AbstractAdhesive bonding is an attractive technique for composite structures due to its high strength-to-weight ratio and excellent stress transfer compared to the usage of metallic fasteners. To ensure the structural integrity of adhesive bonded joints, accurate prediction of fatigue crack growth rate and life is required to determine the inspection regimes.
The majority of laboratory research is based on coupon joints (e.g. Double Cantilever Beam (DCB), End Notch Flexure (ENF), and Single Lap Joint (SLJ)) for fatigue fracture tests. These are narrow in width and may introduce a through-width debond starter. However, realistic structural joints are much wider and may contain process-induced defects and/or accidental damage. Both the patterns are much smaller than the joint width. Consequently, the effect of curvature, constrained edges, and defect size and shape are not studied systematically.
The aim of this research is to develop a finite element model to predict the fatigue debonding behaviour of composite bonded joints with a variety of defects. The key objectives are: (1) to study the effect of curvature, constrained edges and defect size and shape; (2) to investigate the usage of small laboratory coupon joints in establishing the structural integrity of large realistic structures; (3) to assess size scaling to support the Predictive Virtual Testing (PVT) initiative. Work has been undertaken by numerical modelling and experimental testing and validation.
Firstly, a parametric finite element analysis was performed to evaluate the effect of size, curvature and free edges on the fracture behaviour of bonded joints in an assumed large aircraft fuselage panel joint using Cohesive Zone Modelling (CZM). Key findings are that very large curvature has negligible effect on the peak static load capability. The free edge effect is a key issue in predicting the strength of realistic joints using laboratory single lap coupon joints. The size effect is relevant when tackle in different defect scenarios in different joint geometries.
Secondly, static and fatigue tests have been carried out in the laboratory on single lap joints with through-width or semi-circular defects, in conjunction with testing single and mixed-mode fracture specimens to establish the fatigue debonding growth rate of the FM94 adhesive. Linear elastic fracture mechanics (LEFM) based method has been employed to simulate both adhesion and cohesion debonding in non-linear analyses using the Virtual Crack Closure Technique (VCCT). Key findings and main conclusions are: (1) For the FM94 adhesive, the fatigue cracks with mixed-mode ratio smaller than 0.5 behaved very similarly to cracks in mode I. (2) Fatigue damage in wide joints started from semi-circular defect and mainly propagated in the transverse direction until the width of the joints is reached. (3) The predictive model using the normalised fit of the adhesive mixed-mode fatigue debonding growth rate data has demonstrated its capability in predicting the lower and upper bounds of the fatigue crack growth rate for both joints.
Finally, the fracture responses of single lap joints of various sizes containing either a through-width or semi-circular defect were investigated numerically using the VCCT method. It was found that crack tip Strain Energy Release Rates (SERR) of various sized joints can be scaled with the structural length ratio in an empirical relation. The usage of laboratory coupons provides overly conservative prediction of the critical crack length and failure stress if applied directly to the design of large aircraft structures.
|Date of Award||May 2019|
|Supervisor||Xiang Zhang (Supervisor) & Stuart Lemanski (Supervisor)|