Finite element (FE) analyses are carried out on bolt bearing testing scenarios based on data found in the literature. Both layer-by-layer and smeared property FE models are created to calculate the compressive characteristic dimension (CCD) for three GLARE variants. A novel re-definition of conventional CCD is proposed which is governed by the yield strength of aluminum. The new definition also incorporates the two-phase nature of GLARE, as well as the delamination/ buckling phenomenon for pin/bolt bearing, in a bearing failure mode. A previously unconsidered, orthotropic plate buckling analysis is also conducted in a conservative, worst case scenario sense on the laterally unsupported prepreg layers. Results of the buckling analysis suggest that the prepreg contribution to bearing strength, in a bearing failure mode, is at best negligible and joint collapse is governed by the yielding and delamination of the aluminum layers. Calculation of a CCD, based on the new yield strength definition, produced consistent values amongst all GLARE variants considered in the layer-by-layer analysis suggesting that the CCD is a property of the material alone.

A novel experimental methodology extended from ASTM D953 was developed and implemented to conduct pin bearing experiments on GLARE3-5/4-0.3 and GLARE3-4/3-0.3 variants with the aim to examine a bearing load configuration from a local perspective as well as a global one. This was accomplished by introducing previously unconsidered testing fixtures and additional instrumentation, including bonded strain gages, as a means of measuring the local strain field generated by a bearing load. Load—displacement curves were produced as per the standard but were subject to numerous plateaus and significant statistical scatter, attributed to pin seating and global displacement measurement. Conventionally defined bearing yield strengths were calculated but lacked physical meaning, prompting the production of novel bearing strength vs. measured strain profiles. These new profiles, derived from the additionally acquired data, were void of the plateau anomalies and depicted a well defined trend, indicative of a more complete characterization of material response. Examination of these profiles resulted in a new, more intuitive definition of bearing yield strength that incorporated influence from the entire curve rather than idealizing it bilinearly. The phenomenological basis of the new definition suggests an analogous extension to additional methodologies such as standard tensile or compressive tests.

A baseball injury to an instrumented human cadaver maxillae was simulated with a regulation (142 g) baseball traveling at 14 m s(-1). Measurements of strain were obtained with three-axis strain gauge rosettes located at the medial palate and both canine fossae. A three-dimensional finite element (FE) model of a dentate human maxilla was constructed from computed tomography scans of the skull of an adolescent. This three-dimensional mathematical model of the maxilla was deemed geometrically accurate by convergence testing when the model's degrees of freedom approximated 74 000. The simulated load case involved a transient dynamic impact to the medial maxilla with boundary conditions imposed at skeletal buttresses of the model. The model was calibrated through direct comparison with the displacements and principal strains gathered from experimental and epidemiological data. The comparison of experimental and calculated principal strains as a result of the simulated impacts revealed a 1.7-11.4% difference.

Experiments enforcing a pin bearing loading configuration were performed on a fully orthotropic GLARE 4 variant. The protocol employed in such experiments stemmed from a similar methodology performed on quasi-isotropic GLARE variants though now incorporating a local measurement scheme using biaxial strain gauges rather than uniaxial ones. The aim of this local measurement was both the extraction of novel bearing yield strength values and the detection of buckling within the aluminum layers. The encouragement of delamination and buckling was key since they not only form an integral portion of the proposed yielding through delamination buckling (YDB) mechanism but in addition, the pin bearing configuration -which ipso facto, encourages the former(s) -has been identified in the literature as the most conservative and accurate means for analyzing joint behavior and collapse. Analytical calculations previously performed support the empirical findings and provide direct evidence for the hegemony of aluminum yield strength in joint collapse. The proposed and employed protocol has been shown to be effective across a comprehensive range of GLARE variants and may be extended analogously to other standardized testing methodologies.

The development of Fibre Metal Laminates (FMLs) for application into aerospace structures represents a paradigm shift in airframe and material technology. By consolidating both monlithic metallic alloys and fibre reinforced composite layers, a new material structure is born exhibiting desired qualities emerging from its heterogeneous consituency. When mechanically fastened via pins, bolts and rivets, these laminated materials develop damage and ultimately fail via mechanisms that were not entirely understood and different than either their metallic or composite constituents. The development of a predictive methodology capable of characterizing how FMLs fastened with pins behave and fail would drastically reduce the amount of experimentation necessary for material qualification and be an invaluable design tool. The body of this thesis discusses the extension of the characteristic dimension approach to FMLS and the subsequent development of a new failure mechanism as part of a progressive damage infinte element (FE) modeling methodology with yielding, delamination and buckling representing the central tenets of the new mechanism. This yielding through delamination buckling (YDB) mechanism and progressive FE model were investigated through multiple experimental studies. The experimental investigations required the development of a protocol with emphasis on measuring deformation on a local scheme in addition to a global one. With the extended protocol employed, complete characterization of the material response was possbile and a new definition for yield in a pin bearing configuration was developed and subsequently extended to a tensile testing configuraiton. The performance of this yield definition was compared directly to existing definitions and was shown to be effective in both quasi-isotropic and orthotropic materials. The results of the experiments and FE simulations demonstrated that yielding (according to the new definition), buckling and delamination resulting in joint collapse and failure have all occurred within the stipulated predictions of the YDB mechanism.