PSI - Issue 40

A.V. Zinin et al. / Procedia Structural Integrity 40 (2022) 470–476

473

4

A.V. Zinin at al. / Structural Integrity Procedia 00 (2022) 000 – 000

Fig. 1 also shows the dependence of the ratio of plastic p  and total s  strain in the form of the material quality factor of the material ( Q - factor ) p s     depends on the amount of accumulated strain in the form of 1 2 s  . The change in the Q -factor  characterizes the intensity of the decrease of the plastic component p  during cyclic strain. When evaluating the effect of low-cycle overloads on the failure mechanisms and lifetime of structural materials, the staging model predicts the change in the level of material damage due to overloading depending on the stage of the plastic strain process at which the overloading occurs, i.e., the loading history. 3.2. Microstructural analysis Based on physical representations of the destructive nature of the strain, microstructural features of fatigue damage development in the presence of low-cycle overloads were investigated. Fig. 2a shows radiographs of steel in the initial state, and Fig. 2,b-d – after various modes of two-stage fatigue loading, including low-cycle overloads. In the image (Fig. 2b) of the cross-section of the sample obtained after monotonic high-cycle loading before fracture according, the diffraction line (light in the figure) is clearly marked and consists of spots that are slightly blurred compared to the image of the sample in the initial state (Fig. 2a). The radiograph (Fig. 2c) of samples cyclically deformed with a small strain amplitude a  = 0.2% followed by multicycle loading does not differ significantly from the fatigue loading pattern; the diffraction line retains spotting. An increase in the strain amplitude to a  = 0.5% significantly changes the structural state of the material. After five overload cycles, diffraction reflexes, blurring, form an almost continuous line, and when the number of overload cycles increases to 80, a completely blurred, continuous line is observed on the radiograph (Fig. 2d). Such a radiograph corresponds to the structural state of the material at the second stage of elastoplastic strain. It can be assumed that fatigue loading causes softening of such material due to an excess of free dislocations and vacancy type defects and the possible appearance of submicrocracks (Makhutov 2019, Rybakova 1980).

Fig. 2 X-ray diffraction patterns of specimens of steel 10G2S1: a) initial condition; b) high-cycle fatigue loading, σ a = 315 MPa, N p = 1.2x10 6 ; c) low-cycle loading, ε a = 0.2%, N low = 80; d) low-cycle loading, ε a = 0.5%, N low = 80; e) multicycle fatigue loading, σ a = 315 MPa, N p = 9 x10 5 after preliminary low-cycle overloads, ε a = 0.5%, N low = 80. Figure 2d shows a radiograph of the sample after 80 cycles of preloading at the "training" stage with an amplitude a  = 0.5% and subsequent fatigue loading until fracture at p N = 9×10 5 , on which there is a tendency to spot the line, the greater, the higher the degree of preliminary strain. Figures. 3 and 4 show the results of a fractographic study of the fatigue mechanism of 10G2S1 steel samples after preliminary low-cycle loading. Fig. 3a shows the general view of the fatigue fracture of steel after monotonic high-cycle loading before fracture. It follows that the fracture surface has no macroscopic plastic strain. Under higher magnification flat areas of local brittle fracture with fatigue grooves – microscopic (~1 μm) "plateaus" with characteristic microporosity – are visible. Another type has a fracture of the sample subjected to joint low-cycle ( a  = 0.5%, low N = 80) and high cycle loading ( a  = 315 MPa, p N = 9 x10 5 ) (Fig. 3, b). The fracture has radial seams or signs of "slate" (Rybakova 1980), which indicates increased fragility. The fatigue area and the rupture area, which is located close to the sample center, are clearly visible, which is characteristic of low-cycle fatigue. The features of the sample fracture that has undergone preliminary training ( low N = 5 cycles with an amplitude of a  = 0.5%) (Fig. 4) are presented as