Micromechanisms of failure were studied to determine the influence of specimen size and fiber constraint effects with end loading, and with a view to develop the failure laws. For this study, [±10]12s and [±15]12s, [±30]12s, and [±45]12s specimens with various aspect rations (height/width) were tested with an end loading fixture. Failure modes varied depending on the fiber orientation of the laminate layers to the loading axis. Failure transitioned from shear slipping along the fibers in the [±45]12s and [±30]12s samples to global delamination in the 12s samples. Geometry and size effects were also investigated by comparing samples with various aspect ratios. In addition to simply changing the geometry of the samples, adjusting the aspect ratio varied the amount of fibers constrained at the specimen ends against the platen interfaces. For all of the fiber orientations tests, low aspect ratio specimens had a lower elastic modulus but a higher ultimate strength compared to the larger aspect ratio specimens. Stress concentration and fiber constraint effects introduced by the end loading explain these results. These effects highlight the importance of specimen geometry and fixturing considerations for off-axis compression testing and the sensitivity of stress/strain behavior to specimen size.
Biaxial compression testing was also conducted on [±45]12s and [±30]12s samples with a cruciform material testing machine. Samples were loaded with platens slightly smaller than the sample widths. The confinement ratio R, the ratio of stress applied to the secondary to primary sample axes, was varied from 0.25 to 1 to measure the sensitivity of sample failure mechanisms and stress/strain behavior to different stress states. Failure modes for both fiber orientations transitioned from the uniaxial failure mode to massive delamination with an increasing confinement ratio. Results indicate that the shear stress oriented along the fiber direction dictates failure for confinement ratios from 0 to 0.5, since the shear stress at failure is relatively constant. Above R = 0.5, failure moves toward delamination since the fiber aligned axial stresses that produce buckling begin to dominate the decreasing shear stresses along the fibers. These investigations lead to further understanding of the coupling between failure modes and stress/strain results for biaxial compression testing to allow development of failure models.
A simple local matrix shear criterion in the micromechanical model captured the nucleation of fiber-aligned longitudinal cracks rather well. The saturation cracking state that ensured was determined in terms of the length of the two wing-cracks that nucleated from the edges of the main longitudinal crack due to its mode-III sliding along the fiber axis direction. The stress intensity factor at the wing crack tip was determined in terms of the friction sliding resistance, matrix fracture toughness, and the attendant compressive stress state near the main sliding crack. The model captured the results obtained with different biaxiality ratios and also explained the changes in the failure model to axial delamination.