Master of Science

thesisInterlaminar fracture, or delamination, is a common failure mode observed in high-performance tape-laminate fiber-reinforced polymeric (FRP) composites used by the aerospace industry. Typically, initiation or growth of delaminations in FRPs occurs due to mode I (opening), mode II (in-plane shearing), mode III (antiplane shearing), or a combination thereof. To date, a number of experimental test methods have been proposed to measure delamination toughness associated with the diffierent modes of loading, and as a result, several tests for measuring mode I, mode II, and mixed-mode I-II toughness have been standardized by ASTM International. Despite these e orts and nearly 40 years of research, a viable test for measuring the true material toughness associated with mode III delamination has not yet been developed. Among all tests proposed for measuring mode III delamination toughness, GIIIc, the edge crack torsion (ECT) test has received the most attention due to its relatively simple specimen geometry and data reduction procedure. However, the ECT test has recently been shown to produce GIIIc values that are dependent on the geometry and stacking sequence of the test specimen and on the data reduction approach used to calculate GIIIc. This thesis reexamines the limitations and applicability of the ECT test by modifying the specimen geometry and by conducting delamination tests on specimens with two new candidate stacking sequences. In addition, to investigate the dependence of GIIIc on the data reduction method, this work proposes a novel way of calculating GIIIc using finite element (FE) models, coupled with a calculation of critical strain energy release rates using the virtual crack closure technique (VCCT). The results obtained from this work show that the ECT specimen may result in planar delamination growth or delamination migration, depending on the overall torsional and exural rigidity of the samples. The latter failure type is undesirable because it produces artificially high \apparent" GIIIc values caused by additional ply-level failure mechanisms (e.g., intralaminar cracking). Despite this fact, one of the ECT specimen geometries investigated in this study did produce pure mode III delamination growth, and, combined with FE/VCCT-based data reduction, produced specimen-independent values of GIIIc. Although FE/VCCT-based data reduction precludes the ECT test from being used by the broader composites community, this work offers several possible ways of further modifying the ECT test that may ultimately lead to its standardization. i