On the application of fracture mechanics to the size scaling of bonded composite lap joints

The use of large composite materials in large aircraft structures has been widely adopted within the aerospace industry. In the airframe design process, large scale tests (including large components and full scale tests) are required to capture accurately all possible locations of failure. However, large scale tests are expensive and time consuming, which makes it impractical and inefficient to include every possible failure scenario. Current practice is using a test pyramid so that the number of large scale tests can be reduced with the support of testing small scale coupons and components. However, a crucial question still remains: how to relate the small coupon tests to large scale tests.
Previous work conducted numerical and experimental studies of the size effect upon the strength of fibre reinforced composites. These studies showed a tendency for the strength to decrease with increasing specimen volume, where the Weibull strength theory can give a reasonable representation of the size effect. Based on these researches, size effect can be taken into account using the strength properties from small coupons in the design of large structures. However, very limited amount of information is available for the scaling of composite lap joints, which are the primary structural components of aircraft fuselage made of composites. Since aircraft structures are designed for damage tolerance, effect of process-induced defects in bonded joints should be assessed.
The main objective of this paper is to explore the scalability of fracture phenomena in bonded composite joints. A numerical investigation into size effects in composite single lap joints is carried out under the circumstances of two types of defects (strip and semi-circular defects). The in-plane dimensions of the single lap joints with pre-designed defects are scaled up and the crack tip strain energy release rates (SERRs) are determined by the Virtual Crack Closure Technique (VCCT) using Finite Element Analysis (FEA). The baseline geometry of a standard single lap joint is validated with experiments. The study has found that linear elastic fracture mechanics (LEFM) can be applied to the scaling of fracture process of large scale structures. The values of SERR (and also SERR versus crack length relation) can be scaled linearly with the size of the joints, and thus can be used as a normalizing parameter to relate the initial crack propagation and to bridge the dimension gap between large scale composite joints and scaled down lab joints. This study contributes to the concept of Predictive Virtual Testing (PVT) to predict the actual fatigue fracture behaviour of aircraft structures by finite element method for the purpose of replacing or reducing structural tests.