|Institution:||Delft University of Technology|
|Keywords:||composites; interlaminar stresses; free edge effects; stress analysis; damage initiation|
|Full text PDF:||http://resolver.tudelft.nl/uuid:f3b96bcb-4df3-466e-8c2e-0bd37bcd8dba|
In design with composite materials, the capability of tailoring the properties of the structure with respect to the specific load requirements in each direction, allowed by their anisotropic nature, is coupled with the development of interlaminar stresses transferred through the thickness. Specifically, at the free edge of the laminate, due to the change in fiber orientation and the unique response of the constituent materials, stresses are developed in the thickness direction. Moving towards the center of the laminate up to a certain threshold, the stress state of the plies is further complicated by that phenomenon. The primary goal of the thesis project was to develop an analytical stress model for both the in-plane and the out-of-plane stresses acting within the laminate. Furthermore, the model was intended to capture a specific sequence of damage occurrences in laminates of the family [02/θ2/-θ2]s loaded under tension, as observed by O’Brien. Namely, during loading, matrix cracks initiating at the free edge of the off-axis plies and propagating along the respective fiber orientation are anticipated to occur first. While these cracks increase in length and number, at some point a +θ crack will intersect with a – θ one, forming an envelope with the free edge. Continuing, a local delamination is expected within this envelope. In order to predict that, the stress expressions would have to be implemented in a failure theory that is capable of accounting for the ongoing phenomena and of distinguishing between different matrix failure modes. These requirements led to the selection of Puck Failure Theory. The process proved to be more complex than estimated. At first, results from literature could not be duplicated. After an extensive period of thoroughly checking and editing the approach, the expressions and the code in which they were implemented, it was realized that a multiplicative factor is distorting the output. After calculating and applying the factor for each one of the out-of-plane stress expressions, agreement with published results was excellent. Unfortunately, the origin of the error could not be found. Once the error has been corrected, the potential integration of the stress expressions in a fatigue life evaluation model specifically designed for composites, could replace the currently employed semi-empirical theories that where originally formulated for isotropic materials and rely mainly on curve-fitting.