PSI - Issue 5

A. Eremin et al. / Procedia Structural Integrity 5 (2017) 889–895

894

A. Eremin et al/ Structural Integrity Procedia 00 (2017) 000 – 000

6

(a)

(b)

Fig. 5. Crack-tip hysteresis loops for test 2 (with the single overload).

(b)

(a)

Fig. 6. Graphs of the COD (ε max ) maximum value at the various numbers of cycles after the single overload: measurements (a) at the crack tip, (b) at 500 μm behind the tip.

The revealed difference of ε max values in contrast with the reference measurements at the crack tip makes approximately 4 %. In doing so, this value is suggested to come from the measuring technique error (under loading parameters used), while the variation caused by the overload possesses a peak exceeding by 16 % (for the 0 cycle) and by 30 % (for the 100 cycles) that of the reference values. Crack propagation distance measured over digital images (at microscope gained magnification) after 100 cycles of loading does not exceed 2-4 microns. The value of ε max reaches the maximum at 100 cycles while after, it decreases by 8 % with further nearly constant value during next 2 000 cycles. However, at the end of this experiment it did not reach the reference level. The total crack length after 2 000 cycles was 90-100 μm. At the same time after 100 cycles the crack opening displacement measured at the distance of 500 μm behind the tip possess a descending trend. Once again it does not reach its “basic” value. It means that the single overload still affects the crack growth mechanisms even after its extension in plastic zone at the distance of 100 μm (14 % of the monotonic plastic zone). The maximum COD value was registered after 100 cycles when crack has propagated in the overload plastic zone by a few microns. This process should be accompanied by crack blunting and increase of near-tip residual compressive stresses which reach the maximum after propagating by a few microns into the plastic region. The utilized technique for data acquisition and processing, including analysis of crack opening displacement via DIC, makes it possible to estimate the single overload/underload cycles impact on the hysteresis loops shapes and COD parameter (ε max ). The obtained results allow concluding that the ε max parameter is significantly affected by the single underload which results in its sharp increase. The latter is related to the not closed loop induced by residual tensile strains due to the underload. However, these strains are relaxed during next five “basic” loading cycles, then, however, the ascending trend in its behavior is observed that may be attributed to the in crease of the Δ K eff . 4. Summary and conclusion