Residual stiffness based modeling of fatigue damage and the influence of porosity in thin laminated fiber reinforced composites

Research contributions witnessed in the last three decades have lead to a significantly better understanding of the fatigue damage modeling and life prediction methodologies in laminated fiber-reinforced composites. According to Van Paepegem and Degrieck it is possible to broadly classify these models into three categories: fatigue life models, phenomenological models, and progressive damage models. It was observed during a review of these models available in the open literature, that the presence of porosity was not taken into account for the prediction of fatigue damage and fatigue life assessment. This generally leads to unexpected failures and could have catastrophic consequences in terms of time, money or loss of life. To fully understand the influence of porosity on the fatigue life of laminated composites it is first necessary to have reliable models that can predict the residual strength and residual stiffness as a function of life fraction, applied load, and level of porosity for any given ply sequence. The residual stiffness models will allow for accurate prediction of stress redistribution within the laminate as the loading cycles are accumulated, while residual strength models will make available adequate failure functions through the life span of the laminate.;In this dissertation parametric models were determined, using the non-linear formulation of least squares method, that relate residual stiffness, also used in here as an indicator of fatigue damage, to the life fraction with a 95% confidence level while taking the level of porosity in the laminate into account. These models are empirical in nature and confined to the 8-ply graphite fiber reinforced laminates with quasi-isotropic layup of [0 45 90--45] symm with AS-4 fiber and 3501-6 resin. Three metrics were used in estimating the residual stiffness: one from direct modulus measurements, and two from impact vibration response. These three metrics were scaled to allow easy understanding of relative variations in the stiffness of the laminate as it accumulates loading cycles. Fatigue testing of specimens with three levels of porosity (low, medium and high) was conducted at two load levels (70% load level and 80% load level) with a stress ratio of R=0.1. It was revealed in this work that while residual stiffness modeling was suitable at 70% load level it had limitations at 80% load level. Metric 3 (direct modulus measurement) was found to have the least variability in the data but it was also least sensitive for a given life fraction. Metric 1 (spectral moment) is most sensitive of the three metrics for a given life fraction and also has the most variability. Metric 2 (area below the accelerance FRF) was able to differentiate the residual stiffness trend or fatigue damage behavior of the low porosity specimens from medium and high porosity ones.