To enhance the bondability of polyetheretherketone (PEEK), surface treatment techniques of plasma and corona discharge have been evaluated. The results have shown that these two methods are effective and practical. The treated materials not only reach their highest possible joint strength, but also show very promising joint durability. In addition, the various environments, e. g. atmosphere, heat, water, and solvent, have little effects on the enhanced bondability of the treated materials. By using Contact Angle Measurement, Scanning Electron Microscopy (SEM), X-ray Photoelectron Spectroscopy (XPS) and Time of Flight Secondaty Ion Mass Spectrometry (TOF-SIMS) to characterise the PEEK surfaces before and after treatment, it was established that the lack of active chemical groups, which if present can form strong interatomic and intermolecular forces across the adhesive/PEEK interface, is the main cause for the poor bondability of the untreated PEEK. Both plasma and corona discharge treatment introduce such active functional groups, for instance, hydroxyl, carboxylic acid, amine and etc., onto the surface of PEEK film and so greatly enhance the intrinsic adhesion at the interface between treated PEEK surfaces and epoxy adhesive, as confirmed by the TOF-SIMS interfacial analysis. It is deduced that low molecular weight molecules (LMWM) are formed on treated surfaces, which contain high concentration of oxygen and/or nitrogen, and can be removed by solvent washing. The removal of LMWM will drastically reduce the wettability of the treated surfaces, but does not impair the enhanced bondability. It has been found that both plasma and corona discharge treated surfaces are in a thermodynamically unstable state. When exposed to the atmosphere, the treated surfaces tend to lose their improved wettability and decrease their surface polarity. Increasing temperature can not only accelerate these processes but also change the surface chemical structures of the treated materials back to that of the untreated films to some extent, as revealed by the TOF-SIMS analysis. Water immersion, on the contrary, tend to reverse the above processes.

The main objective of this research is to investigate the effect of moisture on the degradation of adhesively bonded aluminium joints under both static and fatigue loading. This has been achieved through a combination of experimentation and progressive damage finite element modelling (using a cohesive zone approach). Moisture uptake behaviour in the adhesive was studied to obtain the coefficient of moisture diffusion and of moisture expansion. The coefficient of thermal expansion was also measured to provide the data for the finite element modelling, which included residual stresses due to cooling from cure temperature and swelling of the adhesive layer. The moisture dependent properties of the adhesive were obtained from bulk adhesive tensile tests. The joints investigated included monolithic single lap joints loaded in tension and laminated doublers loaded in bending and tension. Various widths were used to get the full and partial saturation of the adhesive layer. Under both static and fatigue loading, the degradation increased with increasing moisture content and tended to level out when the moisture content approached the saturation level. Most of the failures were cohesive in the adhesive layer, showing that the degradation was due to adhesive, rather than interfacial degradation. Calibration of the cohesive properties was achieved by combining backface strain and load data measured in the monolithic single lap joint. These were then utilised to predict the residual strength of the doublers in bending. In fatigue, the calibrated cohesive zone properties were integrated with a strain-based fatigue damage model to simulate the fatigue response of the monolithic single lap joint and doubler loaded in bending both in the unaged and aged condition. The backface strain technique has been successfully used to monitor the fatigue damage evolution in the joints considered, to calibrate the parameters in the strain based fatigue model, and also to study the effect of adhesive fillet on the fatigue damage evolution in the doubler loaded in bending. Doublers loaded in tension exhibited an entirely different failure mechanism including rupture of both the adhesive and aluminium layer. The effect of moisture on the 11 degradation of this joint was not significant. The butts between aluminium sheets that inevitably exist in laminates were shown to affect the strength of the joint. For these joints, the progressive damage modelling used a cohesive zone approach for the adhesive layer and the butt, and continuum damage for the aluminium layer. In fatigue, the cohesive zone and the continuum damage were integrated with the strain-based fatigue damage model to predict the response. The predicted response under both static and fatigue loading was found to be in good agreement with the experimental data. Finally, experimental and numerical studies have been undertaken on hybrid fibre-metal (aluminium-Glare) laminate (FML) doubler joints to investigate their response under static and fatigue tension loading. The specimens had fibres either parallel to the loading direction (spanwise) or perpendicular to the loading direction (chordwise). Again, the effect of the butt position was investigated. The spanwise specimen was found to have the highest strength followed by chordwise specimens without butts and finally chordwise specimens with butts. The most critical position for a butt was found to be adjacent to the doubler end. Without butts, the static strength for spanwise and chordwise specimens was controlled by the failure in the Glare layer whilst the fatigue failure was precipitated by failure in the aluminium sheet. Where butts are present, they significantly influence the joint response. A progressive damage numerical analysis was undertaken and was found to be in good agreement with the experiment data in terms of both the strength and the failure mechanisms.

A literature survey of adhesion was undertaken with special emphasis on plastics, glass fibres and existing methods of measuring adhesion and comparing physical properties. The effectiveness of adhesion in carbon-fibre reinforced plastics has hitherto been determined by inter-laminar shear strength testing, although the true meaning of this test may still be obscure. The present work set out to measure adhesion between single carbon fibres and polymers, (a) by pulling single carbon fibres from solidified drops of plastic (a technique not previously reported in the literature) and (b) by stressing clear thermoset resins in which single fibres were embedded.

Solid oxide fuel cells are a developing technology where advanced ceramics play a crucial role. Within the fuel cell the most common electrolyte material is yttria-stabilised zirconia (YSZ). The use of nano particles may provide lower sintering temperatures enabling the possibility of co-sintering the electrolyte with the anode layer. One route for deposition is screen-printing, which is scalable and cost-effective. In this research an aqueous screen-printing route was adopted, which adds the benefits of reduced health and safety risks with reduced costs. The aims of the work were to produce aqueous zirconia inks suitable for screen-printing and to understand the factors involved in ink formulation. The overall objective of the project was to produce a dense impermeable zirconia layer by screen-printing. Sub-micron (~100 nm) and nano (20 nm) primary particles were used to prepare suspensions and inks. The suspensions were optimised for dispersion with the addition of a dispersant (Darvan C or TAC) by zeta potential and rheological methods. The addition of a binder (Rheolate 216) to the suspensions was used to create a 60 wt% screen-printing paste. The inks were characterised by rotational and/or oscillatory rheometry techniques and printed layers were characterised by optical microscopy where cracking was not observed. The input parameters for ink formulation have been investigated. The binder concentration increased the structural properties of the ink and above 5 wt% produced level prints. The dispersant structure was shown to influence the rheological characteristics of the ink. This was linked to the interaction of binder and dispersant. This thesis has shown that it was possible to formulate an aqueous sub-micron zirconia ink that levels after printing and does not crack during drying. Defects in the form of pinholes still remained. Excessively high drying rates were obtained from the aqueous ink making it difficult to process over long time periods and this was overcome through the use of co-solvents at a cost of print quality. Surfactants have been shown to improve the printed layer, but a fully optimised ink has not been found.

Urushi is a complex natural polymer that has been used to protect and decorate objects for many hundreds of years. Urushi is an important material as decorated objects can obtain great value and historical worth. A star item of urushi lacquerwares, the Mazarin Chest, property of Victoria & Albert (V&A) Museum in London, is famous as one of the finest pieces of Japanese lacquerware in the world, dating back to as early as the late 1630s. These urushi lacquerwares are often exposed to environments that are detrimental to both their aesthetic appeal and structural performance, and restoration and conservation procedures are needed to preserve these objects over long periods of time. However, the precise behaviour of urushi lacquers is not sufficiently understood to allow accurate prediction of the material response to environmental effects or of the effectiveness of any proposed conservation procedure. Thus a need exists for a comprehensive understanding of this material and a finite element (FE) model to predict the mechanical response to varying environmental conditions. The aim of this research was to model the hygro-mechanical effects during the environmental ageing process of urushi films by means of FE modelling, which will help to make decisions about the environmental conditions required for storage and exhibition. This was achieved by a combination of experimental and computational methods. A synthetic thermosetting (crosslinked) lacquer, polyurethane (PU), was initially studied to develop the methodology of the experimental and numerical studies and to provide a comparison to the natural lacquer. Experimental work was carried out to characterise the material behaviour of both materials, such as hygroscopic expansion, moisture diffusion kinetics and mechanical behaviour under various environmental conditions. Moisture diffusion in the two lacquer films was well described by Fick s law. However, different sorption isotherm models, Flory-Huggins model for PU and Guggenheim-Anderson-de-Boer (GAB) model for urushi, were required to model the equilibrium sorption of these two materials. The mechanical properties of the lacquer films were found to have a complex dependence on environmental conditions. The tensile properties of both lacquer films were shown to change significantly after UV ageing. With increasing time of ultraviolet (UV) irradiation, Young s Modulus and tensile strength increased dramatically, but the maximum strain decreased. With water absorption, both Young s modulus and tensile strength decreased, and maximum strain increased. The two lacquer films were found to behave with a non-linear viscoelasticity, which was highly dependent on environmental conditions. A modified Burger s model was found to provide a good fit to experimental creep data for the PU lacquer well at different stress levels, suggesting this is a satisfactory method for characterising PU rheological behaviour. A novel modified generalized Kelvin fluid (MGKF) model was found to be a powerful non-linear viscoelastic model capable of representing the rheological behaviour of the urushi below the yield stress. However, in order to include the post yield behaviour a visco-elastic-plastic model is required. A hygro-mechanical model of the urushi behaviour based on the MGKF viscoelastic model was developed and tested. Through careful determination of the mechanical behaviour the constitutive properties of a thin layer of lacquer were determined and used as an input to a FE based model of the deformation and stresses that develop in response to changes in the environmental conditions. The model was validated using experimental results that showed the depth averaged stress in a thin layer deposited on a glass substrate, which allowed insight into the time dependent and spatially varying stresses within the layer. It was shown that the regions of highest stress were found in areas of highest moisture ingress, emphasising the need for control of the environment in which urushi covered artefacts are stored. In order to perform a predictive study on the bending behaviour of urushi films with cracks, as can be seen in aged lacquers, a model with a grooved urushi film on an aluminium substrate was created and subjected to bending loads. The time-dependency of the urushi material properties seems to hardly affect the bending behaviour of the model, however, the profile of the displacement field around the groove was found to be considerably affected by the geometry of the groove. To evaluate the effectiveness of a traditional Japanese consolidation method for lacquerware objects, known as urushi-gatame, a strain-based progressive damage failure model was used to model the continuum failure in the bi-material strip under an increasing bending load. The behaviour of damage initiation and evolution was modelled for an unfilled groove, a filled groove with fresh filler and filled groove with UV aged filler. From the finite element analysis (FEA) results, the introduction of the filler, as a simple mimic of the consolidation method, does enable the strip to sustain a higher bending load. However, this effectiveness is weakened as the material is aged, with it behaving similar to a groove without any filler.

Composite materials consisting of two or more different phases are very extensively used in modern society. If the composite is designed and fabricated correctly, then desirable properties not available in any single conventional material can be achieved. Ceramic reinforced aluminium alloys are desired for high performance applications due to their superior properties compared with the soft, unreinforced metal. However "traditional" particle or fibre reinforced composites suffer from a limited ability to achieve high reinforcement levels. Interpenetrating composites (IPCs) have 3-3 connectivity, with both the matrix and reinforcement phases being fully connected; they are expected to provide truly multifunctional properties. Whilst pressure is normally needed for the processing of IPCs due to the poor wetting between most aluminium alloys and ceramic materials, it raises the risk of ceramic preform damage and limits the component shape. In this research, interpenetrating composites were produced at atmospheric pressure by infiltrating 10 wt% magnesium content Al-Mg alloys into 15-40% dense, gel-cast ceramic foams with average cell sizes from 100 to 500 μm, made from three different ceramics. Previous research at Loughborough University focused on the aluminium / alumina system. In this study, the ceramic foams were made from spinel, mullite and silicon carbide. Effects of processing parameters, including atmosphere, temperature and time, were investigated. The results showed that heating the metal-ceramic couple in Ar and infiltrating in N2 followed by cooling in Ar was a better approach for the infiltration process than heating in N2 during the whole process. The Al(Mg)/spinel system was observed to require the lowest processing temperature and shortest time compared with the Al(Mg)/mullite and Al(Mg)/alumina systems. Microstructures of IPCs were characterised using a series of techniques, including optical microscopy, field emission gun scanning electron microscopy (FEG SEM), X-ray diffraction (XRD), dual beam focused ion beam (DBFIB) and transmission electron microscopy (TEM). A continuous nitride layer consisting of AlN and Mg3N2 was observed at the metal-ceramic interface of the spinel and mullite-based IPCs ABSTRACT ii with MgO and MgAl2O4 observed at localised positions, similar to alumina-based composites. Based on these results, a two-step nitridation infiltration mechanism has been proposed for oxide ceramic foam / aluminium-magnesium IPCs. The infiltration is believed to be dependent on the reaction between Mg and N2 to form Mg3N2, which then deposits onto the oxide ceramic foam surface; once in contact with molten Al, Mg3N2 reacts with the Al to form AlN, which is wetted by the liquid aluminium and induced the infiltration. In the case of mullite-based composites, a small amount of Mg2Si was observed as a result of the reactions between the SiO2 phase in the mullite foam and the liquid metal. The feasibility of fabricating SiC foam / Al-Mg and SiC foam / Al-Si IPCs by pressureless infiltration of molten Al alloys into gel-cast SiC foams has been also evaluated in this research. Serious degradation of the SiC foam was observed in the SiC / Al-Mg IPCs, resulting in the formation of Mg2Si and Al4C3, whilst the SiC foam could not be spontaneously infiltrated by the Al-Si alloy without the presence of Mg. A modified pressureless infiltration technique was developed to allow the manufacture of fully infiltrated SiC foam / Al-Si interpenetrating composites, with little degradation of the SiC foam and very little formation of detrimental phases. Preliminary property characterisation showed that the ceramic-foam based IPCs were up to twice as wear resistant as composites made by infiltrating a bed of ceramic powder. Effects of parameters on wear resistance have been investigated, including the ceramic material, foam density, cell size and degree of sintering. The denser the ceramic foam, the stronger the foam struts, and hence the more effective the composites were in resisting wear. However, a non-linear relationship between the foam cell size and the wear rate was observed; the composites with moderate mean foam cell sizes exhibited better properties than composites with smaller or larger cell sizes. Thermal expansion behaviour of the IPCs has been also studied; a clear hysteresis was observed in the strain curve between heating and cooling. The coefficient of thermal expansion (CTE) was observed to vary as a function of temperature.

No description available.

The concrete lining of traffic tunnels is normally exposed to combined carbon dioxide (C02) and chloride environments. The ingress of CO2 and chloride ions into concrete in these environments could be different from that in I their independent effects, but relatively little is known about this, particularly for high performange concretes (HPCs) that are commonly used in tunnel linings. In this project, changes to hydrated cement pastes and concretes containing pulverised fuel ash (PF A) and microsilica (MS) during both independent and combined exposures to CO2 and chlorides were studied. Effects of these test regimes on the rate of carbonation and chloride ingress were quantified by analysing profiles of the consumed hydroxyl ions and chloride ions. The suitability of using Autoclam Permeability System, Permit Ion Migration Test and Covercrete electrode array for assessing the extent of carbonation and chloride ingress in normal and high performance concretes in the above exposure environments was explored. The results indicated that exposure to chlorides before subjecting to carbonation significantly slows down the progress of carbonation in concretes. However, the influence of carbonation on chloride ion ingress depends on concrete type and carbonation degree. It was found that this combination could severely compromise the tunnel lining durability, particularly for concretes containing PF A, due to the formation of microcracks in carbonated concretes. Autoclam and Permit enabled an assessment of the variations in permeation properties due to the test regimes, and the Covercrete electrode array could identify the carbonation depth and the extent of chloride ingress.

The use of rotating cylinder electrodes for the investigation of aqueous erosion-corrosion has determined that they are suited to the task with some minor modifications. The requirements that need to be met to ensure that true turbulence is produced within a cell to allow reproducible investigation of a range of erosion corrosion parameters have been elucidated. The cell must operate above Taylor numbers of 15000 and as such large annular gaps and higher temperatures are beneficial. Careful consideration of the conditions required needs to be made before designing a rotating cylinder for the study of erosion corrosion. The use of outer cylinder static electrodes has been determined, through mass transfer tests to be valid for a small annular gap of 5mm, allowing experiments to be set up in autoclaves that reproduce hostile process plant parameters. Fixed potential studies of the erosion corrosion of mild and chromium steel in deaerated 1M NaOH have determined that over the majority of the potential range studied there was no synergistic effect between corrosion and erosion. The erosion-corrosion rate was just the sum of the corrosion rate during erosion and the erosion rate in the absence of corrosion. Over a small potential range in the transpassive corrosion region of the chromium steel a negative synergistic effect was determined. In this region, the transpassive corrosion of chromium from the steel inhibits the erosion of the metal by impacting particles. The effect of particle size on total erosion-corrosion, and its separate erosion and corrosion components has revealed that for the same addition of particles by weight, the corrosion rate under erosion is independent of the particle size. However, the erosion produced is heavily dependent on the particle size, being approximately constant for large sizes and decreasing rapidly below 100jm. These findings have been corroborated by microscopy of the sample surfaces.

The papers presented here have been grouped into the following main themes: A. Atmospheric corrosion; B. Surface engineering; C. Corrosion in concrete; D. Degradation of polymer composites; E. Solid-state science and technology; F. Miscellaneous.