Along these lines, Lukas [8] as shown that at least in Cu in the high cycle regime the “persistent slip bands” are composed of localized regions of very high vacancy concentration. It seems that a threshold exists, in iron, for the formation of the quasi-PSB. It is found that below 70 MPa no PSB occurs up to 109 cycles. However, the number of PSB increases with the number of cycles, but at present no experimental evidence is available to prove that some PSBs cannot form at 1010 cycles or more. No fish eye occurs in these conditions, probably due to the strong effect of the plane stress field and the very small size of the inclusions in a low yield stress Armco iron. It is of interest to compare iron and martensitic steel, with the same flat specimen (1mm) and the same frequency, in push pull loading at 620 MPa, for a fatigue life beyond 108 cycles. Figure 7 shows that the initiation in martensitic steel starts from a fish eye and not from the surface. Reasons for this location must be addressed. The fish eye occurs in the high strength steel from a small inclusion of oxide, in plane strain conditions, located inside a plastic zone around an inclusion. In this case, the plane stress effect at the surface is not efficient in causing plasticity due to a defect in steel where the UTS is close to 2000 MPa. There is a competition between plane strain plasticity and plane stress plasticity. In this case the micro plasticity is not governed by the Von Mises criteria but by the stress concentration effect. In plane strain plasticity the PSB or quasi PSB are not observed in the fish eye. The mechanism of plasticity seems relevant to polygonisation; grain refining or phase transformation, around the inclusion. These results show that initiation in gigacycle fatigue is explained by a fish eye formation except when the effect of plane stress occurs in thin sheets or in thick bars when the yield point of the metal is low. In this condition the surface governs initiation. Otherwise in high strength alloys the initiation, in the gigacycle range, is sub-surface and always depends on defects: inclusions, pores, super grains. The surface effect is less important except if the residual stresses are important. According to these observations more attention must be paid to the microplasticity (or damage accumulation) inside the fish eye in high strength alloys such as martensitic steels. 5-Instability of microstructure in VHCF Our own results and those from the literature it is observed that the micro plasticity in gigacycle fatigue is more than simple dislocation slip. Sometimes phase transformation, refining of the grain, twinning, and instability of the yield point, occur even at low loads for a very high number of cycles. Several observations are enumerated up below: -In austenitic stainless steels the austenite is not stable in the gigacycle fatigue regime even if the plastic deformation is theoretically very small. There is also a large thermal dissipation [9]. -When the amount of retained austenite approaches 10% in martensitic steels, a large thermal dissipation is observed at the beginning of the test followed by a high temperature rise when the fish eye propagates. [10] -In low carbon steels the mobility of dislocations is affected by interstitial atoms, depending on the strain rate and the grain size. Several types of instability are observed in monotonic loading such as Luder’s bands, Portevin-Le Chatelier bands or Neumann’s bands, twinning, etc. It seems useful to consider similar effects in ultrasonic fatigue. [11]
RkJQdWJsaXNoZXIy MjM0NDE=