The sensitivity of composite scarf joints to the height of the blunt tip

Abstract

As advanced composites see increased use in military and commercial aircraft, there is a growing need to understand the structural behavior of these composite structures after repair. In this work, a numerical and experimental investigation was conducted to characterize the sensitivity of the behavior of the composite scarf joint, particularly with regard to failure, to changes in the height of the blunt tip, the adherend thickness, and the ratio of the two parameters, defined as the "blunt height ratio". All specimens manufactured for the experimental testing and modeled via finite element analysis were constructed from the same adherend material and adhesive material: Toray Industries T800/3900 pre-impregnated carbon fiber/epoxy composite material, and 3MTM Scotch-WeldTM Structural Adhesive Film AF 555M, respectively. In the numerical work, linear elastic plane strain analyses were performed via finite element modeling to determine the stress and strain fields for a homogeneously orthotropic configuration with moduli equivalent to a [±15]nS configuration and for cases of n equal to 1, 3, and 5; blunt tip heights of 0.0.6tply, and tply; and blunt height ratios of 0%, 5%, 8.3%, and 25%. In the experimental work, all configurations considered were tested in uniaxial tension to failure, and have adherends of a [ 15]3s laminate, which are bonded via a single layer of film adhesive with different scarf angles of 3°, 5°, and 10°, and have blunt tip heights of 1, 2, 3, and 4 plies. The results from the numerical investigation indicate that the strain fields of the scarf joint depend on the blunt height ratio, and not any individual lengthscale as the finite thickness of the joint affects the load traveling through the joint. At the corner of the blunt tip, there are large gradients of longitudinal and shear strain that dissipate along the length of the bondline and away from the blunt tip and this magnified strain at the corner is not due to any specific lengthscale, but is simply due to the existence of the corner. As the blunt height ratio increases, the size and magnitude of this gradient increases and approaches the center of the joint due to the effects of the finite thickness. The general response of the scarf joint with a blunt tip can be modeled as a hybrid of an "ideal" scarf joint, dominated by shear, and a butt joint, dominated by longitudinal strain, acting in parallel. In the experimental work, the load-displacement responses of all specimens exhibit the same general characteristics, with an initial linear region that transitions to increasingly nonlinear behavior to the point of maximum load. The maximum load-carrying capability of all test cases indicate that decreasing the blunt height ratio or the scarf angle increases the load-carrying capability of the scarf joint. In addition, variations of the maximum load amongst specimens of the same test case and differences in the type of failure throughout the joint region are shown to relate to the cured film adhesive thickness. Recommendations for future work are presented, particularly the need to perform additional numerical and experimental work using thicker laminates with a constant blunt height ratio on the order of 1%, and with varying heights of the blunt tip in order to determine any sensitivity of the response of the scarf joint to the height of the blunt tip.