This thesis investigated different quantitative methods for dental plaque detection using digital imaging. Firstly, based on a commercially available two-tone disclosing, the concentration of the dyes in the blue disclosing solution was calculated. This blue dye was used to disclose dental plaque accumulated on natural teeth and complete upper dentures (on two different backgrounds). Digital images were acquired under visible light, in the n-IR spectrum and with a narrow band-pass interference (NIB) filter tuned to the absorption spectrum of the blue dye. The results showed that disclosing dyes and disclosed dental plaque are transparent in the n-IR spectrum whilst the NIB filter maximised the contrast of dental plaque in the images when using the blue stain. A number of computerised segmentation methods were applied to these images showed automation of dental plaque detection to identify reliable methods to quantify plaque coverage. Although minor human intervention was still required in the segmentation process, the continuous development of new software promises that full automation in plaque quantification is almost a reality. Finally, analysis of the inter- and intra-examiner reliability of the commonly used Quigley and Hein index showed moderate reliability, highlighting the need for automated, quantitative and more reproducible methodology.

Filamentous micro-organisms, grown in submerged culture, are used in a large number of fermentation processes, such as the production of penicillin and citric acid. Generally, production takes place in baffled sparged vessels agitated by either single or multiple impellers, traditionally Rushton turbines mounted on a common shaft. This configuration is satisfactory for low viscosity systems, e.g. yeast and some bacterial fermentations. However, in cultures of filamentous fungi, as the fermentation time progresses, the broth may become increasingly viscous, and possibly shear thinning, with the consequent onset of dead zones or stagnant regions. The resulting concentration gradients may cause a reduction in the productivity. It may be possible to overcome these problems by using a fermenter which has two independently driven impellers and a cylindrical draft tube (IDDIDT i.e. an acronym for "Independently Driven Dual Impeller with Draft Tube") Studies have been undertaken in a 0.75 m Perspex proto-fermenter employing a Rushton turbine (6RT) and a Scaba 3SHP1 axial flow impeller, with test fluids such as water and CMC, under unaerated and aerated conditions. In order to fully characterise the system, important mixing aspects such as mass transfer (hold-up, k\(_L\)a) and bulk blending (mixing times, circulation time distributions) were investigated and related to the power input from each impeller. At low gas rates, moderately increased hold-up and k\(_L\)a values were observed with the IDDIDT compared to single or dual Rushton turbine systems. At higher gassing rates the performance was similar for all configurations. For the IDDIDT, regardless of the proportion of the total energy dissipation rate contributed by each impeller, the k\(_L\)a was the same. In addition, very rapid mixing times approximately 3 times faster than predicted by correlations for single impeller, single shaft systems and experimentally determined values for the dual Rushton system, were observed. Mixing times could be directly controlled by the 3SHP1 axial flow impeller, and confirmation of the increased exchange flow rate throughout the vessel was achieved via circulation time distribution (CTD) experiments. Since very little data has been published regarding the use of fermentation broths in these systems, hygienic Aspergillus niger fermentations were carried out in the fermenter to clarify any improvements when using the novel reactor and also to highlight any discrepancies in the use of test fluids, such as CMC, as model broths. Although unaerated and aerated power characteristics were similar for both the test fluids and fermentation broths the hold-up values varied considerably, probably due to differing coalescence characteristics. Productivity (g/l/kWh) of batch phase Aspergillus niger fermentations was unaffected by system geometry, at least up to concentrations of circa 15 g/1.

Diffraction in pp collisions contributes approximately 30 % of the inelastic cross section. Its influence on the pseudorapidity density is not well constrained at high energy. A method to estimate the contributing fractions of diffractive events to the inelastic cross section has been developed, and the fractions are measured in the ALICE detector at 900 GeV (7 TeV) to be f\(_D\)=0.278\(\pm\)0.055 (f\(_D\)=0.28\(\pm\)0.054) respectively. These results are compatible with recent ATLAS and ALICE measurements. Bjorken’s energy density relation suggests that, in high multiplicity pp collisions at the LHC, an environment comparable to A-A collisions at RHIC could be produced. Such events are of great interest to the ALICE Collaboration. Constraints on the running conditions have been established for obtaining a high multiplicity pp data sample using the ALICE detector’s multiplicity trigger. A model independent method to separate a multiplicity distribution from ‘pile-up’ contributions has been developed, and used in connection with other findings to establish a suitable threshold for a multiplicity trigger. It has been demonstrated data obtained under these conditions for 3 months can be used to conduct early strangeness analyses with multiplicities of over 5 times the mean. These findings have resulted in over 16 million high multiplicity events being obtained to date.

The cleaning of pipework from fluids of high viscosity is a significant problem in many food and personal care industries which can cause considerable economic and environmental impact. Three stages are identified in the cleaning of straight pipes; (i) a short core removal stage of product recovery, before water breaks through the filled pipe, (ii) the 1st cleaning stage (film removal stage) when there is a continuous wavy annular film on the wall, and (iii) the 2nd cleaning stage (patch removal stage) in which the material is present as patches on the wall. The product recovery stage is found to influence the overall cleaning process. Conducting product recovery at low temperatures and high flow rates cause the formation of a wavy wall layer which leads to more rapid subsequent removal. A two step CIP protocol is proposed to decrease the environmental impact of cleaning of a viscoelastic material (toothpaste) from pipework. Applying cold water in the 1st cleaning stage and hot water in the 2nd cleaning stage results in 40 % energy saving without affecting the cleaning performance significantly compared to traditional CIP protocols used in plants. Yield stress of deposits is the key effect on the cleaning of deposits. A new dimensionless number is explored. It is physically a ratio of flow energy to the yield stress of a deposit. It has collapsed cleaning time data onto a one curve for different deposits cleaned at different velocities. It can be used to predict cleaning times of deposits or identify the governing cleaning mechanisms in cleaning. It is found that the magnitude of pressure loss in flow can be used to quantify the flow and turbulence effect on cleaning and scale up lab scale data. The results show that although fluid mechanical removal is a complex process, cleaning time of deposits can be predicted by knowing its rheology and fluid mechanical parameters of cleaning fluid.

The static and dynamic properties of carbon fibre composites of varying orientation, stacking sequence and geometry has been analysed in terms of modulus and material loss factor up to strain rates applicable to golf club shafts. No noticeable change in modulus or damping was seen at strain rate applicable to golf club shafts. All panels tested strain rate sensitivity onset occurred at around 0.4 s\(^{-1}\), which is above the maximum observed during a golf swing (0.1 s\(^{-1}\)). The strain rate sensitivity was found to be sensitive to aspect ratio (for strain rates above 0.4 s\(^{-1}\)). Two 20° panels of the same fibre interfacial area, but with different aspect ratios (length/width) showed different strain rate sensitivities. The short wide panel (aspect ratio 1.5) showed a higher stiffness and lower strain rate sensitivity when compared to a panel with an aspect ratio of 2.6. A model was created to predict the modulus and damping of lay-ups for laminates and golf club shafts. The model was validated against three composite systems at varying orientations and stacking sequences. The software agreed well with laminate experimental data (data sets showed a RMSD of less than 5 %). From this an optimising software was developed to provide the user with a stacking sequence that will optimise modulus, damping or the product of both. This thesis also evaluated commercial shafts in order to determine the models applicability to this application. Commercial shafts were tested for both stiffness and damping, where a number of aspects such as inter-ply resin rich regions and seam were observed as possible areas for discrepancy with the models prediction. Shafts were fabricated in order to analyse these aspects in greater deal, and to determine the models limits for this application. The model accurately predicted the stiffness of the shafts however the model failed to predict the damping of the shafts when comparing to the average values taken. When damping was compared to the areas where no seams were present, the model agreed well except for in two cases, which have been attributed to shafts flaws (cracks or excess inter ply resin). The model presented in this research consistently characterised the stiffness of fabricated shafts, however the seams proved too dominant a feature to be neglected in the prediction of damping.

The research presented in this thesis discusses the design of a new configuration of a hybrid parallel robot with nine degrees of freedom. Parallel robots have been studied due to their high stiffness and accuracy rather than serial robots. However, parallel mechanisms have not been widely used in industry due to the lack of workspace. A major advantage of the proposed hybrid parallel robot is to increase the work volume while the stiffness of the system remains suitable for a range of industrial applications. Novel dynamic formulations have been developed by using Newton Euler and inverse kinematics in order to identify the best configurations based on stiffness and velocity of system in any particular position. The developed program determines the velocity and stiffness of the system for each configuration. A physical robot prototype built to test the developed theoretical model. The control strategies of the robot were developed and tested based on both point to point control and continuous path applications. The path between two points was selected based on the stiffness of the system in a particular position and orientation. A robotic ankle rehabilitation application was successfully used to verify the design of the proposed hybrid parallel robot.

The continual development of precise timing has always pushed technology forward, and the latest generation of clocks uses electronic transitions within cooled atoms as their frequency reference. For atoms where the clock transition is at an optical frequency an ultra stable laser is used to probe the transition; the construction and development of this laser for use with strontium is the subject of this thesis. The system consists of a laser, an optical cavity and the electronics to lock the laser to the optical cavity. The theoretical component of this thesis looked into the possible materials used to make the optical cavity from and how the optical cavity should be supported to minimize the frequency noise of the laser. A compact external cavity diode laser was constructed and shown to have a linewidth of 26 ±15 kHz. To lock and narrow the linewidth of this laser, two ultra stable optical cavity assemblies were constructed with vibration and thermal stabilisation enclosures. To characterise the laser stability a beat measurement was performed between the two stabilised lasers and a beat note linewidth of 1:37 ± 0:28 Hz was measured. The instability of the beat note was measured; an Allan deviation of 2 x10\(^-\)\(^1\)\(^4\) was obtained between averaging times of 0.2 s and 1 s.

The primary purpose of this thesis is to present a comprehensive analysis of occupant kinematics and spinal injuries, during road traffic accidents in Saudi Arabia from the points of view of statistical analysis, modeling of occupant kinematics, and biomechanics. An in-depth database containing information on 512 real world vehicle crashes was constructed. The study identifies the characteristics of the collisions and occupant spinal injuries in Saudi Arabia, and suggests measures to mitigate them. A logistic model has been presented which can be used to provide information about the crashes and spinal injuries. The model may serve as an initial prediction to establish the risk of spinal injury sustained by occupants at road crash, and a paramedic’s protocol, as part of the emergency response, could be revised according to the developed model. State of the art techniques for accident reconstruction have been demonstrated as a tool to investigate the crashes, and the probable cause of crashes, and to make recommendations to prevent crashes and/or mitigate the severity of the accidents and resulting spinal injuries. Computational simulations of crashes provide a tool for understanding the dynamics of crashes and injuries, and are being used worldwide to study dynamics of crashes and efficacy of safety devices. The work conducted here has demonstrated how crashes can be simulated to estimate the injury parameters, and the likelihood of injuries on various parts of the body. While this study presents a detailed multi-dimensional study on road traffic crashes and spinal cord injuries therein, it remains a pilot study for Saudi Arabia. It demonstrates how this type of study can have far reaching consequences and the need to collect such data and carry out this kind of a study on a regular basis at the national level.

This thesis presents a systematic study on the characterising ultrafast dynamics of photo excited carriers in hydrogenated nano crystalline silicon (nc-Si:H) and porous silicon (PS) materials using the ultrafast optical pump probe spectroscopy approaches. The work involves the ground state optical property detection and optical model simulation, ultrafast time resolved pump probe measurements and pump probe data analysis processes. Applying the novel pump probe detection methods, the study explores the carriers concentration, carriers recombination property, scattering rate and conductivity of photo excited semiconducting nano materials.

Preclinical bioluminescence tomographic reconstruction is underdetermined. This work addresses the use of prior information in bioluminescence tomography to improve image acquisition, reconstruction, and analysis. A structured light surface metrology method was developed to measure surface geometry and enable robust and automatic integration of mirrors into the measurement process. A mouse phantom was imaged and accuracy was measured at 0.2mm with excellent surface coverage. A sparsity-regularised reconstruction algorithm was developed to use instrument noise statistics to automatically determine the stopping point of reconstruction. It was applied to in silico and in simulacra data and successfully reconstructed and resolved two separate luminescent sources within a plastic mouse phantom. A Bayesian framework was constructed that incorporated bioluminescence properties and instrument properties. Distribution expectations and standard deviations were estimated, providing reconstructions and measures of reconstruction uncertainty. The reconstructions showed superior performance when applied to in simulacra data compared to the sparsity-based algorithm. The information content of measurements using different sets of wavelengths was quantified using the Bayesian framework via mutual information and applied to an in silico problem. Significant differences in information content were observed and comparison against a condition number-based approach indicated subtly different results.