PSI - Issue 68

Lukas Dominik Geisel et al. / Procedia Structural Integrity 68 (2025) 1273–1279 L. D. Geisel and S. Marzi / Structural Integrity Procedia 00 (2024) 000–000

1276

4

Fig. 2: Labeled photography of the 4-OSLB setup during one of the conducted tests.

The experiments were carried out with specimens that were bonded with the elastic-plastic structural adhesive SikaPower ® 498 . This adhesive was chosen since a lot of previous works can be found in the literature (see Lad wig et al. (2022); Loh and Marzi (2018b,a, 2019); Schrader and Marzi (2022, 2023); Stamoulis et al. (2014, 2016); Kididane et al. (2024)), providing fracture mechanics data for comparisons. The adherends were manufactured from the high-strength aluminium alloy EN AW-7075 T6. The long and short adherends have lengths of 1000 mm and 640 mm respectively. The seesaw span is l s = 300 mm. The second moment of area I of the adherent cross-section is (19 . 6 ± 0 . 5) × 10 3 mm 4 . The flexural modulus E of these adherends was measured to be (72 ± 1) GPa. The thickness t adh of the adhesive layer was set to a nominal value of 0.3 mm. Further details regarding the experimental setup can be found in Geisel and Marzi (2024). A typical backface strain measurement within the interval x ∈ [ l , l + l s ] of one specimen at particular piston dis placement is shown in Fig. 3. The fracture process zone, along which the traction-separation laws are calculated, is highlighted by the gray background. For more details regarding the detection of this process zone, see Geisel and Marzi (2024). Using Eqs. (6) and (8), values of w COD and J OFS can be calculated from the shown backface strain measurements. However, the only measurements that have some physical meaning with respect to cohesive laws are obtained within the highlighted fracture process zone. Furthermore, if only measurements are considered which were taken after the start of crack propagation, the whole range of possible w COD -values (i.e., zero up to a maximum COD at failure) as well as their respective J OFS -values is contained within the obtained results. This allows the calculation of a single traction-separation law for every single backface strain measurement. Such a traction-separation law is depicted in Fig. 4. Since each strain measurement yields one traction-separation law, each specimen can in turn potentially yield hundreds of cohesive laws as shown in Fig. 5 for one representative specimen. A mean cohesive law was calculated from the results of all examined specimens. This TSL is depicted in Fig. 6 together with previously obtained cohesive laws using di ff erent specimens (see Schrader and Marzi (2023); Loh and Marzi (2019)) as well as the law obtained with the same specimen using force and angle measurements in conjunction with extensometer measurements at the crack tip (see Geisel and Marzi (2024)). The results of this study show some consistency with these previous results. However, one di ff erence can be seen in the latter half of the cohesive laws. 4. Experimental results