PSI - Issue 7

Y. Nadot et al. / Procedia Structural Integrity 7 (2017) 530–535

533

4

NADOT Yves / Structural Integrity Procedia 00 (2017) 000–000

Fig. 2. SN curves for different defect sizes (a) fine microstructure (b) coarse microstructure.

To determine the influence of defect size in relation to grain size upon fatigue limit, it is necessary to experiment a wide range of surface defect size introduced in the two microstructures. The Electric Discharge Machining (EDM) technique is used because it enables to introduce reproducible calibrated defects. The EDM process was calibrated to manufacture surface defects with the same hemispherical morphology: the defect depth equals half the surface diameter of the defect as illustrated in Table 1. Thereby, all defects present the same stress concentration coefficient K t = 2:05 (on defect surface and in the plane perpendicular to the loading direction), only the defect size changes from one specimen to the other. This defect size is defined by the surface diameter of the defect, noted D. The EDM process is acceptable to reproduce a calibrated surface defect as shown Table 1. In addition, this process does not affect the grain size at the interface with defect in terms of grain morphology and size. 3. Results In order to carry out the study of notch sensitivity, several S–N diagrams were obtained by varying the defect-size to grain-size ratio over a wide range. Endurance tests were performed for the two microstructure sizes and for various defect diameters: respectively five and four defect diameters for fine and coarse microstructures. For fine microstructure and a defect size given, the defect has more influence on the reduction of maximum stress for high lifetime (superior to two million cycles) than for small lifetime. For example, in a case of D g = 30 µ m, for a defect diameter fixed D = 310 µ m and for lifetime inferior to 200,000 cycles, the maximum stress has a reduction of 14%, whereas for lifetime greater than two million cycles, the maximum stress exhibits a reduction of 25%. For fine microstructure, the percentage of maximum stress reduction doubles from the small to high lifetime. Indeed, for small lifetime, the propagation governs lifetime, whereas for high lifetime, the initiation seems to govern lifetime. For coarse microstructure and a defect size given, the reduction of maximum stress seems to have the same tendency. For example, in the case of D g = 340 µ m, for a defect diameter fixed D = 640 µ m and for lifetime inferior to one million cycles, the maximum stress exhibits a 4% reduction; whereas for lifetime superior to two million cycles, the maximum stress has a reduction of 9%. For fine microstructure and for high lifetime (greater than two million cycles), the presence of defect reduces maximum stress and this reduction increases when defect diameter grows. In a case of D g = 30 µ m, for a defect diameter D = 130 µ m, there is a 6% reduction in maximum stress; whereas for a defect diameter D = 990 µ m there is a 32% reduction. As expected, an increase in the defect size results in a marked decrease in fatigue resistance. For coarse microstructure and for high lifetime, the influence of defect on maximum stress seems to have the same tendency.