The anodic oxidation of titanium, in the majority of electrolytes (sulphuric, phosphoric, boric acid, etc.) results in the formation of a thin homogeneous dielectric oxide film of titanium dioxide (amorphous or crystalline) . At constant current the potential increases with increasing film thickness. If the potential is then maintained constant the ionic current decreases rapidly and the current becomes almost entirely electronic. In these circumstances, providing film breakdown does not occur, film growth practically ceases and oxygen evolution becomes the predominant electrode process. In chloride electrolytes, however, film growth proceeds up to a potential of 12-14V. If the potential is increased above this value film breakdown occurs and titanium ions pass into the electrolyte and precipitate as titanium dioxide. Corrosion of the titanium is confined to a relatively few, widely dispersed, small areas of the metal surface with the consequent formation of gross pits while the rest of the surface, which is coated with the initially formed anodic oxide, remains passive. During a study of the anodic oxidation of titanium in various electrolytes it has been found that formic acid (and formates) give rise to a phenomenon which does not appear to have been reported previously. Within a certain range of temperature and concentration of formic acid, film formation (at constant current) proceeds up to a voltage of 60 V, but as the potential increases above this value very fine pits are formed uniformly over the metal surface (micropitting) and simultaneously titanium oxide is precipitated from the electrolyte. The potential slowly increases to 120 V, indicating film growth, and then remains practically constant and if anodizing is continued the pits are propagated into the metal to a considerable depth. A microscopical examination of the metal surface during the initial stages of micropitting has revealed that different areas on the metal surface have different interference colours showing that the thickness of the surface oxide is not uniform. If, however, the temperature of the electrolyte is greater than 50 C or if the concentration is in the range of 70-96 per cent by vol formic acid, micropitting does not occur and a homogeneous dielectric oxide film is formed. It should be observed that at 25 C, at a concentration of 96 per cent formic acid, film formation can proceed to 400 V without film breakdown, whereas in boric acid the maximum formation voltage is only 250 V. At very high concentrations of formic acid (96.5 per cent to concentrated), a thick dark-yellow non-adherent oxide forms on the metal surface. The phenomenon of micropitting appears to be confined to the titanium-formic acid system as similar results could not be achieved with niobium or zirconium. The mechanism of the phenomenon described is discussed and some tentative theories have been proposed to explain the behaviour of titanium during anodic polarisation in formic acid at various concentrations and temperatures.

Some of the scientific and technological limitations in the field of austenitic thermochemical treatments involving the diffusion of nitrogen have been investigated. New'methods of accurately analysing the nitrogen concentration gradient in surface hardened components have been evolved, using standards based on a controlled nitriding process. As a result of this work, (a) emission spectrometry is now being used as a quality control tool in industrial austenitic carbonitriding and (b) quantitative electronprobe microanalysis for nitrogen is now a powerful tool for physical metallurgical research into nitrogen-rich iron-based systems. In addition, methods of accurately determining the residual ammonia level in heat treatment atmospheres have been evolved using both infra-red gas analysis and gas chromatography. Using the analytical methods just described, an investigation of the iron-carbon-nitrogen system was undertaken. This has resulted in the following information about the system: (a) the diffusion coefficient of nitrogen in austenite is compositionally dependent, in a similar manner to that for carbon in austenite, (b) using existing DY data for very low nitrogen levels, it is possible to estimate nitrogen concentration gradients in surface hardened components, by means of equations based on compositionally dependent diffusion coefficients, (c) Dy, at a fixed nitrogen level, increases with increase in base carbon level of iron-carbon alloys,(d) the iron-carbon-nitrogen ternary system at temperatures in the range 748°C to 845°C has been investigated and the austenite phase field partly delineated, (e) the mechanism of gas void formation in the austenitic nitriding of pure iron, which has in the past restricted its application, has been investigated and shown to be due to overlapping of advancing austenite interfaces in thin sheet material. This has important industrial implications in. ýi. a. relatively thick components can be surface hardened with hi&t: nitrogen levels without fear of blistering due to gas void formation.

The transition from localised corrosion to environmentally assisted cracking (EAC) is a phenomenon that remains poorly understood despite its importance to many industrial applications. Although the transition stage is often noted, very few detailed studies have been performed to quantify it. Some authors believe that the local environment is the fundamental contributor to EAC initiation in structural materials, while others consider the geometrical discontinuity introduced by a localised corrosion site such as a pit or crevice to be the significant factor. Assessment of the transition stage is often based on the competing growth rate concept originally proposed by Kondo [1], where it is assumed that crack initiation occurs at the base of a corrosion pit when; (i) the pit achieves a critical depth and hence critical stress intensity for crack initiation, and (ii) the crack growth rate exceeds the pit growth rate. This project aims to develop techniques and methodologies of measure that can be used to provide a more meaningful analysis of this transition effect. Two independent investigations of the transition phenomena have been presented which rely on the emerging field of X-ray tomography that can be used to study localised corrosion and environmentally assisted cracking events. InPart A of this dissertation, real time intergranular corrosion and stress corrosion crack growth rates have been quantified via synchrotron X-ray tomography in sensitised high carbon (0.056 wt.% C) type 304 stainless steel exposed to dilute potassium tetrathionate solutions. Crack growth rates, in both the long and short crack regimes, have also been measured using traditional fracture mechanics based tests. The transition from localised corrosion to EAC is discussed on the basis of Kondo's competing growth rate model. Results illustrate the unique power of the X-ray tomography technique in acquiring growth rates for localised corrosion and environmentally assisted cracking. In Part B of this dissertation, focus tube X-ray tomography is utilised for studying the threedimensional evolution of pitting and stress corrosion cracking in 3NiCrMo V steam turbine disc steel. Results indicate that pit and crack depth are not necessarily the same at the point of transition, and that cracks nucleate more readily at the pit side, just below the sample surface, in contrast with the assumption of existing deterministic (mechanistically based) models [2-6] developed to address SCC initiation and propagation in low-pressure steam turbines. With this in mind it is suggested that the existing pit-to-crack models based on Kondo's criteria should be reassessed and modified to incorporate empirical aspects specific to the system of interest.

Surface melting and alloying of commercial complex aluminium bronze CA104 has been carried out using an electron beam. In addition, other aluminium bronzes - Ampco 18, Ampco 25, CMA1 and AB2 have also been surface melted. The structural changes have been studied using optical microscopy, scanning and transmission electron microscopy, energy dispersive X-ray analysis and X-ray diffraction analysis. The hardening response to the treatments and the hardening mechanisms have been investigated. Tempering of the surface melted CA104 has also been carried out systematically to investigate the responses of the structure and hardness to subsequent heat treatment. The wear behaviour of the materials has been studied in both dry and lubricated rolling-sliding contact against hardened En19 steel and under two-body abrasive conditions against silicon carbide grinding papers. The electrochemical corrosion properties in simulated cooling water and 3.5% NaCI solution have been studied using AC impedance and DC polarisation techniques. Experimental results show that the surface layers have been transformed to martensitea fter surfacem elting due to rapid solidification and fast subsequenct ooling. The martensite in the surface melted CA104 has a 3R structure with globular precipitates dispersed in it, which are based on (Ni, Fe)A1 with a B2 structure. On subsequent tempering treatment, these precipitates grow steadily with increasing temperature and time and the martensite gradually transforms to copper-based solid solution. The hardness has been increased by surface melting as a result of the martensitic transformation and precipitation hardening. During tempering, further precipitation hardening occurs and the hardness shows a peak plotted against temperature. Surface alloying with aluminium can increase the hardness beyond that achievable by surface melting. The tungsten carbide retained after surface alloying further strengthens the alloy. The wear rate and the friction coefficient in both the dry and lubricated rolling sliding wear have been reduced by surface melting. In the dry wear test, adhesive and delamination wear are the main wear mechanisms while it is abrasive wear that dominates the lubricated wear. The abrasion resistance against silicon carbide has been increased by both surface melting and alloying, with surface alloying much more effective. The best abrasionr esistanceis obtainedf rom the material alloyed with WC particles. The results also show that the electrochemical corrosion properties are improved by surfacem elting in both electrolytesu sed. The charget ransfer resistanceis increaseda nd the corrosion current is reduced. In the simulated cooling water, the corrosion of the materials during anodic polarisation follows three steps: the formation of cuprous oxide layer on the surface; the damage and destruction of the oxide layer; the pitting and the formation of the main corrosion product, CuCl.

This study examinest he creepb ehaviour of austenitic steels under service temperatures, to determine the effect of creep on material performance. Nuclear power plant components are in regular use at temperatures greater than 450°C, where creep deformation plays a dominant role in limiting the lifetime of the material. The prime aim of this study was to characterise the effect of load-reductions on the creep behaviour of austenitic steels (AlSI type 316H). In-service materials seldom operate at a constant load and/or temperature. The supply demand, maintenance operations, refuelling, etc. will result in large variation of load and temperature acting on the material. Experiments where load/temperature removals during a creep test were therefore conducted. These unloading procedures result in material recovery of the accumulated creep strain (anelasticity). This phenomenon will influence the material properties such as creep life and ductilities. Creep life was found to increase by 2-3 times whereas creep ductilities decreased by 50% when compared to steady-load creep data under identical conditions. The occurrence of anelasticity suggested the presence of a material backstress. The origin and evolution of this internal stress was investigated using neutron diffraction and TEM microscopy. Lattice strain measurements were conducted in-situ using neutron diffraction during a creep test which consisted of load/unload cycles. Experimental results suggest that creep strain is equivalent to plastic strain at a granular level. The data also shows intergranular micro-stressesa re introduced into the material by primary creep. Anisotropic behaviour of the individual crystal planes results in formation of tensile and compressive intergranular stressesin individual grain families. Residual compressives tressesd rive this anelastic deformation. TEM examinations of samples stopped during the unload show changes in dislocation and precipitate morphologies during the plastic strain recovery phase. Evidence of a changing dislocation substructure during the load-reduction period was found. Examinations have also shown carbide densities change during the unload. Pipe diffusion is a possible mechanism which can be used to explain this occurrence. The changing precipitate and dislocation state will influence the strengthening mechanisms, which in-turn will affect the deformation characteristics. These microstructural observations were introduced into a damage mechanics model. Predictions of material behaviour using this model have shown good agreement with experimental data. Outcomes of this project, have established that changes in creep deformation mechanisms will greatly influence material properties. Deformation history of the material will affect the intergranular stress state which in turn will affect the elastic and plastic response of the material. The effect of plastic strain history must be considered and incorporated accounted in any design and assessment procedure.

The constitution of the gold-rich region of the gold—cadmium diagram has been etudied by means of thermal analysis, optical microscopy and Debye-Scherrer X-ray films of high-temperature and quenched specimens. Five phases have been identified, α face-centred cubic, α₂ close-packed hexagonal, β' ordered body-centred cubic as well as two phases at low temperatures, α₁ face-centred tetragonal and α' face-centred cubic. The liquidus and solidus relationships have been found for the α, α₂ and β-phases by thermal analysis, the results being in good agreement with the X-ray and microscopical results. A eutectic horizontal representing the reaction α₂ + β' ⇌ liquid is found at 626.5°C. The β'-phase maximum is only just detectable, rising to 629°C. From the X-ray evidence a peritectic horizontal, corresponding to the reaction α + liquid ⇌ α₂, is believed to exist between 633°C and the eutectic temperature; the thermal arrests place it at 627°C. The α₂(α₂ + β') and (α₂ + β')β boundaries have been determined microscopically. Only a few experimental points have been obtained for the α + α₂ boundaries but the high-temperature X-ray films indicate that the boundaries of previous workers are in need of revision. The maximum solid solubility of cadmium in gold is 34 atomic % at 620°C to 630°C. The β'-phase narrows in composition range as temperature falls. The high-temperature X-ray films showed superlattice lines up to 624°C and an estimate of the degree of ordering at 620°C suggests the β-phase to be about 90% ordered at this temperature. The α₁-phase at 23 atomic % cadmium is disordered face- centred tetragonal when photographed at elevated temperatures. The absence of superlattice lines is contrary to previous work which describes the unit cell in quenched α₁, specimens as ordered (an anti-phase domain structure based on four fundamental tetragonal cells). It is conjectured that quenched α₁ specimens may be ordered but as temperature is increased the long-range order is broken down, producing small domains each highly ordered. At about 25 atomic % cadmium the structure is face- centred cubic. A new phase α' is proposed at this composition. A tentative diagram of the Au₃Cd region, linking the four phases α, α₁, α' and α₂ is suggested with peritectoid horizontals at 412°C and 407°C. Factors affecting the behaviour of the liquidus and solidus of the gold-cadmium system are revealed by comparing with other systems of copper, silver and gold. The comparisons show the influence of the electron concentration, modified by lattice distortions and the electrochemical factor. The α₂-phase (close-packed hexagonal) is shown to contain numerous stacking faults; this may be a consequence of the α(face-centred cubic) and α₂-phases having nearly equal free energies. The occurrence of the 3/2 electron compounds in gold is discussed and suggestions are given for their bearing on maximum solid solubilities in gold.

No description available.

Ni-based superalloys are primarily used in the manufacturing of critical gas turbine components, such as rotating blades. The drive for increased efficiency has led to a continuous rise in engine operating temperatures, and therefore these components are exposed to conditions that can ultimately compromise their mechanical integrity and therefore limit their service life. There is a desire to extend the component's life and also to avoid any premature failures, and therefore understanding of the material's stability and properties throughout component life is increasingly important. The identification of features in the microstructure of Ni-based superalloys which change systematically with time and temperature may allow it to be used as a 'time-temperature' recorder. This could enable a determination of the effective temperature a component will have experienced, and in conjunction with known values of operating stress, an estimate of the remaining service life can be made.

Alloys of the elements Mo, Pd and Rh have been studied since complexes of these with Ru and Tc are found as metallic inclusions in irradiated oxide fuels, and it is considered that knowledge of the constitution of these multicomponent alloys may assist in understanding the history and character of such fuel. Samples were prepared either by arc melting or rapid solidification from Mo-Pd and Mo-Rh and ternary Mo-Pd-Rh systems. Rapid solidification was employed in an attempt to overcome the problem of slow diffusion in these systems particularly at lower temperatures. The usefulness of rapidly solidified materials in the constitutional investigation of these particular systems has been proved. Constitutional studies were made of binaries Mo-Pd and Mo-Rh using optical microscopy SEM analyses and X-ray diffraction methods. The phase boundaries in both binaries at temperatures ranging from 881 to 1100 °C were established satisfactorily. No intermediate phase was found at the temperatures of interest. Differences from published diagrams are suggested. An isothermal section of Mo-Pd-Rh ternary system was established at 1100 °C by constitutional studies employing the same techniques as utilised in the binary investigations. The section involves only the bcc fcc and hcp phases.T he bcc solid solution is restricted to the Mo comer while the fcc and hcp phases are stable over wide ranges of compositions. A three phase region (bcc + hcp + fcc ) is located near the Mo-Pd edge binary. The phase boundary composition limits found in the relevant binaries agree very well with those of the two-phase fields in the ternary . Thermodynamic coefficients for the phases in the relevant binaries Mo-Pd , Mo-Rh and Pd-Rh were derived by the assessment of available phase diagram and thermodynamic data using the Lukas program ; the resulting coefficients were input to reestablish the binaries. Very good agreement was found with the experimental phase diagrams and the comparison of integral molar free energies and the partial free energies of Mo with experimental values was very satisfactory. The coefficients established in the binary computations were combined with the lattice stability values of components and the ideal term to compute ternary isothermal sections-No ternary interaction terms were involved. The isothermals between 1373 - 2673 K were constructed. The isothermal calculated at 1373 K matches very well with the constitutionally established one except for the prediction of a more restricted hcp phase. The location of the three phase region is the same on both the experimental and calculated isothermal-Solidus , liquidus and secondary surfaces counters were constructed satisfactorily. Some attempts have been made to improve the 1373K isothermal section calculated employing the relevant binary coefficients only , by introducing ternary interaction terms. But these attempts were ineffective because of the lack of ternary thermodynamic values and insufficient ternary phase diagram data.

This thesis presents a microstructural evaluation of the gamma titanium aluminide alloy Ti-48Al-2Mn-2Nb (at.%) joined by diffusion bonding, electron beam welding and linear friction welding. The aim of the project was to evaluate a number of different joining techniques in order to gain an understanding of the type of microstructures generated and, ultimately, the advantages and disadvantages of each technique. The diffusion bonding behaviour of the intermetallic compound Ti-48Al-2Mn-2Nb (at.%) was studied as a function of various parameters including; starting microstructure, temperature, surface finish and time. During diffusion bonding, small y-based grains nucleated at the interface and grew into the matrix forming a 'double necklace' grain structure. Backscattered electron imaging (BSEI) and transmission electron microscopy (TEM) identified the presence of oc 2 -based particles, both at along the original boundary and between the newly formed interfacial y grains and the matrix. The formation of a 2 particles was believed to be related to the redistribution of a 2 -stabilising elements, notably oxygen, during bonding. Post bond heat treatment in the a phase field («1400°C) was found to be an effective way of replacing the original double necklace grain structure with a fully lamellar structure that exhibited good grain growth across the original boundary. The electron beam weldability of the alloy was studied with particular emphasis on reducing cold cracking associated with the weldment cooling through the ductile to brittle transition temperature (DBTT) at approximately 700°C. It was found that the amount of cold cracking could be reduced and eliminated by reducing the cooling rate of the weldment through the DBTT via in-situ electron beam heating. Tensile testing of the welds showed that although the welds appeared to be crack-free they were considerably weaker than the matrix, this suggests the welds contained a high degree of residual stress. The application of different preheats to the samples produced a range of cooling rates within the weldment which proved ideal for studying the various transformation products of the high temperature a phase. The retained ot 2 , massively transformed y and wheatsheaf / feathery structures were all studied in detail by TEM, with particular emphasis on the oc-»massive y transformation. The analysis culminated with the proposal of a qualitative CCT curve for the various transformation products from the a phase. Analysis of the a -> massive y transformation suggests the massive phase developed in the parent a from a coherent, grain boundary nucleus which subsequently grew into an adjacent grain by movement of an incoherent interface. A feathery structure was also observed in the welds. This structure consisted of lamellar colonies that appeared to gradually Tan out' at angles between 1-30°. No conclusive results were obtained to explain how the feathery structure developed, although a number of mechanisms were proposed including; the growth of lamellae on distorted a planes, non-coherent growth and the growth of colonies via an intermediate massive transformation. Examination of linear friction welds identified the presence of three different types of microstructure. The high deformation associated with the forging cycle produced extensive dynamic recrystallisation at the weld centre. The microstructure within this region was predominantly fine grained equiaxed y (50-100 nm). Away from the weld centre the matrix appeared severely deformed, causing extensive twinning and bending of the original lamellar structure. In some cases, regions were observed containing a high density of oxide - nitride based particles indicating the welding parameters were insufficient to cause complete extrusion of flash material from the weld.