum shear stress. After reaching a certain length (that is shorter than the distance from initiation site to surface) the crack branches on the plane of maximum normal stress. Sometimes there is a sort of competition between two maximum normal stress planes that produce a ‘wing’ like structure at the surface. Excluding this ‘wing’-like structures, a further branching is limited in the case of internal crack initiation and no branching ‘threshold’ (like in case of surface crack) can be found at the fracture surface. In contrary, another notable zone or critical crack length can be reported. As shown in Fig.7b the crack front shape is changing as soon as it arrives at the surface, probably because the stress intensity factor is increasing and consequently the crack growth rate too. (a) (b) Figure 10 : Fracture surfaces of specimens (a) S=; Nf= (b) S=; Nf=; with clear border of fatigue crack branching. Finally, for forged titanium alloy loaded under push-pull fatigue crack initiations were mainly associated with macro-zones and smooth facets cracking [16], while in the case of torsion loading these features were not found. C ONCLUSIONS ased on the results of fatigue tests under pure ultrasonic torsion the next conclusions can be drawn up: (1) SN curves of VT3-1 titanium alloy under torsion R=-1 loading have a more significant slope, compared to results under fully revered tension. However, the relation between axial and torsion fatigue strength obtained for HCF seems to be applicable for VHCF data. (2) Independently on the production process (forging and extrusion) two different crack initiation sites were found: surface and subsurface crack. The subsurface crack initiations were observed for specimens failed at longer fatigue life. Transition from surface to subsurface crack initiation was found well after the same transition for push-pull loading: about 10 8 cycles for torsion and 10 6 cycles for axial loading. (3) Qualitatively, a surface cracking under torsion loading in VHCF is similar to HCF results, except an additional possibility for internal crack to branch in VHCF regime. (4) A sort of branching ‘threshold’ or critical crack length can be found at the fracture surface, beyond which a torsion crack shows several branches in another 45° plane of maximum normal stress. (5) In the case of internal crack, an initiation and early crack growth is also being in Mode II. Bifurcation to a normal stress plane growth happens before the crack reaches the specimen’s surface. Further crack growth is processed along the specimen surface. At later stage of subsurface growing a crack turns to the surface. The moment when subsurface B