PSI - Issue 42

Mike van der Panne et al. / Procedia Structural Integrity 42 (2022) 449–456

455

M. v.d. Panne, J.A. Pascoe / Structural Integrity Procedia 00 (2019) 000–000

7

The presented results also imply that testing multiple di ff erent fibre orientations, and possibly di ff erent widths, is necessary to fully understand delamination growth. For practical applications this is not desirable. One option is to only test a small number of interfaces and then try to interpolate for other interface angles, for example by employing an approach similar to those discussed by Jones et al. (2021). However, the change of crack migration behaviour at di ff erent fibre orientations might prevent this from working. Alternatively, one could conduct planar delamination growth tests, where the fibre orientation with respect to the delamination varies around the circumference of the delamination. Another option would be to only perform tests on an interface that will give ‘worst-case’ crack growth results, i.e. with the least amount of fibre bridging. Of the interfaces presented in this paper, the 0 // 0 interface clearly presented the worst case, especially at long pre-crack lengths. However it is not certain this indeed represents the absolute worst case. For example a 0 // 5 interface might prevent nesting, and thus suppress fibre bridging, which at the same time also blocking the delamination migration behaviours seen in the 0 // 45 and 0 // 90 specimens. Further research is needed to establish what is the worst-case delamination growth scenario. In the present research, fatigue driven delamination growth was investigated using double cantilever beam speci mens with a 0 // 0, a 0 // 45, and a 0 // 90 delaminating interface. The results show that: • The fibre orientation a ff ects the nature and strength of the fibre bridging and crack migration phenomena and consequently the resulting delamination growth rate at a given fatigue load; • The delamination growth rate is more sensitive to the fibre orientation for longer initial crack lengths; possibly this is related to the crack extension needed to develop the fibre bridging zone; • Of the tested interfaces, the 0 // 0 interface produced the fastest crack growth rates, especially for long initial crack lengths. It is clear that further investigation of the e ff ect of fibre orientation on delamination growth under fatigue loading is required. Relying solely on unidirectional coupons to understand and predict the behaviour of full-scale structures is likely not su ffi cient, unless worst-case data is used, which could result in over conservative designs. To prevent the neccesity of testing many di ff erent lay-ups, it would be preferable to develop our understanding of the underlying physics governing fatigue delamination growth. This would then allow us to a predict the delamination growth rate for an arbitrary combination of fibre angles at the interface, based on only limited material data. 6. Conclusions

Data access statement

The data underlying this research will be made publicly available via: https://dataverse.nl/dataverse/ FibreOrientationFatigueDelam . If needed, a copy of the data can also be requested directly from the corre sponding author via e-mail: j.a.pascoe@tudelft.nl

Acknowledgements

The authors express their thanks to Dr. ir. R.C. Alderliesten (TU Delft) and Dr. A. Brunner (formerly EMPA) for providing the draft test protocol used in this research. The authors are also grateful for useful feedback on the manuscript from Dr. ir. R.C. Alderliesten and Mr D. Biagini (TU Delft).

References

Alderliesten, R., 2018. Fatigue delamination of composite materials - Approach to exclude large scale fibre bridging. IOP Conference Series: Materials Science and Engineering 388. doi: 10.1088/1757-899X/388/1/012002 . Alderliesten, R., Yao, L., 2017. Mode I fatigue delamination growth in composite laminates with fibre bridging. - Dataset. URL: https://doi.org/10.4121/uuid:6da548f6-f801-41b4-8d88-db9ae81f6913 , doi: https://doi.org/10.4121/uuid: 6da548f6-f801-41b4-8d88-db9ae81f6913 .