PSI - Issue 38

7

S. Häusler et al. / Procedia Structural Integrity 38 (2022) 230–237 S. Häusler et al. / Structu al Integrity Procedia 00 (2021) 00 – 000

236

lines (6k and 19k cycles) are caused by cracks in multiple cross sections, which can be described as a complete degradation of the measurement area. The others showed cracks running through the width at only one or two locations. For the orange line ’ s specimen (215k cycles), the crack was slightly under the bottom side speckle pattern and thus not within the DIC measurement area. This indicates that the material does not undergo any significant degradation besides the one directly caused by the cracks. Compared to the representative line in Cadavid et al. (2017) the black and lilac lines show similar stiffness degradation magnitudes close to final failure. Apart from this, only the lilac line has a comparable progress while the other two differ due to delayed degradation onset. In contrast to the tests above, the shear tests show only little variations in the degradation trend. Furthermore, the graph in Figure 5 (a) shows for R = -1 no evidence of a stress amplitude dependency. In general, the Θ = ±45° specimens were visibly degrading all-over within the examination area with multiple cracks parallel to the fibre direction. Figure 4 (b) shows a backlight image with a representative damage state close to final failure. Compared to this image, the blue line ’ s specimen (151k cycles) in Figure 5 (b) showed a more broadly distributed damage over the whole examination area, which can explain the high loss in stiffness. On the contrary, the orange line ’ s specimen (984k cycles) showed only small degradations around the dominant crack crossing the whole width in a 45° angle. The remaining lines between the blue line (151k cycles) and the orange line (984k cycles) show comparable scatter as the lines in the tension-compression ( R = -1) tests. However, the stiffness in the tension-tension tests ( R = 0.1) undergoes the final drop at around 5-10% less degradation. It should be noted that a few lines in Figure 5 (b) had to be cut at the end when surface cracks crossed the lateral speckle pattern and thereby disrupted the lateral strain measurement to an unusable state. The initial stiffnesses for normalisation in Figure 5 are presented in Table 3: Table 3a: Initial stiffness used to normalise the curves of the experiments shown in Figure 5 (a) and the respective stress amplitude R = -1 Curve 14.1k 14.3k 23.7k 65.4k 112k 134k 953k E init GPa 4.2 4.2 4.2 4.5 4.6 4.4 5.0 MPa 20.5 20.3 18.4 16.3 16.0 15.8 13.0 Table 3b: Initial stiffness used to normalise the curves of the experiments shown in Figure 5 (b) and the respective stress amplitude R = 0.1 Curve 11.7k 19.2k 151k 155k 202k 427k 662k 987k 1.25M E init GPa 4.5 4.3 4.7 4.7 4.6 4.6 4.9 5.2 5.2 MPa 12.8 12.3 10.6 10.0 9.9 8.9 8.8 8.5 8.1 Figure 5: Relative stiffness degradation of the Θ = ±45° specimen tests a) at an R -value of -1 and b) at R = 0.1. The number of failure cycles as well as the stress amplitude are given in the legends.