PSI - Issue 38

S. Häusler et al. / Procedia Structural Integrity 38 (2022) 230–237

233

4

S. Häusler et al. / Structural Integrity Procedia 00 (2021) 000 – 000

whole examination area. Therefore, the degradation of the material properties due to cracks and delamination within the area of interest are measured in an average sense. To obtain the shear stiffness degradation, the longitudinal and lateral strains are averaged before calculating the shear deformations. To keep the autogenous heating at the surface bellow 10°C, as suggested in ISO (2003), the frequencies were controlled. Moreover, within each measuring campaign the strain rate was kept comparable, as recommended in Wedel-Heinen et al. (2006). For this, the strain rates of the first load cycles were used for comparison, since the strain rate in the ongoing test depends on the individual damage state. This resulted from high to low stress amplitudes in frequency ranges from 2.2 to 10 Hz at R = -1 and from 5.5 to 14 Hz at R = 0.1. However, if overheating was detected in the examination area of the initial tests, the frequencies were adjusted accordingly for subsequent fatigue tests. 3. Results and discussion Besides the determination of the S-N curves, the mater ial’s stiffness degradation is analysed. Therefore, the peak and valley stresses and strains of certain load cycles are needed. The stiffness degradation is considered with the cyclic modulus, here defined as the linear gradient between the respective stress-strain pairs under cyclic loading. 3.1. Fatigue testing and S-N curves The measurements are presented in S-N curves, see Figure 2. As a result, from the coefficients of determination it can be seen that both shear sets as well as the two sets with unidirectional specimens show a similar scattering. For the three runouts, it is noticeable that the two tests with unidirectional specimen do not show any visible degradation in the backlight images and thereby it is assumed that no macroscopic cracks are initiated. In contrast, the shear test shows progressed damages in the examination area. The dependence of the onset of degradation caused by crack initiation for the tests with specimen of Θ = 90° was also observed in the other tests and will be discussed later.

Figure 2: The S-N curves for the specimen with Θ = 90° are shown in the left column (a, c), the ones with Θ = ±45° are in the right (b, d). The upper graphs show the results for R = 0.1 and the lower for R = -1. The formula to calculate the semi-logarithmic approximation is given in each graph. The blue marked data were excluded from the stiffness degradation analysis and the orange marked data have been locally overheated.