A critical review of the current knowledge of PEO coating characteristics and properties is made, and several areas requiring more accurate or more detailed investigations are identified. A leading commercial product – the <i>Keronite^TM</i> coating for aluminium alloys – is the used as a basis for the investigation of the microstructure and properties of PEO coatings. Coating structure and morphology are studied quantitatively to investigate their growth mechanism. Composition is accurately determined for the first time, allowing predictions of physical, mechanical and thermal properties to be made. Particular attention is also paid to the presence of amorphous material and fine-scale porosity – properties which had previously been neglected. The latter is critical to the understanding of coating formation and the capacity for coating impregnation, and is measured and characterised using numerous porosimetry techniques. Mechanical properties of the coatings are characterised using indentation and macroscopic techniques such as beam bending. Correlations are established between the observed structure and measured physical properties such as hardness, local modulus and global stiffness. It is found that wear resistance can also be explained on the basis of the measured mechanical properties and structure. The discovery of low coating stiffness means that high-temperature applications, which had previously been dismissed on the basis of thermal expansion mismatch between the coating and substrate, may indeed be possible. The thermal stability of the coatings is therefore investigated and their stability up to 800°C is demonstrated. Residual stresses are measured and explained in terms of the postulated coating growth mechanism.

No description available.

No description available.

A method devised by Edwards and Gupta for nematic polymers is implemented here for semiflexible filaments which enforces mean inextensibility. This linearises the path integral so that analytic results can be obtained. I argue that the expressions for the probability density obtained using this approach have merit – successfully reproducing the limiting cases of flexible and rigid chains – and have the advantage of being algebraically simple. I develop an algorithm based on the Metropolis method that can calculate the statistical properties of single chains subject to arbitrary interactions. I use this algorithm to check the analytical results obtained using the mean inextensibility approximation and find they are in close agreement. I highlight the applicability of the algorithm to arbitrary interactions by considering the statistical properties of a single chain which is stearically confined to a tube. I develop a model for the elasticity of semiflexible polymer networks based on the assumption of affine strain that uses the simple analytical expressions for probability densities obtained using mean inextensibility. Under such assumptions, I find that mechanical stretching of filaments must be considered in the network energy. This in turn implies that there are three distinct scaling regimes of the modulus which are relevant to stiff filament networks. I check the validity of the homogenous strain assumption by developing an algorithm capable of calculating the elasticity of a network of filaments subject to arbitrary strains and interaction energies. The effect thermal fluctuations, non-homogeneous deformations and variations in filament length have on elasticity is then explored. I find that all are in fact crucial in determining elasticity correctly, particularly at small strains.

The continuous research of lightening mechanical and aeronautical structures is leading to the adoption of innovative materials and structure designs. and this development has to be necessarily followed by research in optimising structural health monitoring techniques. The set up of a reliable and quick monitoring technique. together with the enhancement of the mechanical performances of structures, is the tUl1ung point to the usage of even lighter and longer-lifed structures. Being impossible to build up a monolithic structure, it is common sense to understand that the most critical part of a structure is the joint between two or more different parts. For this reason, research is being focused on both the improvement of existing joint configuration, and the development of automatic systems which are able to monitor the health state of these joints when the structure is in service. The main reason urging to improve the current composite fastening joint and to develop innovative in situ monitoring techniques is linked to the continuous pursuit of cost reduction. In fact, a stronger and longer lifed fastening configuration, together with a reliable. quick and real-time non-destructive monitoring system applied on in service structures, would allow a considerable cut in maintenance time, with a consequent drop of costs. The objectives of this PhD work are manifold. On one side it focuses on the development of an innovative structural coupling configuration which is able to improve the mechanical properties of the joint; on the other hand, it aims to define reliable indexes able to assess the state of health of a joint structure. More precisely. the entire part concerning the monitoring side of the problem, it focuses on the definition of different possible method capable of determining the loosening state of a bolted connection. These methods are all based on the vibration and ultrasonic based techniques. The first step consists in the development of indexes able to predict the healthy state of a joint by studying some properties related to acoustic signals passing through the part's connection. After theoretically defining the parameters involved, experimental campaigns aimed to validate the models have been carried-out. Once the indexes were found, and confirmed to work on a small simplified structure, the most reliable and accurate IS chosen to be applied for the monitoring procedure on larger and more complicated structures. Both theoretical and experimental approaches were adopted to justify the reliability of the used method. Ultimately it was proved that with the adoption of few sensors and natural frequencies based vibration, with the developed method, it is possible to assess the location of a loosening bolt also on a multibolted structure. Additional research has to be done on real larger scale structures, hut promising results based on solid theoretical principles have been achieved. With regards to the necessity of enhancing existing fastening coupling configuration, all the efforts concentrate on the adoption and optimization of hybrid joints. Hybridization joint process takes advantages of positive aspects of both composite materials and fastening joints. Most important, the rising use of composite parts contributes to the reduction of structures' weight, with all the advantages involved, such as for instance, fuel consumption reduction with a consequent decrease of running costs. For this reason the second part of the thesis will discuss possible solution able to get out the best from coupling together composite materials and fastening configuration. More precisely all the attentions will focus on how to reduce problems generated by fastening composite parts, such as low bearing and shear strength, notch sensitivity and high dependency of joint strength from the lay-up configuration. The proposed solution will be as strong as the hybrid solutions developed so far, but also lighter at the same time

We present asymptotic approximations of Green's kernels for operators of linear elasticity in planar and three dimensional domains containing small subdomains, with either Neumann or Dirichlet boundary conditions whic~ correspond to solids with voids or rigid inclusions. The main feature of these approximations is their uniformity with respect to the independent variables. The asymptotic formulae are supplied with rigorous remainder estimates. We present numerical experiments, comparing the asymptotic approximations and finite element numerical simulations, which show the effectiveness of our approach.

No description available.

No description available.

A study has been completed investigating the elastomer seal failure phenomenon termed Explosive Decompression. In attempting to examine this phenomenon, a two pronged investigation strategy was initiated. The first of these was to develop a method for quantifying cavitation damage within elastomer materials as an attempt to differentiate the controlling parameters. The second strategy was to measure the effects on the mechanical properties caused by high pressure CO<SUB>2</SUB> swelling. These changes, if found to be significant, would have to be considered during any future developments of a predictive model. After the successful development of a quantification method for measuring cavitation damage and the design and development of equipment for high pressure (50 bar) and high temperature (180 °C) CO<SUB>2</SUB> tensile and tear testing (Rubber Tensometer), ten engineering elastomers were examined in "O" ring and sheet form. The results showed that a drop in sealing force is experienced and can be measured after elastomer seals are exposed to explosive decompression from saturation with 40 bar CO<SUB>2</SUB>. Moreover, in all ten elastomers studied, ultimate tensile stress, ultimate tensile strain and tear strength were significantly reduced after saturation with 40 bar CO<SUB>2</SUB>. In light of these results it seems reasonable to expect that both the method and subsequent results should be considered during any future investigation into explosive decompression or into any area where elastomers are dynamically used whilst in contact with high pressure swelling gaseous environments.

No description available.