13th International Conference on Fracture June 16–21, 2013, Beijing, China -8- Fig.9. Evaluation of yield stress variation on S-N properties Such a diagram gives us two very interesting points, which have to be discussed. First, if we consider only results from tests performed at 0.2, 2, 20 and 140 Hz, one can see that fatigue strength of S15C steel gathers in a thin area. Besides, fatigue limits from 2 and 140 Hz fatigue tests become really close each other. As a consequence, such a diagram is significant in order to evaluate the yield stress influence on the S-N property of S15C steel. The second main point is of course the case of fatigue results from ultrasonic tests. Due to relatively high difference of yield stress introduced before between ultrasonic tests conditions and usual test frequency conditions, the huge gap of fatigue strength seen in Fig. 3 is significantly lower by taking into consideration the effect of strain rate. Nevertheless, it is not enough to claim that S-N properties are similar with other results from usual range frequency. So, the strain rate effect on yield stress is not sufficient to give an entire explanation of fatigue strength gap found in Fig. 3. 4.2. Cyclic hardening / softening behavior In Figs. 6 and 7, a difference of stress-strain hysteresis phenomenon can be highlighted between 0.2, 2, 20 and 140 Hz fatigue tests. In other words, cyclic softening and hardening behavior seems to be slightly influenced by fatigue test frequency in this range. Of course, it is impossible to undertake similar stress-strain hysteresis study in the case of ultrasonic testing method, as real stress applied on specimen in unknown under ultrasonic technique. So, we cannot assess in the same way a potential discrepancy between ultrasonic and fatigue tests conducted in usual frequency range. In order to highlight anyway a possible frequency effect on the cyclic hardening and softening, misorientation observation by EBSD method has been performed. This comparison involves 20 kHz and 20 Hz fatigue tests. The samples used in this study have been prepared following hereafter instructions. For each frequency, two particular stress levels have been chosen, which are the fatigue limit plus 10 and 30 MPa. These stress levels are called Sw+10 and Sw+30 in the rest of the present paper. For each stress level, corresponding fatigue strength is given by S-N model shown in Fig. 3. Three different specimens were stopped at 5%, 10% and 25% of the calculated fatigue strength. So, 12 different specimens have been fatigued in this process considering both 20 kHz and 20 Hz fatigue tests. After fatigue test has been stopped, center section of the specimen has been extracted, and then polished. Sample final condition has been prepared by OP-AA polishing method in order to
RkJQdWJsaXNoZXIy MjM0NDE=