Influence of cyclic stress intensity threshold on the scatter seen in cyclic Mode I fatigue delamination growth in DCB tests

Abstract

Whilst composite materials are now widely used in aircraft structural components for certification purposes current designs are such that any delamination/disbond will not grow. Despite this limit fleet data and data obtained from full scale fatigue tests reveal that small sub mm initial delaminations/disbonds can grow when subjected to operational flight loads. To account for this the US Federal Aviation Authority advisory circular AC20-107B outlines a slow growth approach for certifying composite/bonded aircraft structures. A key point in AC20-107B is that growth must be both slow and predictable. Since the life of an airframe is determined by the growth of the fastest (lead) crack/delamination/disbond any analysis tool developed for a composite/bonded airframe requires the ability to predict the both the scatter in delamination/disbond growth and thereby the growth of the lead delamination/disbond in the airframe. To this end the present paper reveals that the Hartman-Schijve variant of the NASGRO equation would appear to show promise for capturing the scatter seen in Mode I delamination tests associated with specimens fabricated from nominally identical material supplied by two different suppliers. It is shown that the values of the constants in the Hartman-Schijve equation associated with these two sources are consistent and the scatter in growth can captured by allowing for small variations in the fatigue threshold term.