PSI - Issue 13

Tsubasa Kumamoto et al. / Procedia Structural Integrity 13 (2018) 710–715 Author name / Structural Integrity Procedia 00 (2018) 000 – 000

714

5

Fig. 3. SEM images of the uncharged specimen deformed at a strain rate of 10 − 4 s − 1 in (a) the damage arrest regime and (b) the damage growth regime; the H-charged specimen deformed at strain rates of (c, d) 10 − 4 s − 1 and (e, f) 10 − 2 s − 1 in (c) the damage arrest regime, (d) the damage growth regime, (e) the early damage arrest regime, and (f) the late damage arrest regime. F and M indicated ferrite and martensite, respectively. Observations were made in the TD cross-sections at mid-width locations in the specimens. Fracture surface morphology was observed to support the discussion on the micro-mechanisms of the damage evolution. Fig. 4 shows the detailed fractographic feature at the center of the fracture surfaces. The fracture surfaces of all specimens show dimples. In the specimen fractured at a strain rate of 10 − 4 s − 1 with hydrogen pre-charging, a flat surface with a scale of grain-size was rarely observed (Fig. 4b ′ ); however, the fracture surface was dominantly covered with dimples. The flat fracture surface implies the damage nucleation associated with the decohesion of the martensite region. The appearance of dimples indicates that the coalescence of voids and cracks contributed to the final fracture. These facts indicate that the damage evolution after its initiation stage is dominated by ductile mechanisms even at the strain rate of 10 − 4 s − 1 with hydrogen pre-charging where a significant reduction in elongation appeared. More specifically, according to the damage analysis results shown in Fig. 3, the effects of hydrogen on the plasticity-driven damage growth in ferrite need to be analyzed to understand why there is a significant difference in the elongation between the strain rates of 10 − 4 and 10 − 2 s − 1 with hydrogen pre-charging. In this context, we consider two factors causing the strain rate sensitivity of the hydrogen-assisted damage growth. The first factor is the increase in the damage number density at a given strain. When hydrogen was introduced, the damage number density at 10 − 4 s − 1 was higher than that at 10 − 2 s − 1 at an identical local strain, as shown in Fig. 2a, which indicates the average distance between respective voids at a given strain decreased with decreasing strain rate. Therefore, damage coalescence can occur early, at the lower strain rate, which causes a reduction in elongation. The second factor is the damage arrestability of ferrite. Lowering the strain rate with hydrogen pre-charging shortened the damage arrest regime, as shown in Fig. 2c, and the damage arrestability was predominantly affected by the plasticity of ferrite. These facts indicate that the damage arrestability of ferrite was reduced by lowering the strain rate with hydrogen pre-charging. More specifically, the plasticity-driven deterioration of the damage arrestability of ferrite can be caused by a reduction in the critical strain for damage coalescence or propagation owing to the localized plasticity at the damage tip. In the context of the arrestability of ferrite, hydrogen diffusion and associated hydrogen localization in the ferrite matrix at a damage tip significantly affect the damage growth, resulting in the strain rate sensitivity. Hence, the damage arrestability would decrease with decreasing strain rate, which explains the disappearance of the negative hydrogen effect on elongation when the strain rate was increased from 10 − 4 to 10 − 2 s − 1 with hydrogen pre charging. In order to clarify the underlying mechanism of the strain rate sensitivity of the hydrogen effect, we will perform further detailed microstructure characterization near the damages and more statistically-reliable damage quantifications in future.