Scaling global fracture behavior of structures-sized laminated composites

The propagation of damage in laminated fiber composites similar to a crack-like configuration is studied experimentally. Of particular interest is the behavior of a 'macro-crack' or 'global crack' that propagates through a large scale structure (e.g., airplane fuselage or wing) with the potential interaction of stringers or other reinforcements. The question is considered whether such damage propagation can be understood in terms of classical fracture mechanics. Because the damage zone (wake width) can be very extensive and may be measured in terms of inches, the question arises as to the scalability of associated 'fracture' phenomena. An integral part of this investigation is thus an examination of the size of the test specimens to establish whether a minimum size is required to relate fracture at larger scales by laboratory specimens. Using (globally anisotropic) 32-lamina composite specimens, proportioned like compact tension specimens, it is found that test 'coupons' on the order of 45 to 50 cm on a side (18 x 18 in.(2)) or larger are needed to begin simulating large scale structures for establishing a global equivalent of a fracture energy for 'crack propagation' simulations. Displacement controlled tests on groups of three specimen sizes (15, 30 and 46 cm on a side) indicate that scaling can be accomplished through the square root of the linear specimen or crack dimension. In certain lay-ups 'run-away delamination' severs the surface laminae so that the reinforcement action of stringers is jeopardized. Damage at the 'global crack front' is quantified through an effective area relation, with a characteristic value for the mostly intra-laminar initiation of damage (intra-lamina cracks) at a sharp notch, and another value conceived of as governing the onset of unstable growth.