Stochastic Damage Evolution under Static and Fatigue Loading in Composites with Manufacturing Defects

In this dissertation, experimental investigations and theoretical studies on the stochastic matrix cracking evolution under static and fatigue loading in composite laminates with defects are presented. The presented work demonstrates a methodology that accounts for the statistically distributed defects in damage mechanics models for the assessment of the integrity of composites and for the structural design of composites.

The experimental study deals with the mechanisms of the formation of a single crack on a micro-scale and the stochastic process for the multiplication of cracks on a macro-scale. The defects introduced by the manufacturing processes are found to have significant effect on the matrix cracking evolution. Influenced by the distributed defects, the initiation and multiplication of cracks evolve in a stochastic way. The experimental study on the in-plane shear stress finds the detrimental effect of the shear stress on the fatigue performance of composite laminates. Combined with the transverse tensile stress, the in-plane shear stress induces multiple inclined microcracks in the matrix, which enhance the initiation and propagation of the major matrix cracks.

Based on the experimental investigations, a statistical model for the stochastic matrix cracking evolution on the macro-scale is developed. Simulations based on the statistical model yield accurate predictions for both static and fatigue loading compared to the experimental data. The Weibull distribution of the static strength is estimated by the statistical model by comparing against the experimental crack density data. The estimated Weibull distribution of the static strength provides an efficient approach to characterize the manufacturing quality of composite laminates. Compared to deterministic approaches, the Weibull distribution of the static strength provides comprehensive information of the strength property of composite laminates.