Issue 21

D. Croccolo et alii, Frattura ed Integrità Strutturale, 21 (2012) 13-20; DOI: 10.3221/IGF-ESIS.21.02 16 I06 Point # 1 Point # 2 Point # 3 Point # 4 Point # 5 Point # 6 Point # 7 External load (kN) 46.8 44.1 41.4 38.7 36 33.3 30.6 Remote maximum stress (MPa) 650 612.5 575 537.5 500 462.5 425 Cycles 11,710 98,239 136,289 232,356 684,844 692,557 1,347,271 Table 2 : Stress results for the I06 specimens. I2 Point # 1 Point # 2 Point # 3 Point # 4 Point # 5 Point # 6 Point # 7 Point # 8 External load (kN) 52.2 49.5 46.8 44.1 41.4 38.7 36 30.6 Remote maximum stress (MPa) 725 687.5 650 612.5 575 537.5 500 425 Cycles 35,260 70,017 112,800 136,490 193,510 300,535 430,877 2,000,000 (run out) Table 3 : Stress results for the I2 specimens. Figure 4 : S-N Diagram. D ISCUSSION y focusing attention on the S-N diagram reported in Fig. 4, it is possible to confirm that the interference fit level has a strong and positive influence on the fatigue strength of the OH specimen, as widely demonstrated in [5, 7]. However, a remarkable difference can be noticed between the two levels, I06 and I2 : the higher is the interference level, the higher is the fatigue strength at the highest levels of remote stress. Conversely, by reducing the remote stress, the fatigue strength and the Endurance Limit is quite the same for different interference levels. This experimental evidence can be explained by advocating the actual amplitude of the local stress field in the vicinity of the hole. The observed cracking behaviour indicates a crack initiation close to the cross section of the holed surface (see Fig. 3), so that, it is clear that the actual amplitude of axial stress, normal to the cross section of the specimen, is the driving force of the crack propagation. The amplitude depends on both the remote stress and the local residual (compressive) stresses due to overcoming of the Yielding point. For this reason the local amplitude, depending on both the remote stress and the interference level, has been calculated and analysed. A number of Finite Element Analyses ( FEA ) have been carried out in order to confirm what stated above by directly following the method proposed by [6, 7, 13]. In Fig. 5 an example of the FEA model (mesh and contour plot of results) is reported. The analyses are conducted on a 2D elastic-plastic plane stress model (6 nodes triangular elements), in which three different steps have been simulated: i) the interference (if present), ii) the application of the maximum remote stress and iii) the unloading with R=0.1. The actual stress amplitude along the cross section of the specimens, which can be related to fatigue life, has been calculated as the difference between the axial stresses of the phases ii) and iii) and reported in Fig. 6a, 6b and 6c for the OH , I06 and I2 specimens respectively: the B