13th International Conference on Fracture June 16–21, 2013, Beijing, China -6- and microscopic inspection of the failure sites revealing that the practical totality of the fatigue failures started from the specimen surface. In the exceptional case that a failure could be identified to have its origin in a volume defect, it was not considered in the present analysis. a) Rod d=25 mm (used for d3 and d8 specimens) b) Rod d=45 mm (used for d22 specimens) Figure 3. Microstructure of AlMgSi1 4. Results and model application The experimental data sets for each specimen geometry were individually fitted to the model given by Eq. (6) of section 2.2. to obtain the parameters referred to ΔS=9 mm2 as shown in table 3. The choice of ΔS is free, so that larger values of ΔS will only result in smaller values of δ, remaining the other parameters unchanged. Table 3. Parameter estimates for each data set dmin [mm] ΔS [mm2] B exp(B) [cycles] C exp(C) [MPa] λ β δ 3 9 11.57 105873 5.40 221 0.01 2.42 0.16 8 9 9.95 20952 5.27 194 0.21 4.31 1.65 22 9 10.63 41357 5.27 194 0.00 3.41 4.09 The fatigue test data and their corresponding Wöhler fields are represented in Fig. 4 for the specimen geometries d3, d8, and d22. The percentiles are computed replacing into Eq. (6) the parameter estimates referred to the area ΔS given in table 3 and the specimen geometries of table 1. In a second step, the SN fields for the d3 and d22 specimens are predicted based on the parameter estimates, found by fitting the data of another specimen geometry, by substituting the corresponding radii and lengths in Eq. (2). Figs. 5a and b show the Wöhler fields for the d3 and d22 specimens using the parameter estimates obtained by fitting the d8 data. The extrapolations from the d3 to the d22 specimen and from the d22 to the d3 specimen are given in Figs. 5c and d, respectively.
RkJQdWJsaXNoZXIy MjM0NDE=