PSI - Issue 7

A. Giertler et al. / Procedia Structural Integrity 7 (2017) 321–326

324

4

A. Giertler et Al./ Structural Integrity Procedia 00 (2017) 000–000

could be divided into an athermal σ G and a thermal component σ*, Seeger 1954. The athermal fraction σ G is dependent on the Peierls stress and describes the stress proportion which is necessary for moving dislocations to bow out and to overcome the elastic interaction with other dislocations. The thermal component is a supplement stress, necessary to allow screw dislocations to move at a given temperature and strain rate. According to an analysis by Seeger, this proportion is negligibly small above a transition temperature T 0 or below a transition rate ε̇ T . Fig. 1b) shows the S-N data for the 57HRC hardness condition. Due to the lower tempering temperature and the resulting higher hardness as well as higher strength of the material, the fatigue life shifts as expected to higher values of the stress amplitude σ a . The increase in hardness has led to a change in the failure mechanism from pure surface failure to internal crack initiation at nonmetallic inclusions. The data points shown in Fig. 1b correspond only to internal crack initiation on non-metallic inclusions of the type Al 2 O 3 . The results for the test frequency of f =20,000Hz show a shift to higher values for the fatigue life. However, the reason for this could not be the influence of the frequency described above. The increased strength of the martensitic microstructure for the 57HRC hardness condition leads to a pronounced increase in the at hermal fraction σ G by an increase in the Peierls stress. As a result, the thermally activated stress component σ* shifts to lower temperatures and therefore to higher values for the strain rate ε̇ . Here, the small deviation of the fatigue life for the two test frequencies can be attributed to a statistical influence. The change of the test frequency from f =20,000 Hz to f =95 Hz required a change of the test machine and associated with this was a change in the specimen geometry. The critical volume in the gauge length of the sample was V 20kHz =125mm 3 for the test frequency of f =20,000Hz and V 95Hz =402mm 3 for the test frequency of f =95Hz. As a result, due to the increase of the critical volume by a factor of 3.2 the presence of larger nonmetallic inclusions in the area of the gauge length is promoted, which leads to the premature failure of the material for the test series of f =95Hz. The failure mechanisms in the VHCF fatigue regime are strongly dependent on the nature of the material, as well as on its condition. Therefore, a distinction is made between specific failure mechanisms in this fatigue regime. Mughrabi proposes a differentiation between type I and II materials, Mughrabi 2006. Accordingly, Type I materials are pure single-phase ductile metallic materials without internal extrinsic defects. Type II materials, on the other hand, have such defects as non-metallic inclusions or pores, Mughrabi 2002. In the following, a distinction is made between type I and type II behavior, based on the behavior with regard to the damage mechanisms observed in the fatigue test for the material conditions of 37HRC and 57HRC. 3.1. Type-I Fatigue behavior Detailed studies using high-resolution scanning electron microscopy (SEM) on the fatigued specimens for the 37HRC condition reveal a characteristic damage pattern for type I materials. Within the martensitic microstructure, regions with local plastic deformation can be identified at the surface. A detailed examination of the run-out specimens showed, that even on specimens with more than 10 8 number of cycles, local damage occurred within the microstructure due to the formation of slip bands, Fig. 2a.

a)

b)

Fig. 2. a) Slip band formation on the surface of a 37HRC fatigue specimen after N =2·10 8 number of cycles loaded with a constant stress amplitude of σ a =490MPa and b) FIB cross section through a slip band showing a microcrack.