Bearings as a machine element are designed to facilitate relative motion. The expected life span for bearings in every each application, depending on operation conditions, is different. Whilst numerous efforts have been focused on increasing the bearing life-time, over the last few decades, premature failures have been reported especially in automotive and wind turbine gearbox bearings i.e.e.g. hydrogen embrittlement and micropitting surface fatigue. Furthermore, Zinc Dialkyl Dithio-Phosphate (ZDDP) as an anti-wear additive, which is crucial to supress destructive wear, accelerates micropitting. Moreover, water contamination is common in wind turbine gearboxes especially in off- and near-shore turbines which can further promote micropitting occurrence. Therefore, it is significantly important to understand the influence of ZDDP and water on micropitting and investigate the effect of low Sulphated Ash, Phosphorus and Sulphur (SAPS) additives capable of suppressing the micropitting-enhancing behaviour of ZDDP. In this study a modified PCS instrument MicroPitting Rig (MPR) is utilised to perform micropitting experiments. The influence of ZDDP on micropitting is investigated over a range of different contact cycles and the tribochemical phenomena involved in the micropitting is addressed using X-ray Photoelectron Spectroscopy (XPS) and Energy-Dispersive X-ray (EDX) in a Transmission Electron Microscope (TEM). Employing wWhite lLight iInterferometry (WLI) a promising procedure for micropitting mapping is implemented in order to achieve a micropitting surface area in the wear scar. Experiments are undertaken in order to understand the impact of water as a contaminant on micropitting where ZDDP was used in the lubricant formulation. The influence of free and dissolved water on tribocorrosive micropitting is differentiated and clarified. A mechanistic understanding of effect of water and relative-humidity on micropitting is elucidated using XPS. A range of organic Nitrogen-Containing-Additives (NCA) with a friction reduction capability is proposed to diminish the micropitting-enhancing behaviour of ZDDP and evaluated using MPR. The affinity of the functional group of NCAs, nitrogenN, to ZDDP which brings about a delay in the tribofilm formation, their friction modification properties, changes in the tribofilm composition and being an additive with no SAPS content were the rationale behind employing NCAs in combination with ZDDP. The tribofilm formation and frictional properties of the different lubricant formulations are examined using a mMini tTraction mMachine (MTM-SLIM). To gain a full understatingunderstanding of NCAs behaviours in combination with ZDDP, tribofilms are studied using XPS and EDX-TEM and aAtomic fForce mMicroscopy (AFM). The results in this study showed that interaction of ZDDP with the steel surface under severe rolling-sliding contacts is the a influential derivativeprofound factor to induce micropitting. While dissolved-waterdissolved water increases micropitting nucleation and expands the nucleation across the wear scar, in free-waterfree water condition micropitting appearance is suppressed due to a dominant action of the mild wear. The chemical structure of NCAs significantly influences the tribological results. A careful selection of NCAs with a tailored concentration successfully reduces micropitting and protects the surface from wear simultaneously and thus NCAs are desirable additives in rolling-element bearing applications under certain conditions.

The high prevalence of osteolysis and aseptic loosening associated with the wear particles of conventional metal-on-UHMWPE (MoP) total hip replacements (THRs), and concerns over the release of metal wear particles and metal ions around the body from metal-on-metal (MoM) THRs, led to the development of alternative ceramic-on-ceramic (CoC) THRs. CoC bearings are of great interest due to their superior wear properties, compared to MoP and MoM bearings. Historically, ceramic THRs had a reputation for fracture, and recent issues have centred around surgical positioning and squeaking. The development of improved manufacturing methods allowed major improvements of ceramics which led to the introduction of composite ceramics for example, zirconia-toughened, platelet reinforced alumina or ZTA, otherwise commercially known as BIOLOX® Delta. The wear performance of composite CoC THRs such as those using BIOLOX® Delta has been extensively investigated, however no studies have reported the combined characteristics and biocompatibility of the wear debris generated from these bearings. Understanding wear particle characteristics and their biological activity is an essential step in the pre-clinical testing of joint replacements. However, currently for composite ZTA CoC bearings there is a lack of relevant studies, due to difficulties in generating high volumes of clinically-relevant ceramic wear debris in vitro, in addition current particle isolation methods are not sensitive enough to reliably isolate wear particles from hip simulator lubricants, due to the inherent low wear rates of the composite ZTA ceramics. Hence, the particles have not been systematically characterised and therefore little is known about their size, morphology and biological responses. Therefore, the aim of this study was to investigate the characteristics and biological activity of wear particles generated from composite BIOLOX® Delta ZTA CoC THRs. This study developed a two-step particle isolation method and subsequently applied it to hip simulator lubricants for the isolation of composite ceramic wear particles generated from BIOLOX® Delta CoC bearings tested under edge loading conditions. The high sensitivity of this new particle isolation method coupled with its effective removal of protein, allowed the successful recovery and characterisation of very low volumes of both micro and nano-scale wear particles, generated from composite ZTA CoC hip replacements for the first time. The recovered wear particles demonstrated a bimodal size range, which has been previously reported for wear particles generated from alumina ceramic hip replacements. A comprehensive evaluation of the biological impact of commercially-obtained composite BIOLOX® Delta ZTA ceramic model particles and clinically-relevant composite BIOLOX® Delta ZTA ceramic wear particles was investigated in terms of cytotoxicity, inflammation, genotoxicity and oxidative stress. The clinically-relevant composite ZTA ceramic wear particles were generated in water lubricant using a hip simulator under severe edge loading conditions. The biological impact of the ceramic particles was assessed using L929 fibroblast cells and peripheral blood mononuclear cells (PBMNCs) isolated from healthy human donors. Both the model and clinically-relevant BIOLOX® Delta ceramic wear particles demonstrated significant reduction in the viability of L929 fibroblast cells at very high doses (500μm3 of particles per cell), however no cytotoxic effects were observed at the lower clinically-relevant doses (0.5-0.05μm3 per cell). The BIOLOX® Delta ZTA ceramic model particles failed to stimulate an inflammatory response in terms of TNF-α release and did not cause any significant DNA damage or production of reactive oxygen species (oxidative stress) in PBMNCs from all donors. However, high doses (50μm3 per cell) of clinically-relevant BIOLOX® Delta ZTA ceramic wear particles caused significantly elevated levels of TNF-α release from PBMNCs. But, there were no significant effects in terms of DNA damage and oxidative stress in PBMNCs from all donors. This study demonstrated that there was a threshold volume of clinically-relevant ceramic wear particles required to stimulate significant TNF-α release from PBMNCs. However, these doses were not clinically-relevant and highly unlikely to occur in vivo due to the extremely low wear rates of CoC bearings. This comprehensive study indicated that composite ZTA Delta ceramic hip replacements had a low biological impact, which may enhance long-term clinical performance. The results from this study are only relevant for BIOLOX® Delta ZTA ceramics and not other manufacturers ceramics.

Current methods to diagnose bone diseases like avascular necrosis (AVN) are subjective and a reliable assessment of the fracture risk is not available. A diagnostic fracture prediction tool would aid clinical diagnosis, anticipate disease progression and help with the planning of subsequent interventions. The strength of bones, including the femur, can be calculated using structural mechanics with a view to ascertaining fracture risk. The aim of this thesis was to develop and validate a fracture prediction method based tomographic imaging and beam theory. In-vitro disease models were created from additive manufacturing, explanted porcine and human femoral heads. The disease models contained a simulated lesion that was either lateral or medial to the fovea to analyse the effects of different lesion positions and to verify the ability of the developed fracture prediction tool. Current classification methods rely on the identification of the lesion volume and location to quantify the fracture risk, an approach that is purely based on geometrical information. The fracture prediction method based on structural stiffness also considered material properties which potentially added predictive capability. The tool was subsequently validated by predicting the fracture risk of femoral heads from AVN patients to demonstrate the ability to identify necrotic lesions that were likely to progress to fracture. The predicted fracture risk was compared to the current diagnostic gold standard to diagnose AVN. The beam tool was also compared against another novel fracture prediction tool based on FEA to identify possible advantages of beam theory. The verification tests confirmed that samples with a lesion in the weight bearing area were statistically more likely to fracture at a low load. A low fracture load meant a high fracture risk. However the experimental fracture load of porcine and human femoral heads, even among samples with similar lesions, showed variations indicating that lesion volume and location were not good predictors of fracture risk alone. There was a good correlation between the predicted fracture risk and in-vitro fracture loads of the human femoral head disease model indicating that the developed tool was able to objectively predict the fracture risk. The beam tool had similar good predictive capabilities as current diagnostic methods and fracture prediction methods based on FEA. An objective in-vivo analysis of the mechanical fracture risk helps identifying patients whose disease is at risk of progressing, as well as stratifying surgical interventions.

The IT (Information Technology) infrastructure power consumption constitutes a large portion of global electricity consumption and a large proportion of this energy is to maintain an acceptable thermal environment for the IT equipment. Therefore, it is important to understand and improve the thermal and energy management of data centres for lower cost and higher sustainability. Toward this goal, Direct Contact Liquid Cooled (DCLC) servers, where liquid loop heat exchangers are attached to the CPU, were proposed to study the use of chiller-less energy efficient data centre. Thirty Sun Fire V20z servers in a data centre rack have their CPUs water cooled with the remaining components air cooled, together with a rear door heat exchanger to capture this air heat flow. The heat generated by the servers is ultimately transferred to the environment using an Air Handling Unit (AHU). The AHU was fitted with a water spray system to increase the heat transfer capacity. The designed DCLC system was tested and characterised in terms of power consumption and thermal performance. The design successfully provided stable inlet coolant temperature (±1°C) to the IT despite the variation in the IT workload and environmental conditions. Activating the spray reduced the thermal resistance of the AHU heat exchanger (HE) by 50%. However, the power consumption and pressure drop across the HE was increased. The flow distribution and the coolant pumping configurations of centralised (where the coolant is pumped by two central pumps connected in series) and distributed (where small pumps inside the servers are activated) was investigated. The EPANET software was used to analyse the flow and showed that the servers in the top of the rack receive a higher flow rate (by approximately 30%) than the servers in the bottom of the rack. This resulted in a variation in the CPU temperatures of different servers. Optimisation analysis proposed increasing the manifolds size to improve the flow rate and reduce the flow maldistribution. In the distributed pumping case, the CPUs temperature showed to be 2°C higher compared with the central pumping case for the high IT workload. The rack inlet temperature was tested in the range of the ASHRAE W4 envelope in terms of CPU temperatures, power consumption and computational efficiency. Increasing the coolant inlet temperature resulted in high energy saving in the AHU, while the rack energy consumption increases marginally in idle operation and considerably more in high IT workloads. This results in an improvement in the energy effectiveness of 17% but a deterioration in the computational efficiency of 4%. Finally, a parallel study was carried out to investigate the droplet evaporation over heated surfaces which ultimately be used in studying sprays in the AHU or in direct on chip cooling via evaporation. A novel experimental design was proposed to track the lifetime of any droplet size that span the surface tension to gravitydominated regimes. A theoretical model was also proposed to predict the droplet lifetime based on the initial contact angle, contact radius and the receding contact angle. The model predicted the droplet evaporation over hydrophobic surfaces with good accuracy of an error less than 4% while under estimated the evaporation with hydrophilic surfaces.

A novel low-cost microfabrication technique for manufacturing RF MEMS switches and varactors is proposed. The fabrication process entails laser microstructuring and non-clean room micro-lithography standard wet bench techniques. An optimized laser microstructuring technique was employed to fabricate the MEMS component members and masks with readily available materials that include, Aluminum foils, sheets, and copper clad PCB boards. The non-clean room micro-lithography process was optimized to make for the patterning of the MEMS dielectric and bridge support layers, which were derived from deposits of negative-tone photosensitive epoxy-based polymers, SU-8 resins (glycidyl-ether-bisphenol-A novolac) and photoacid activated ADEXTM dry films. The novel microfabrication technique offers comparatively reasonably yields without intensive cleanroom manufacturing techniques and their associated equipment and processing costs. It is an optimized hybrid rapid prototyping manufacturing process that makes for a reduction in build cycles while ensuring good turnarounds. The techniques are characterized by analysing each contributing technology and dependent parameters: laser structuring, lithography and spin coating and thin film emboss. They are developed for planar substrates and can be modified to suit specific work material for optimized outcomes. The optimized laser structuring process offers ablation for pitches as small as 75 μm (track width of 50 μm and gap 25 μm), with a deviation of 3.5 % in the structured vector’s dimensions relative to design. The lithography process also developed for planar and microchannel applications makes for the realization of highly resolved patterned deposits of the SU-8 resin and the laminated ADEXTM polymer from 1 μm to 6 μm and with an accuracy ±0.2 μm. The complete micro-fabrication technique fabrication techniques are demonstrated by realizing test structures consisting of RF MEMS switches and varactors on FR4 substrates. Both MEMS structures and FR4 substrate were integrated by employing the micro-patterned polymers, developed from dry-film ADEXTM and SU-8 deposits, to make for a functional composite assembly. Average fabrication yield up to 60 % was achieved, calculated from ten fabrication attempts. The RF measurement results show that the RF MEMS devices fabricated by using the novel micro-fabrication process have good figure-of-merits, at much lower overall fabrication costs, as compared to the devices fabricated by conventional cleanroom process, enabling it to be used as a very good micro-fabrication process for cost-effective rapid prototyping of MEMS.

Faecal incontinence is a highly debilitating condition, prevalent across the population worldwide. Coupled with a large unmet need for clinically viable treatment options, a paucity of research into the biomechanics of continence inhibits the development of treatments which address multi-faceted challenges associated with the condition. Consequently, this thesis presents a method to fabricate, measure and control a physical simulation of the human defecatory system to investigate individual and combined effects of anorectal angle and sphincter pressure on continence. To illustrate the capabilities and clinical relevance of the work, the influence of a passive-assistive artificial anal sphincter (FENIX) is evaluated. A model rectum and associated soft tissues, based on geometry from an anonymised computerised tomography dataset, was fabricated from silicone and showed behavioural realism in terms of their morphology to the biological system and ex-vivo tissue. Simulated stool matter with similar rheological properties to human faeces was developed. Instrumentation and control hardware were used to regulate injection of simulated stool into the system, define the anorectal angle and monitor stool flow rate, intra-rectal pressure, anal canal pressure and puborectalis force. Studies were conducted to examine the response of anorectal angles at 80°, 90° and 100° with simulated stool. Tests were then repeated with the inclusion of a FENIX device. Stool leakage was reduced as the anorectal angle became more acute. Conversely, intra-rectal pressure increased. Overall inclusion of the FENIX reduced faecal leakage, while combined effects of the FENIX and an acute anorectal angle showed the greatest resistance to faecal leakage. These data demonstrate that the anorectal angle and sphincter pressure are fundamental in maintaining continence. Furthermore it demonstrates that use of the FENIX can increase resistance to faecal leakage and reduce anorectal angles required to maintain continence. The physical simulation of the defecatory system is an insightful tool to better understand, in a quantitative manner, the effects of the anorectal angle and sphincter pressure on continence. This work is valuable in helping improve our understanding of the physical behaviour of the continence mechanism and facilitating improved technologies to treat severe faecal incontinence.

Adverse biological reactions to orthopaedic wear particles have led to the investigation of silicon nitride (SiN) coatings for implants, which may improve the biocompatibility of hip replacements. However, the analysis of wear particles is impeded by suboptimal methods to isolate wear particles from tissue. The aim of this research was to develop an improved method to isolate wear particles from tissue samples and to evaluate the in vivo biocompatibility of SiN particles. The method involved digestion of tissue samples with papain (1.56 mg.ml-1) and proteinase K (1 mg.ml-1). Samples were then subjected to density gradient ultracentrifugation using sodium polytungstate (SPT) and washing through further rounds of ultracentrifugation. The isolation method was validated by subjecting tissue samples doped with 0.0025 mm3 of silicon nitride (SiN), or 0.025 mm3 of cobalt chromium (CoCr) or titanium particles to the isolation procedure. Particles were filtered and analysed using scanning electron microscopy (SEM), elemental analysis and image analysis software before and after isolation. The method had no significant effect on SiN, CoCr or titanium particle geometries. Volumes of 0.018 mm3 of SiN, CoCr or titanium particles were injected into rat stifle joints. After seven days animals were euthanised, and the stifle joints were formalin-fixed. Stifle joint tissues were harvested and subjected to particle isolation using the validated protocol. No significant changes to SiN or CoCr particle geometries occurred. Titanium particles were similar in size, aspect ratio and circularity before and after isolation, though particle numbers were insufficient for statistical analysis. Histology of stifle joint tissues demonstrated that all three particle types caused a degree of inflammation; in the SiN group, numbers of macrophages but not lymphocytes were significantly elevated. Necrosis and increases in synovial thickness were observed in the CoCr group, but were absent in the SiN and titanium groups. Rabbits were implanted with CoCr stifle joint prostheses, which were either uncoated or coated with SiN. After 12 weeks the rabbits were euthanised. Tissue samples from the joints were formalin-fixed and subjected to the particle isolation process. SiN coating particles, CoCr substrate particles, and zirconium dioxide particles from bone cement were isolated from the samples. Due to apparent coating failure, the coating particles were 1 - 100 μm in size and were columnar or shard-like in morphology. Histological analysis of rabbit samples from the coated group demonstrated a moderate macrophage infiltrate including giant cells, with few lymphocytes and no necrosis. Overall, SiN particles were relatively biocompatible, particularly in comparison to CoCr particles. However, analysis of a functional SiN coating must be carried out over a longer timeframe to enable chronic reactions to SiN coating particles to be fully evaluated.

Most of point-of-care diagnostics and lab-on-chip devices that do on-chip sample preparation require active fluid actuation. In a laboratory setting, this is done via bulky benchtop equipment such as syringe pumps, peristaltic pumps and pressure systems. However, integration of a pumping unit onto the device allows for increased portability and decreased footprint of the device. Although there are multiple examples of realised micropumps based on different technologies, no one solution offers a combination of small footprint, low costs, scalable manufacturing and high performance required for point-of-care devices. Surface acoustic wave (SAW)-based micropumps are an exciting alternative to the current micropump systems due their small footprint and simplicity of manufacturing, yet many of the SAW micropumps presented to date suffer from poor performance and/or utilisation of open channels, which can be a problem regarding contamination. The SAW micropump demonstrated here uses a novel planar design and SAW scattering effects to significantly improve the pump performance and maintain closed channels, which is a pre-requisite for point-of-care applications. This thesis evaluates the fabrication of SAW devices and microfluidic channels using soft lithography. After evaluating the SAW device design concerning electrical characteristics both experimentally and theoretically, the first iteration of SAW micropumps utilising SAW momentum along the piezoelectric substrate is presented and characterised in terms of fluid flow velocity profiles and volume flow rates produced. Subsequently, a concept of a more efficient SAW micropump is presented based on out of the plane interaction between SAW and liquid. To fully utilise this interaction a protocol on the development of 3D microfluidic channels is introduced followed by a discussion on SAW-liquid coupling setting the scene for a demonstration of efficient and closed-loop SAW micropump that delivers pressure gradients up to an order of magnitude higher than the best to-date reported values at a similar input power levels. Finally, the newly developed pump is utilised in an on-chip flow cytometer to showcase the advanced flow manipulations, showing the potential applications of the SAW micropump beyond simple fluid actuation.

Resistive memory is widely considered as a promising non-volatile memory to address the demands for high-density data storage, low power consumption, augment the performance of current transistor-based memories or even replace current transistor-based memories. Main advantages of resistive memory include simple Metal/Insulator/Metal device structure, low switching voltage, fast switching speed, and long data retention. Material properties of the insulating layer play important roles in the overall performance of resistive memory. Among a range of insulator materials of resistive memories that have been reported in the literature thus far, Silicon Carbide (SiC) has shown great promise as the insulating layer which leads to resistive memories with desirable performance including large ON/OFF ratios, excellent data retention, and CMOS compatibility in device fabrication. However, there are still many challenges to be solved in the resistive memories using SiC, especially amorphous (a)-SiC as the insulating layer to be superior to other resistive memories. One of these challenges is to reduce the forming voltage which could affect the power consumption and complexity of the peripheral power-supply circuit. Another challenge is to achieving device structure exclusively using native CMOS back-end-of-line materials which would enable low fabrication cost and low development time to embed a-SiC based resistive memories in the CMOS back-end-of-line layer. Moreover, the existing Electrochemical metallisation (ECM) mechanism cannot precisely predict the switching voltage nor resistance state of resistive memories, and there is a lack of knowledge on how the material properties of the insulating layer affect resistive-switching performance and mechanisms. Further exploration of the resistiveswitching characteristics to improve the understanding of switching mechanism and influence of material and electrical properties of the insulating layer on resistive-switching characteristics are needed from a scientific point of view. This thesis focuses on addressing all the challenges above, highlights the influence of insulator material choice on the performance of resistive memories using SiC as the insulating layer. Amorphous silicon carbide (a-SiC), Cu embedded a-SiC (a-SiC:Cu), CMOS back-end-of-line dielectrics (a-Si(O)C:H), and crystalline SiC (c-SiC) are used as the insulating layer of resistive memories in this thesis. The material and electrical properties of these insulator materials are characterised. Metal/Insulator/Metal resistive memories using these insulator materials as the insulating layer are fabricated and the resistive-switching characteristics of these resistive memories are studied. Ultrahigh ON/OFF ratios up to 109 which enables fast and reliable detection of the states, are achieved. Forming voltage and SET voltage are reduced and endurance is improved by embedding Cu nanoparticles in the a-SiC insulating layer. Non-volatile resistive-switching is observed on resistive memories using exclusively native CMOS back-end-of-line materials including Cu, W, a-SiC:H, aSiOC:H, and a-SiCO:H. The influence of material and electrical properties of the insulating layer on resistive-switching characteristics of resistive memories made exclusively using CMOS back-end-ofline materials is discussed.

The development of preclinical tools to study fluid transport within biological tissue is critical to understanding not only the progression of disease, but the role of the microenvironment in healthy tissue. The limited availability of experimental data across all length scales provides scope for the development of mathematical models to simulate fluid transport throughout the microvasculature and surrounding tissue. Here, the novel REANIMATE (REAlistic Numerical Image-based Modelling of biologicAl Tissue substratEs) platform is developed which, guided by both ex vivo and in vivo imaging data, simulates fluid and solute transport in silico, based on real-world tissue substrates. In this thesis, the intravascular flow model of Fry et al. (2012) and and oxygen transport model of Secomb et al. (2004) are applied to an in vivo cortical microvascular network containing the locations of fluorescently-labelled vascular smooth muscle cells. The simulated results provide insights into the mechanisms underpinning local regulation of cerebral blood flow which would be inaccessible in a conventional experimental setting. Secondly, a transvascular model is developed to simulate the effective transport of fluid through the vasculature and into the interstitium. parameterised against in vivo perfusion data, the model is applied to two ex vivo colorectal tumour datasets to investigate the role of vascular heterogeneity in elevated interstitial fluid pressure within tumours. Next, this platform is used to simulate the steady-state fluid dynamics in a further two murine xenograft models of human colorectal carcinoma, allowing for the prediction of heterogeneous delivery of specific therapeutic agents to be compared with that observed in vivo. Finally, developing upon work by Shipley and Chapman (2010), a discrete-continuum model is developed which allows for the approximation of fluid transport through tissue in the absence of experimental data on tissue-specific vascular micro-structures, thereby providing additional information unavailable in the traditional experimental setting.