This dissertation presents a novel multiscale fatigue life prediction framework for carbon fiber reinforced polymer (CFRP) laminate composites that unifies intra- and interlaminar damage mechanisms. To describe constituent-level degradation, Mori–Ta...
This dissertation presents a novel multiscale fatigue life prediction framework for carbon fiber reinforced polymer (CFRP) laminate composites that unifies intra- and interlaminar damage mechanisms. To describe constituent-level degradation, Mori–Tanaka’s mean-field micromechanics theory was employed, and a modified Chaboche fatigue damage model was formulated to capture nonlinear stiffness and strength reduction throughout the primary, secondary, and tertiary stages of cyclic damage evolution. Fiber–matrix interfacial degradation, which is difficult to quantify experimentally, was directly evaluated using interrupted fatigue test based X-ray computed tomography (CT) image processing, and the resulting interfacial damage behavior was cross-validated with molecular dynamics (MD) simulations. Manufacturing-induced voids were characterized through CT image analysis, and their agglomeration effects on homogenized stiffness and strength were quantified using a finite element based representative volume element (RVE) approach. These results were further incorporated into the stochastic multiscale fatigue framework to account for microstructural variability. The complete framework was implemented in ABAQUS via user subroutines, UMAT for intralaminar behavior and UEL for interlaminar delamination, allowing direct coupling between micromechanical degradation, cohesive damage evolution, and macroscopic fatigue response. Validation was performed through fatigue experiments on T700/Epoxy laminates with various lay-ups ([0]7, [90]14, [±45]4s, [0/90/±45]2s) and stress ratios (R = –1, 0.1, 10). The simulation results showed excellent agreement with experimental S–N curves and failure morphologies, confirming the predictive accuracy of the proposed multiscale intra- and interlaminar fatigue model. Furthermore, the framework was extended to incorporate freeze–thaw (FT) environmental effects based on experimental observations. The FT cyclic environment was found to weaken the fiber–matrix bond, as confirmed by SEM analyses across all tested specimens. However, the fatigue life of CFRP laminates is predominantly governed by the dominant failure mechanisms, which depend on the applied stress ratio and laminate configuration. In shear specimen (SS) configuration, where fiber–matrix debonding is the primary failure mechanism, the FT environment directly accelerates interfacial degradation, leading to a pronounced reduction in fatigue resistance. By incorporating temperature-dependent thermal strain differentials, the proposed framework successfully reproduces the experimentally observed fatigue-life reduction under FT conditions, demonstrating its capability to capture environmental effects on multiscale fatigue behavior. Overall, the proposed multiscale framework provides a physically consistent and experimentally validated approach for predicting fatigue behavior of CFRP laminates, accounting for constituent degradation, fiber-matrix interface damage, void agglomeration, and environmental influences. The results demonstrate its potential as a robust tool for the design, reliability assessment, and lifetime prediction of advanced composite structures operating under complex service conditions.