Effects of specimen size and sample aspect ratio on the compressive strength of graphite/epoxy laminates

Failure mechanisms and stress/strain behavior have been investigated for different geometries of uniaxially-compressed AS4/3502 graphite/epoxy laminates with [0]48, [±10]12s, [±15]12s, [±30]12s, and [±45]12s architectures. Specimens with various aspect ratios (height/width) and heights were tested with an end-loading fixture. This allowed a simultaneous study of the influence of the amount of fibers constrained by the loading platens on the strength and stiffness of these laminates. The failure stress for each orientation was found to be inversely proportional to the sample aspect ratio and height, without any noticeable change in the failure micromechanisms. Except for the [±45]12s samples which showed a damage strain of almost 8–9% at a relatively constant stress, failure in all other orientations was brittle and catastrophic and characterized by a sharp termination in the respective stress/strain characteristics. While [±15]12s and [±30]12s samples failed on a single shear fault aligned with the fiber direction, [±10]12s specimens suffered multiple delaminations and ply separations upon failure. The origin of these delaminations was traced to the deflection of in-plane matrix shear cracks, and not ply bending, fiber scissoring, or interlaminar shearing. [0]48, samples failed by end brooming, matrix crushing and delamination, without registering the widely observed fiber kinking mechanism, which we believe is observed when the specimen ends are gripped. Thus, this paper questions if any of these failure mechanisms for the [0]48, samples are fundamental.