Finite Element Based Methodology For Prediction Of Matrix-dominated Failure In Composites

Abstract

The objective of this work is to develop a computational failure prognosis methodology applicable to virtual testing of composites. The major motivation for the establishment of virtual testing is the ability to reduce actual testing of composite structures, which is oftentimes expensive and time consuming, by numerical simulations. A successful virtual test methodology should be applicable to monotonic and fatigue loading, account for strong anisotropy of composite materials and capture the effects of manufacturing irregularities. This work proposes to use three-dimensional finite element modelling for predicting the onset and progression of matrix-dominated failures modes, such as matrix cracks and delaminations, which are the primary modes of failure in composite laminates utilized in aerospace composite structural designs. The ability of the method to predict strength, damage topography and fatigue cycles to failure is verified using actual standard tests of multi-directional laminates. High-fidelity nondestructive inspection techniques, based on micro-focus X-Ray computed tomography, are used to integrate manufacturing defects including porosity/voids into finite element models, and contribute to greater understanding of the effects of defects on the strength and fatigue performance of composites. Finally, the ability to transition the methodology, developed and verified at the material scale, to larger composite structural elements is evaluated.