This paper presents an investigation on the effectiveness of crack growth retarders bonded to integral metallic structures. The study was performed by both numerical modelling and experimental tests. It focuses on aluminium alloy panels reinforced by bonded straps made of carbon-epoxy, glass-epoxy composite materials or a titanium alloy. The goal was to develop a fail-safe design for integrally stiffened skin-stringer panels applicable to aircraft wing structures. The modelling strategy and finite element models are presented and discussed. The requirements that the models should meet are also discussed. The study has focused on establishing the extent of crack retarder benefits, in terms of fatigue crack growth life improvement, by numerical simulation and experimental tests of various crack retarders. The results of predicted fatigue crack growth retardation have been validated by tests of laboratory samples. This study concludes that by bonding discrete straps to an integral structure, the fatigue crack growth life can be significantly improved.

a b s t r a c tFibre metal laminates (FMLs), such as glass reinforced aluminium (GLARE), are a family of materials with excellent damage tolerance and impact resistance properties. This paper presents an evaluation of the low velocity impact behaviour and the post-impact fatigue behaviour of GLARE laminate adhesively bonded to a high strength aluminium alloy substrate as a fatigue crack retarder. The damage initiation, damage progression and failure modes under impact and fatigue loading were examined and characterised using an ultrasonic phased array C-scan together with metallography and scanning electron microscopy (SEM). After impact on the substrate, internal damage to the GLARE bonded on the opposite side of the substrate occurred in the form of fibre and matrix cracking. No delamination was detected at the GLARE/substrate bond. Before impact the bonded GLARE strap caused reductions in substrate fatigue crack growth rate of up to a factor of 5. After impact the retardation was a factor of 2. The results are discussed in terms of changes to the GLARE stiffness promoted by the impact damage.

This paper explores the viability of the bonded crack retarder concept as a device for life extension of damage tolerant aircraft structures. Fatigue crack growth behaviour in metallic substrates with bonded straps has been determined. SENT and M(T) test coupons and large scale skin-stringer panels were tested at constant and variable amplitude loads. The strap materials were glass fibre polymer composites, GLARE, AA7085 and Ti-6Al-4V. Comprehensive measurements were made of residual stress fields in coupons and panels. A finite element model to predict retardation effects was developed. Compared to the test result, predicted crack growth life had an error range of -29% to 61%. Mechanisms and failure modes in the bonded strap reinforced structures have been identified. The strap locally reduces substrate stresses and bridges the crack faces, inhibiting crack opening and reducing crack growth rates. In the absence of residual stress, global stiffness ratio accounts for effects of both strap modulus and strap cross section area. In elevated temperature cure adhesives, retardation performance was best in aluminium and GLARE strap materials, which have the closest thermal expansion coefficient to the substrate. Strap materials of high stiffness and dissimilar thermal expansion coefficient such as titanium had poor retardation characteristics.