Fatigue damage modeling based on cracking progress in unidirectional and cross-ply FRP composites

The present study intends to investigate the fatigue damage response of various off-axis unidirectional and cross-ply fiber-reinforced polymer (FRP) composite laminates. A fatigue damage analysis for these composites has been developed based on (i) a cracking mechanism and damage progress in the matrix (Region I), the matrix-fiber interface (Region II) and the fiber (Region III) and (ii) the corresponding stiffness reduction of the composite laminate as the number of ccles progresses. The characteristics of damage growth in unidirectional and cross-ply GRP and CFRP composites materials have been studied and compared with those of available experimental data for the respective materials. Experimentally obtained data versus stress cycles was found to be in good agreement with the predicted values.

The predicted fatigue damage results based on the proposed damage model for FRP composites were also found to be in good agreement with experimentally obtained values of fatigue damage at various cyclic stress levels, stress ratios, and off-axis angles for this material reported in the literature. In the cross-ply (0/90) system, the damage function is found to be dependent on the mechanical properties of the fiber and the matrix in O^o and 90^o plies.

The proposed dramage analysis also puts forward the effect of matrix-fiber interface bonding by introducing the parameter "f". This parameter varies between zero and unity. As f approaches zero, the interface strength drops and the load transfer from the matrix to the fiber decreases, while for f approaching unity, the bonding gets stronger and the load from matrix to the fiber is transferred efficiently. A finite element model (FEM) has been developed to estimate the efficiency parameter by using the displacement method in the concentric cylinders model. A comparison between f values obtained from FEM and the experimental data is also carried out. The predicted parameter agrees well with the experimental data.