PSI - Issue 13

Kohei Kishida et al. / Procedia Structural Integrity 13 (2018) 1032–1036 Author name / Structural Integrity Procedia 00 (2018) 000 – 000

1035

4

Notably, the difference between the fatigue limits at 293 and 433 K in the Fe-0.016C-1Si alloy is 40 MPa, which is smaller than the difference of 65 MPa in the Fe- 0.017C alloy. Hence, the robustness of the fatigue limit against temperature was improved by the addition of Si. It is speculated that Si delays carbide formation, and hence, the specimen is significantly influenced by DSA. However, the decrease in fatigue limit upon increasing the temperature from 293 to 433 K indicates that the addition of 1wt% Si cannot suppress the carbide formation at 433 K. In addition, we propose three reasons why the crack did not stop propagating in the Fe-0.016C-1Si alloy: A) Solute hardenability degrades in a time-dependent manner owing to carbide precipitation during the tests. Accordingly, the strength at a crack tip decreases with the number of cycles, causing further propagation. B) A new crack forms in a weak microstructure and subsequently coalescences with a pre-existing crack (Koyama et al., 2017a). As the crack length increases via coalescence, raisedd mechanical driving force enables further propagation. C) The addition of Si may enhance dislocation planarity (Ushioda et al., 2009), assisting shear-type fatigue crack growth. The shear-type crack propagation is considered to be less influenced by crack-closure-induced compressive stress. The identification of the underlying mechanism of this issue requires further investigation, and therefore, we will examine the small crack propagation behavior as a future work.

Fig. 4. Replica images taken at (a) 0, (b) 1.0 × 10 7 , (c) 2.0 × 10 7 , (d) 2.1 × 10 7 , (e) 2.4 × 10 7 cycs in the Fe-0.016C-1.0Si alloy at 200 MPa at 293 K.

3.3 Temperature and aging time dependence of hardness and carbide precipitation Figure 5 shows the variation of hardness with aging time. The hardness at 293 K increased steadily with aging time. In particular, the hardness increased proportionally to 2/3 power of aging time in the early stage, indicating that the increase in hardness is attributed to the segregation of solute atom into the pre-existing dislocations, i.e., strain aging (Abbaschian and Reed – Hill, 2008). The slow increase in hardness in the late stage is also consistent with the conventionally known trend of strain age hardening (Harper, 1951). Accordingly, the serrated flow at the strain rate of 10 -5 s -1 is concluded to arise from DSA. In contrast, the variation of hardness at 433 K did not follow this trend and even decreased after aging for 1 day. Notably, the number of carbide particles increased after aging at 433 K for 92 h.

Fig. 5. Vickers hardness plotted against 2/3 power of aging time in the Fe-0.016C-1.0Si alloy