Assessing failure and delamination growth in composites and bonded joints under variable amplitude loads

Composite and bonded aircraft structures are currently designed such that any delamination/disbond will not grow. However, it is now known that despite this design limit, delaminations and disbonds can both arise and grow when subjected to operational flight loads. In this context in 2009, the FAA Advisory Circular 20-107B introduced a slow growth approach to certification, and this approach is now included in the Composite Materials Handbook CMH-17-3G. To meet the challenges inherent in a slow growth design, this paper summarises selected advances in the state of the art for assessing the effect of damage in composite and bonded airframes. The problem of how to account for the scatter seen in delamination growth is first discussed, and a methodology for determining valid upper bound curves is outlined. Examples that illustrate shortcomings in the current building block approach to aircraft certification are also briefly discussed, and the need for coupon tests to be performed under realistic multi-axial stress states is highlighted. It is also shown that the failure loads predicted using the A4EI computer code, the industry standard approach for determining the failure load of axially loaded bonded joints, and finite element analysis yield very similar failure loads. It is then demonstrated that growth equations that express da/dN in terms of ΔG should not be used to compute crack growth under variable amplitude loads, and an example is presented that illustrates how the Hartman-Schijve equation can be used to accurately compute delamination growth under a variable amplitude load spectra.