Figure 1 shows a schematic of the tested plate. The specimens were pre-loaded by 50 kN in order to ensure that they are in their initial state. However, to understand the fracture mechanisms, it is crucial to examine their response in a different loading regime, therefore, one- and two-way cyclic loading tests were carried out. The latter is important to investigate the impact of the loading direction on the crack progression, as well as the effective displacement-control imposed on the crack surfaces. The pre-loaded plates are then subjected to loading under one- or two-way cyclic loading schemes, which involves a displacement-control regime of 20 m or 5 mm, respectively. The values are sufficiently high to allow for visible crack growth and to avoid plastic yielding, yet low enough to ensure that the damage accumulation process is not influenced by the plastic properties of the composite.
The final fracture resistance of the plates is plotted in figure 3 for the different cyclic loading conditions. The main conclusion that can be drawn from this figure is that the loading direction has a significant influence on the crack growth. Indeed, the average resistance of the pre-cracked specimens is about 30% higher than the average resistance of the virgin specimens, which indicates that the cracks are much slower to progress in the negative (virgin) loading direction. This is also observable in the scatter plot of the resistance of the pre-cracked specimens compared to the resistance of the virgin specimens in figure 4. The latter shows that indeed, the progression of the cracks is much slower in the negative loading direction.
The pre-cracking brought the test specimen into the non-linear domain. As the crack width was limited to ~1 mm, the tests were carried out at the compressive strength (effective stress). The displacement-controlled test was performed at a rate of 20 mm/min with a displacement of 10 mm. The two-way cyclic loading tests were performed at a rate of 10 mm/min with a displacement of 50 mm. In both cases, the rate of loading was selected to be significantly slower than the displacement-controlled test. The test was continued for more than 100 cycles in order to observe the material behaviour and extent of damage under a realistic seismic excitation. 827ec27edc