A literature survey on the whole field of pore formation is assembled into the form of a general theory of the causes of porosity in castings. The conventionally accepted modes of feeding are assessed: liquid-, mass-, and interdendritic- feeding; and two further mechanisms are proposed: burst- and solid-feeding. The latter is investigated theoretically using various flow models: elastic-plastic, viscous, creep and Bingham flow. A new theory is proposed for the origin of layer porosity in castings. Experimental work on a wide variety of alloys: Al-Cu, Fe-C, Complex Ni- and Co-base alloys, cast both in air and in vacuum are investigated for the effect of section thickness, taper, and mould and metal temperatures. The formation of porosity appears to change from a non-nucleation to a nucleation mechanism as section thickness increases. A new method of interpreting radiographs based upon a longitudinal line count reveals that solid feeding becomes important in reducing porosity at high mould temperatures. Experiments on the effect of composition of an alloy on porosity cast doubt on the widely accepted theory that the presence of non-equilibrium eutectic liquid reduces porosity, but indicate that the non-equilibrium freezing range of the alloy may be the critical parameter. The effect of pressure on porosity is investigated utilising pressures below atmospheric; the results are inadequately explained by current theories and are discussed in terms of the nucleation and growth of pores; the effect would also appear to have considerable industrial potential for reducing porosity in vacuum cast components.

Aluminium and their alloys readily react with air to form surface oxides and films. If they become entrained in a casting, they form double oxide film defects that reduce the mechanical properties of the component. It is important to determine the nature and behaviour of the films formed, so that defects can be avoided or their internal gas volumes reduced. The reaction of common additions and impurities that may be present in aluminium alloys were melted at 750 °C in air and nitrogen, to determine thickness, composition and structure of possible films that may form in a defect. The effects of molybdenum were also investigated. It has also been suggested that oxides formed on aluminium alloys may act as heterogeneous nucleation points for solidification. Spinel containing copper layers were manufactured by electro-codeposition, and added to commercial purity aluminium to test its potential as a grain refiner. It was compared with non-refined aluminium and aluminium refined with standard Tibor refiner. The lattice parameters of oxides formed were also compared with that of aluminium to ascertain their potential as nucleation centres. The results are discussed in this work.

Pitting corrosion of iron has been studied via artificial pits. Solid corrosion products were observed within the pit, which was characterised as disordered-carbon and Fe\(_3\)C, acted as diffusion barriers for metal ions and slightly increased the solution resistance. Its formation depends upon the interfacial potential. High purity iron was used to calculate an effective diffusion coefficient for metal ions (a combination of self-diffusion and electrical migration) in different MgCl\(_2\) concentrations. The contribution of self-diffusion increases with increasing MgCl\(_2\) concentration if Mg\(^2\)\(^+\) depletion in the pit is considered. The Tafel kinetics of iron dissolution in metal ion saturated solutions was studied in different MgCl\(_2\) concentrations. The Tafel slope (56 to 70 mV/decade) was independent of MgCl\(_2\) and FeCl\(_2\) concentration when FeCl\(_2\) is saturated. The effect of nitrate on the composition/structure of salt layers was characterised. The salt layer is composed of FeCl\(_2\).4H\(_2\)O in chloride-based solutions (isotropic in HCl and anisotropic in HCl with trace nitrate), and Fe(NO\(_3\))\(_2\).6H\(_2\)O in nitrate-based solutions. The dissolution behaviour of iron in chloride/nitrate solutions was studied. Dissolution is uniform under a salt layer, but crevice formation and surface roughening. Potentiodynamic measurements can induce abrupt dissolution/ passivation transitions, resulting in deep and localised attack.

An investigation into the optimisation of magnetic properties of two types of magnetic alloys based on Sm(Co,Cu,Fe,Zr)_z (z = 7.4) and Nd₁₅Fe₇₇B₈ has been carried out. The optimum heat treatment which gave the best magnetic properties for the Sm(Co,Cu,Fe,Zr)_z alloy depended primarily on the sintering and solid solution temperatures. After quenching the alloy from the solid solution temperature, and then ageing isothermally, it was found that the intrinsic coercivity, microhardness and electrical resistance changed with ageing time, and the ageing temperature also affected the rate of these changes. The magnetic properties of specimens produced from the Nd-Fe-B type alloy were also shown to be affected by different heat treatments. The microhardness, microstructure and intrinsic coercivity were affected by an isothermal ageing treatment and the temperature of the treatment. Two routes were used for powder preparation 1n the production of the sintered specimens: i. Mechanical ball milling of crushed ingot material. ii. Hydrogenation followed by ball milling. It was found that higher energy products were achieved in the ball milled material for the two alloys investigated. Two reasons for this were proposed: i. The hydrogenated and ball milled powder had a larger size than the ball milled only powder. ii. The surfaces of the hydrogenated powder were highly reactive in air and some oxidation of the surfaces occurred. Optical and electron microscope studies showed that the variations in the ageing behaviour of both the alloys was concurrent with subtle differences in the microstructure between the samples aged at different isothermal ageing temperatures.

Magnesium hydride is a promising material for hydrogen storage. However, the thermodynamic stability of the hydride coupled with slow reaction kinetics means that it is currently infeasible for application as mobile hydrogen storage media. Thin films of magnesium have been investigated to ascertain the influence of stress on the effects of the sorption thermodynamics and kinetics. It was found that: • The intrinsic stress state can be heavily influenced by deposition conditions. • The influence of the underlying microstructure heavily influences hydrogen sorption mechanics. A model has been proposed which shows regimes dominated by porosity, elastic and plastic contributions and their likely effects on sorption thermodynamics. • Stress analysis using different film thicknesses has been used to model the diffusion behaviour within the films, demonstrating how differential rates and directionality of absorption also affect both kinetic and thermodynamic properties of the films. • A novel approach to tailor thin film stress using flexible substrates was used to investigate and tune thermodynamics. The use of an opto-mechanical hydrogen sensor based on this system design. • Y/Mg multilayer systems were investigated for the first time. The characterisation of these shows that hcp/fcc yttrium occurs at different layer thicknesses, altering the lattice spacings at coherent interfaces with Mg.

Steel properties are controlled by its microstructural parameters, such as grain size, phase balance and precipitates. It is desirable to monitor microstructural changes during processing, allowing in-situ feedback control, or microstructure characterization in a non-contact and non-destructive manner. Electromagnetic (EM) sensors are sensitive to changes in magnetic (relative permeability- dominant effect) and electrical (resistivity­ minor effect) properties, which in steels, vary with microstructure and temperature. EM sensors have been shown to have great potential for assessing steel microstructures (austenite to ferrite transformation or decarburisation). However, the influence of key microstructural parameters is not yet fully understood. This thesis presents a study of the effect of individual microstructural parameters on relative permeability and hence sensor output. In particular the ferrite grain size, pearlite interlamellae spacing, as-quenched martensite carbon content and phase balance were independently studied. The relative permeability of certain steel microstructures was determined using a finite element (FE) model fitted to experimental data. These values agreed with the literature and were used to predict the relative permeability of complex microstructures using an embedded microstructure FE model. Finally a case study on commercial steels was carried out, where the phase balance and tensile strength of dual phase steels were accurately predicted.

Interstitial free (IF) steel is used extensively throughout applications in the automotive, packaging and furniture industries due to its excellent formability and ductility. The manufacturing process ensures excellent material properties for subsequent forming processes are developed through the formation of a fine equi-axed grain structure and crystallographic texture. The annealing process improves the formability of the cold rolled IF sheet, whilst also reducing strength through the recovery and recrystallisation process. After the cold rolling process the grain structure is heavily deformed. During the recovery process the dislocation density is reduced through annihilation and redistribution of dislocations to form sub grains. During the recrystallisation process new grains nucleate and grow into new, strain free, grains. Magnetic properties of ferromagnetic material are known to be affected by microstructural phenomena such as dislocation density, grain boundaries, grain size and texture. It is therefore possible to monitor the recovery and recrystallisation processes using sensors that are responsive to changes in magnetic properties. The purpose of the research completed was to establish whether it would be possible to use electromagnetic (EM) sensors to monitor recovery and recrystallisation processes in-situ during heat treatment, such that EM sensors could then be deployed in a continuous annealing line.

High strength strip steels with a yield strength of ~700 MPa are used in the yellow goods industry, with recent trends requiring thicker strips with an improved toughness. Strip steel was produced to thicknesses of 10 – 17 mm, with thinner strips showing improved Charpy impact toughness compared to the thicker strip. It was found that all samples (different thicknesses and test orientations) showed splits on the fracture surfaces in the upper transition region. The hot rolled strip steels showed a predominantly fine-grained (~4 µm ECD) ferrite microstructure, with some coarse grain patches (grains ≥5 µm grouped together). No significant differences were noted for the grain size, yield stress, tensile strength, hardness and texture for the different compositions. Splits were characterised and produced successfully by low blow Charpy testing with acoustic emission (AE) sensors indicating the presence of splits which occurred in the upper transition region predominantly without main crack propagation. Analysis by x-ray tomography and sectioning was carried out showing splits propagate by transgranular cleavage, preferentially following coarse-grained regions, with the 10 mm strip producing deeper splits which was related to elongated grains (high aspect ratio). Therefore the presence of deep splits improved toughness through a reduction of the impact transition temperature, which was achieved by ensuring the presence of a bimodal grain structure, containing coarse elongated grain patches.

Poly(hydroxybutyrate) copolymers are sustainable and biodegradable, but they are known to exhibit secondary crystallisation, which severely reduces the ductility of these materials, thus hindering their current commercial use. Therefore, the main focus of this research was to explore a number of strategies to control the secondary crystallisation behaviour of two Poly(hydroxybutyrate) based copolymers. Blends of P(HB-co-HV)(3 wt % HV) with carbohydrate molecules of varying chain lengths were prepared by melt blending, characterised, and monitored over time to assess their capability to reduce secondary crystallisation. Additives were found to hinder the secondary crystallisation process, demonstrated by a reduction in the percentage change of mechanical properties as the concentration and chain length increased. The effect of storage temperature on the secondary crystallisation behaviour of P(HB-co-HHx)(33 % HHx) was also reported. Samples were stored at a range of storage temperatures and the effects on thermal, chemical and mechanical properties discussed. Increasing storage temperature caused the secondary process to occur to a greater extent, with greater increases in the melting temperature recorded in samples stored at 100 \(^∘\)C (128 \(^∘\)C - 135 \(^∘\)C) compared to samples stored at 7 \(^∘\)C (128 \(^∘\)C – 128 \(^∘\)C). Sub-melting point degradation of the material was also noted.

In this work, the deformation behaviour of Ti-6Al-4V micro-pillars was studied via compression tests at room temperature. Micro-pillars containing either the single α phase region or both α and β phases were prepared and either prismatic < a > slips or < c+a > slips were activated. Thin foils for TEM study were prepared using the FIB from the selected deformed pillars. STEM was used to image dislocations and their interaction with the α/β interfaces. Sample yield strength has been observed in all groups of micro-pillars and its relationship with the β phase, α/β interface, orientation and extrinsic sample size was discussed. The β phase is proved weaker than the α phase and the α/β interfaces behaviour as both the barriers and sources of dislocations. The magnitude of influence depending on the volume fraction, orientation and morphology of the β phase and the α/β interfaces. The work-hardening rate is related to the orientation of micro-pillars and the α/β interfaces in the micro-pillars. The strain bursts can be suppressed by the α/β interface and smaller sample size. The sample with smaller size expressed the larger strain burst in the stress-strain curves. The orientation influences the number and magnitude of strain bursts.