The validity of the theories of quantum corrections to the electrical conductivity, namely weak localization and enhanced electron-electron interaction, has been tested quantitatively in well characterized, free-electron-like Mg-Cu and Mg-Zn metallic glasses containing various amounts of Ag and Au through measurement of the electrical resistivity between 1.5K and 20K in magnetic fields up to 5.6T. It is found that the theories give an excellent description of the magnetoresistance at low fields, in both the weak and strong spin-orbit scattering limit but that at higher fields they break down. The electron spin-orbit scattering and dephasing rates have been deduced. Above 4K the dephasing rate is controlled by inelastic electron-phonon scattering, below 4K it saturates to a value consistent with a new model of dephasing of the quantum back scattering interference by ionic zero-point motion. The first direct measurement of the effect of superconductivity on the magnetoresistance in bulk amorphous metals is also presented. The temperature dependence of the resistivity between 1.5 and 6K is in qualitative but not quantitative agreement with the quantum correction theories.

In recent years, nanostructured and amorphous metals have become more prominent in scientific research. Their unique characteristics and their vastly superior mechanical properties have lead to emerging technical applications however the use of these materials is only in its infancy. To further growth the industrial applications for these advanced materials, joining processes capable of maintaining the unique aspects of the microstructure are required. / Joining of two beveled plates has been accomplished by use of the Cold Spray process. This process uses a converging diverging nozzle to accelerate micron sized powder towards a substrate. The strength of the resulting joint has been compared to freeforms made of powder of the same composition. No significant difference was observed in the mechanical properties between the freeforms and the weldments and examination of the fracture surface showed that the strength of the interface is higher than the strength of the freeform itself. / A weld repair experiment has also been performed where the ElectroSpark Deposition process (ESD) has successfully been used to transfer amorphous material from an electrode to an amorphous substrate without crystallizing either material. This result is of crucial importance as it signifies that these extremely expensive and heat sensitive materials can be repaired when in-service wear causes damage. This process, due to its inherently fast cooling rate, has also successfully been used to transform a crystalline AlCoCe alloy into an amorphous deposit atop both amorphous and crystalline substrates.

No description available.

No description available.

The superior qualities of ceramics such as high hardness, chemical stability and wear resistance make them promising tribological materials for machine elements, for example, pumps, bearings and heat engines. Ceramics are hard and brittle, therefore, machining such materials is time-consuming, difficult and expensive. A low cost machining process that can remove material rapidly while maintaining a good surface finish is required. The general aim of the studies presented is to find a correlation between the wear mechanisms and surface modifications induced by interactions of tribochemical nature and to identify the most effective combination of parameters involved in producing high material removal rates of ceramics. This study concentrates on silicon nitride, as this material has most potential for use in industry. The influence of lubricant chemistry on the friction and wear of silicon nitride is investigated using a ball-on -plate tribotester and a modified four-ball machine. A wide variety of liquid lubricants and additives supplied by Castrol International are tested to evaluate the role of the chemical characteristics of the lubricant on the friction and wear of silicon nitride. Surface and chemical analysis results reveal that by using different chemistries of liquids, the material removal rate and the surface finish of the silicon nitride can be significantly altered. The highest material removal rate is obtained when using the ester base fluid T80884 + 0.3% triethanol amine. This gives an increase by four fold when compared to the material removal rate obtained with the industrial reference slurry Kemet. The topography of the silicon nitride ball after the grinding test is found to be very smooth indicating that the predominant mechanism of material removal rate could be due to a tribochemical reaction occurring at the contact interface.

Ceramic/ceramic and ceramic/steel contacts under lubricated rolling conditions are studied. This work is of interest to ball bearing manufacturers as the use of ceramics in the design of these components has some advantages over traditional bearing-steel materials. Low density and increased stiffness are the mechanical properties which gas-turbine and machine tool manufacturers are most likely to realise. Much research over the past two decades on material structure, quality control and manufacturing techniques has produced a material which can seriously challenge bearing steel in ball-bearing design. This is especially the case for hybrid ball-bearings, ie ceramic balls with steel bearing races which are now used as standard components. The purpose of this study is to examine the rolling contact fatigue failure modes of ceramics. This study concentrates on silicon nitride as this material has most potential for use by industry. The primary reason for studying ceramic balls is because of interest in ball-bearing applications, hence a modified four-ball machine is employed which correctly models ball motions and precisely defines ball load. Experimental and theoretical kinematic analysis of ball motion during modified four ball machine tests is presented. The kinematic analysis reveals that in practice, lower ball tracking exists at high speeds. Test conditions of lubricated contacts under high compressive stress show delamination type failures. Delamination failures are classified in terms of propagation and initiation from scanning electron microscope observations. Residual stresses are measured on delaminated surfaces, which implies plastic deformation of the ceramic. Also, chemical analysis implies that disruption of silicon, nitrogen and oxygen levels may take place on delaminated surfaces. Experiments illustrating various fatigue failure modes using artificially pre-cracked ceramic balls in contact with a steel upper ball are presented.

In this thesis we study several different models which display glassy behaviour. Firstly, we investigate a simple, purely topological, cellular model for which the Hamiltonian is non-interacting but the dynamics are constrained. We find a non-thermodynamic transition to a glassy phase in which the energy fails to reach the equilibrium value below a characteristic temperature which is dependent on the cooling rate. This model involves activated processes and displays two-step relaxation in both the energy and the correlation functions; the latter also exhibit signs of aging. The relaxation time can be well-fitted at all temperatures by an offset Arrhenius law. Some predictions of Mode-coupling Theory are tested with some agreement found, but no convincing evidence that this description is the most fitting. By defining a suitable response function, we find that the equilibrium Fluctuation-Dissipation Theorem (FDT) is upheld for all but very short waiting-times, despite the fact that the system is not in equilibrium. This topological model is simplified to a hexagonally-based spin model, which also displays glassy behaviour, involves activated processes and exhibits two-step relaxation. This is a consequence of reaction-diffusion processes on two different time-scales, one temperature-independent and the other an exponential function of inverse temperature. We study two versions of this model, one with a single absorbing ground state, and the other with a highly degenerate ground state. These display qualitatively similar but quantitatively distinct macroscopic behaviour, and related but different microscopic behaviour. We extend this work to a square lattice, and find that the geometry of the lattice has a considerable impact on the behaviour, and to three dimensions, which provides support for the reaction-diffusion classification of the early behaviour. We find observable-dependent FDT plots; the observable can be chosen such that FDT is upheld for a region whilst the system is out of equilibrium — this observation is supported by some preliminary results for one-dimensional kinetically-constrained Ising chains.

Two methods of measuring the dielectric properties of materials - matched termination and coaxial line resonance- have been developed and used to study the properties of two groups of nitrogen ceramics, namely, silicon nitrides and oxynitride glasses. In these methods advantage is taken of the wide frequency range, from 500 MHz up to about 9 GHz, covered by a single apparatus - the General Radio slotted coaxial line. Previous measurements in this Department have indicated the difficulties in the determination of the loss tangent of low to medium loss samples (tan Ỏ ~ 10(^-3) - 10(^-2)). The two methods developed reduced these difficulties. The applicability of these methods was assessed using known materials including the high loss liquids water and chlorobenzene, medium loss solutions of chlorobenzene in cyclohexane and the low loss solids polymethyl methacrylate and polytetrafluoroethylene. The silicon nitride ceramics were in various degrees of nitridation given by the weight gain which ranged from 38% for partially-nitrided to a maximum of 63.2% for the fully-nitrided samples. The dielectric constant at 1 GHz increased from 4.51 for fully-nitrided to about 9.9 for the 38% weight gain samples. The fully-nitrided material has a loss factor of 7.6 x 10(^-3); this increased to 1.85 x l0(^-1)as the weight gain decreased to 38%. These values have been extrapolated to 'zero weight gain' and compared with pure silicon for which the dielectric constant is 11.7 and the loss factor approximately 0.2. The oxynitrides have dielectric constants between 6.5 and 7.5 depending on the cation present and the percentage of nitrogen substituted for oxygen in the glasses. Є' increased in the cation order Mg, Y, Ca and increased with increasing nitrogen substitution in each cation series. The loss factor, however, depends not only on these cation types but also on the other constituents of the glasses. The dielectric constant of both the silicon nitride ceramics and oxynitride glasses fitted the Jonscher universal law of dielectric response (Є' - Є) w(^n-1), where the exponent n - 1 for all the samples. Similarly, the loss factor for these materials showed a frequency dependence Є" w(^n-1), again with n approximately 1. This is a limiting case of dielectric behaviour corresponding to a frequency independent loss where most dipolar processes have been eliminated.

No description available.

No description available.