The directional solidification process is used to produce single crystal turbine blades for modern aero engines. The spacing of primary dendrites (PDAS) within the microstructure are an indication of the thermal gradient attained during solidification; process optimisation is being driven by the desire to refine the single crystal microstructure and maximise the thermal gradient which can reduce the propensity for defects, such as freckles, forming during solidification. Optimisation of the casting process also allows experimental materials to be investigated such as eutectic alloys, which require more stringent process conditions. A CFD-based method has been used in order to predict the thermal field within the ‘single shot’ Bridgman furnace and validated experimentally. The CFD model has been used to design an optimised heating element with the aim of increasing the thermal gradient achievable when casting single crystal alloys. Dendrite arm spacing has been measured from alloys cast using both heating element designs, with the concept experimental heating element showing a reduction in PDAS of up to 30% compared to the standard design for the same furnace conditions. The viability of casting eutectic alloys with such a furnace has also been assessed both experimentally and analytically using a CFD approach.

The present study is concerned with the single crystal growth of functional oxides by the floating zone technique using a four mirror image furnace. This thesis attempts to extend our knowledge of the subject by firstly studying the optical floating zone method as a technique in terms of the experimental growth parameters and then by performing crystal growth and characterization on a number of functional oxides such as rutile (TiO\(_2\)), Co-doped rutile and rare earth orthoferrite crystals (RFeO\(_3\), R=Er and Y). When growing high quality, large and homogeneous crystals using an image furnace, finding the optimum growth parameters (such as growth speed, rotation rate, gas pressure and type of atmosphere) is a crucial first step since it has been found that these parameters can all affect the eventual crystal quality and properties. Comprehensive characterization of an image furnace revealed that the growth behaviour, and consequently, the quality of the as-grown crystal, can be explained by the effects that varying these growth parameters have upon the temperature profiles within growing crystals. Characterization of as-grown crystals has been performed using a combination of different techniques such as scanning electron microscopy, energy dispersive x-ray analysis, x-ray diffractometry, back reflection x-ray Laue, polarised optical microscopy and vibrating sample magnetometry. The best quality TiO\(_2\) crystals were prepared when neither the growing crystal nor the feed rod were rotated during crystal growth. When rotation was employed, this was found to introduce sub-grain boundaries in crystals when the linear growth rate was low, or bubble inclusions when the linear rate was high. The use of lower molten zone temperatures was found to give pale-yellow crystals, while crystals grown using higher molten zone temperatures tended to have the blue colour characteristic of oxygen deficient material. A study of Co-doped TiO\(_2\) single crystals, prepared from the melt using the FZ technique, revealed that Co does not dissolve into the rutile matrix and that the type of atmosphere used during FZ affects the type of cobalt-based second phases formed. Magnetization measurements of Co-doped rutile bulk samples (both single crystals prepared using the FZ technique and polycrystalline powders prepared using a solid state reaction) also indicated that the magnetic properties depend upon the material preparation conditions. An oxygen deficient environment during the preparation of Co-doped TiO\(_2\) powders and single crystals was crucial for the observation of room temperature ferromagnetism, since preparation in oxygen rich conditions led to the formation of paramagnetic material that included the second phase CoTiO\(_3\). Finally, growth of ErFeO\(_3\) and YFeO\(_3\) single crystals by the FZ technique revealed that the best quality crystals were prepared when the growth rate was 6 mm/h or less, since the use of higher growth rates (12, 18 and 24 mm/h) was found to result in the formation of second phase inclusions within the crystals. The magnetic properties of these crystals, as represented by the hysteresis loops, showed that crystals of both materials grown at higher rates have significantly lower coercivities; this was attributed to the effect of the second phase in nucleating reverse magnetic domains and so aiding the demagnetization process.

Production of gas turbines for jet propulsion and power generation requires the manufacture of turbine blades from single crystal nickel-based superalloys, most typically using investment casting. During the necessary subsequent solution heat treatment, the formation of recrystallised grains can occur. The introduction of grain boundaries into a single crystal component is potentially detrimental to performance, and therefore manufacturing processes and/or component geometries should be designed to prevent their occurrence. If the boundaries have very low strength, they can degrade creep and fatigue properties. The root cause for recrystallisation is microscale plasticity caused by differential thermal contraction of metal, mould and core; when the plastic deformation is sufficiently large, recrystallisation takes place. In this thesis, numerical and thermo-mechanical modelling is carried out, with the aim of establishing computational methods by which recrystallisation during the heat treatment of single crystal nickel-based superalloys can be predicted and prevented prior to their occurrence. This work shows that recrystallisation is most likely to occur following 1.5-2.5% plastic strain is applied at temperatures between 1000oC and 1300oC. This emphasises that high temperature deformation is more damaging than low temperature deformation.

Weldable 13%Cr supermartensitic stainless steels are commonly used for subsea pipelines in the oil and gas industry. Although classified as corrosion resistant alloys, these steels can be susceptible to Sulphide Stress Cracking (SSC) when exposed to wet environments containing chlorides, carbon dioxide and low levels of hydrogen sulphide. Standard guidelines stipulate that laboratory SSC tests are performed at 24 °C and at the maximum design temperature, however some studies suggest that the risk of SSC could be greater at temperatures below 24 °C. Seabed temperatures can be as low as 5 °C, so in-service cracking could occur following shut-down conditions even if the material has been qualified at 24 °C. Four-point bend SSC tests performed at 5 °C and 24 °C in simulated seabed environments showed the material was more susceptible to SSC at 5 °C, but only when the as-received pipe surface was compromised. A supporting stress and strain investigation highlighted strain concentrations on the test surface which were coincident with the location of cracking observed in the SSC tests. Finite element simulations were used to demonstrate that tensile stress-strain data should be used over flexural bend data to load four-point bend specimens to the desired loading strain.

In the present study, the effect of hot isostatic pressing (HIPping) variables such as HIPping temperature, HIPping cycle and powder particle size, on the microstructure and mechanical properties of HIPped samples of two Ti alloys have been assessed. Powders of the most commonly used (α + β) alloy, Ti-6Al-4V and one specific beta alloy, Ti-25V-15Cr-2Al-0.2C wt% (burn resistant titanium alloy, BuRTi) were studied. The Ti-6Al-4V powder was made by the plasma rotating electrode process (PREP). BuRTi powders, which were made both by gas atomisation and by PREP were HIPped to investigate the influence of the initial structure of the powder on the microstructure and associated mechanical properties of the HIPped alloy. The PREP Ti-6Al-4V powder was shown to be fully martensitic in the as-atomised condition. The gas-atomised and PREP powders of BuRTi showed very different as-atomised structures, but in both cases the structure was, as expected single phase beta, with the carbon retained insolution. The individual particles of gas-atomised BuRTi powder were always polycrystalline, although there was a significant scatter in grain sizes within different particles. In contrast the individual particles in the PREP powder were either coarse grained polycrystals or single crystals. These differences led to significant differences in the microstructures and properties of HIPped samples. It was found that HIPping of Ti-6Al-4V samples resulted in the formation of equiaxed regions and lath-like microstructure. The small equiaxed regions are formed by recrystallisation which occurs at original particle boundaries where most of the deformation occurs during HIPping; the lath-like microstructure is formed by simply tempering the (less deformed) original alpha prime martensite within the central part of original particles. Among the three HIPping temperatures used, samples machined from powder HIPped at 930°C exhibited a better balance of properties than those HIPped at 880°C or 1020°C. The fatigue properties of samples HIPped at 930°C, made using different HIPping procedures were compared. It was found that samples which contain the as-HIPped surface, which were made using a new HIPping procedure, have better fatigue properties than samples with as-HIPped, machined or electro-polished surfaces which were produced by conventional HIPping . The properties of optimally HIPped Ti-6Al-4V samples are as good as or better than ingot-route samples. In the case of BuRTi the original single crystals or coarse grained polycrystals in the PREP powder are retained after HIPping and limited grain growth occurs in the gas-atomised samples. The tensile strength is comparable for the gas-atomised and PREP samples, but samples tested to failure showed a significant scatter in ductility (a larger scatter in the PREP powder samples) and all fracture surfaces contained large circular fracture initiation sites, with larger sites associated with lower ductility. Initiation occurs in the centre of these circular regions in large grains or in adjacent grains which have similar orientations and the failed region expands symmetrically in powder samples where no texture is expected. The fatigue properties of the PREP samples are much lower whereas the fatigue properties of the gas atomised samples are better than those of samples from ingot route. This behaviour is associated with obvious facetted failure sites in the PREP powder samples where it is suggested that the coarser microstructure has allowed persistent slip to occur leading to localised deformation and to premature failure. These observations are discussed in terms of the potential of net shape HIPping for the production of engineering components and in this context the fact that a new HIPping schedule has been developed during this study, where the fatigue properties of samples containing an as-HIPped surface are excellent, is very significant.

A high critical current density Jc is crucial for the application of HTS YBa2Cu3O7-δ (YBCO) in the fabrication of energy efficient power devices and wires. We have prepared and studied YBCO films with nanoinclusions for increasing the current-carrying capability. All films were prepared by pulsed laser deposition (PLD) on single crystal SrTiO3 (STO) substrates at optimised condition parameters. We found that the substrate temperature Ts of 780° C, laser energy E ~ 218 mJ/pulses, distance between target and substrate Dt of about 55 mm, annealing oxygen pressure Oap 450 Torr and cooling rate Ct 8°/min were the optimum conditions for making good films with Tc = 91 K. We have used a method for introducing artificial pinning centers which has been shown to be successful in the nanotechnology of pinning centres: distributing a secondary phase, YBCO with 1% Gd2Ba4CuWOy (2411W) nano-inclusions, YBCO with 2% wt BaZrO3 (BZO) and YBCO with 4% wt BZO nano-crystalline particles in the film. The superconducting properties were determined by AC susceptibility, magnetisation loops and transport measurements using a Quantum Design Magnetic property measurement system (MPMS) and a Physical properties measurement system (PPMS). Scanning and transmission electron microscopy (SEM) and (TEM), Atomic force microscopy (AFM) and X-ray were also used to characterise the micro-structure of the films. Both 2411W and BZO nano-inclusions led to increased Jc in applied fields and self-field. Scaling of the flux pinning force based on the Dew-Hughes model has been used in this work for investigating the flux pinning mechanism.

High frequency ultrasound is receiving increased attention for medical imaging in areas such as ophthalmology and dermatology. Recent advances in the manufacture of fine scale piezocomposite materials mean there is great potential for commercial transducers far superior to conventional devices currently in clinical use. This Thesis reports the fabrication and characterisation of passive materials suitable for use in high frequency piezocomposite transducer devices. Epoxy composites have been fabricated using tungsten and alumina as filler material with volume fractions up to 0.4. Acoustic impedance and attenuation has been determined for different filler volume fractions to provide data for modelling to aid transducer design and also determine the filler volume fraction required to provide optimal results. Acoustic impedance values of 3-15MRayl were measured for the materials made in this work and the influence of filler particle size and shape is also discussed. Piezocomposite transducers have been constructed using material developed in this work and compared to devices made with more conventional passive materials. In addition to the fabrication of the composite samples for characterisation a process for incorporating material into transducers is described showing how the fabrication can be a part of the transducer construction resulting in an efficient and neat package.

The transition toughness of low alloy steels used in pressure vessels is of key importance to establishing the safe operation of a number of structures; of these, the integrity of a nuclear reactor vessel is of greatest concern. The through life toughness of such vessels is a combination of the start of life properties and irradiation damage response of the material. Modelling of the inherent scatter of toughness measurements has received much academic interest since the mid-twentieth century and is found to be dependent on a number of metallurgical factors and failure modes; therefore, the micro-mechanisms of the ductile and brittle failure are explored and an assessment of the current best thinking on the modelling of crack arrest toughness is also considered. It has been established in this work that a highly accurate representation of a large toughness database can be achieved by the inclusion of constraint loss effects and the interaction between initiation and arrest toughness distributions.

This present study is concerned with developing a new alloy system which is capable of forming a metallic glass on rapid solidification of the melt, rather than modifying a known glass forming composition, and assessing its glass forming ability. Iron (Fe) was chosen as the solvent element because it is significantly cheaper than the base elements found in some other metallic glasses and does not require the addition of large quantities of expensive alloying elements to enable vitrification. A ternary system using carbon (C) and boron (B) was studied initially as these metalloids are known to aid glass formation in other systems. Manganese and molybdenum were selected as secondary alloying additions in order to determine if they would have an effect on the Fe-C-B alloy with the best glass forming ability. A combination of optical microscopy, scanning electron microscopy, transmission electron microscopy, X-ray diffractometry and secondary ion mass spectroscopy was used to investigate the microstructure of as-cast and rapidly solidified alloys. Differential scanning calorimetry (DSC) was used to investigate the thermal behaviour of the alloys. The ability of the iron-based alloys to form a glass on rapid solidification from the melt could not be predicted by observation of the as-cast microstructure or through computational methods. It was found that vitrification of the ternary system was only possible for compositions which were close to a eutectic point and that stabilisation of the supercooled liquid was caused by competition for nucleation between austenite and metastable phases, rather than between primary equilibrium solidification products. Of the ternary compositions where an amorphous phase was produced it was concluded that Fe₈₀.₉C₅B₁₄.₁ had the best glass forming ability (GFA). It was determined that the addition of manganese and/or molybdenum to the base composition generally had the effect of improving the GFA through the increased complexity of the system making it more difficult for recrystallisation to occur. Of the multi-component alloys it was concluded that Fe₆₀.₉Mn₁₀Mo₁₀C₅B₁₄.₁ had the best GFA as it had the highest values for each of the parameters used to describe GFA. It is believed that this is due to competition between the austenite and alpha stabilisers (manganese and molybdenum respectively) causing enhanced stability of the supercooled liquid.

Laboratory and synchrotron X-ray fluorescence (XRF) have been used to investigate the microscopic and macroscopic distribution of metallic contaminants in membrane electrode assemblies (MEAs) which were used in proton exchange membrane fuel cell (PEM FC) stacks. The laboratory XRF results were consistent with the synchrotron XRF results. Higher levels of contaminants observed for the areas near to the coolant outlet than the areas near to the coolant inlet. The cathode side of MEAs showed higher levels of contamination than the anode side of the MEAs. Synchrotron XRF maps of MEAs generally showed higher levels of contaminants on the cathode side compared with the anode side. Fe was mainly observed in the cathode side microporous layers, whereas Ni, Cr and Cu were mostly accumulated in the cathode side or in the membrane. Synchrotron XRF maps of MEA plan views showed a crack-like distribution for Fe and Pt which were similar to cracks in the microporous layer of the MEAs. A novel electrochemical cell that simulated galvanic and crevice corrosion, temperature cycles for a PEM fuel cell, and pressure across the stacks was designed and used to discriminate between the corrosion behaviour of candidate coatings for bipolar plates.