Tidal stream energy is an emerging low-carbon technology which could meet 5% of UK electricity demand. Current developments use ‘axial-flow’ rotors, which are efficient but limited in size, to generate electricity from ocean currents. This thesis investigates the hydrodynamics of a previously undeveloped rotor concept which has two rows of blades and also has no inherent size limit, hence it might achieve greater economies of scale. The rotor concept, called the ‘Moonraker’, is a cross-flow device with an oval blade path in the horizontal plane. This thesis presents research into the hydrodynamic performance of the Moonraker, focussing on the forces exerted on the blades by water currents and thereby deriving the thrust on and power generated by a Moonraker. The point vortex method was used to model the Moonraker and predicted high power coefficients when compared to a conventional cross-flow turbine with a circular blade path. A lab-scale Moonraker device was built and tested in the towing tanks at UCL and QinetiQ. The device was 2 m wide, 0.5 m high, with up to six blades and was towed at up to 0.7 m/s (blade Reynolds numbers were in the range 65,000--112,000). One of the blades was instrumented with strain gauges so that two components of blade loading could be recorded. Comparisons of predictions and measurements of blade loading showed some encouraging agreement, but also some disagreement, leading to suggested improvements in the modeling of the blade forces. The vortex model was subjected to further verification and validation tests in order to explore the issue of double actuator surfaces in close proximity. The extension of this work could help optimise the spacing between the two rows of blades on a Moonraker.

Future fuels will be developed from a variety of biomass and fossil sources, and must seek to address the adverse environmental impacts of current fossil fuel usage. To this end, understanding how the molecular structure of a fuel impacts on the processes of combustion and emissions production is critical in selecting suitable feed-stocks and conversion methods. This work presents experimental studies carried out on a compression ignition engine equipped with a novel low volume fuel system. This system was designed and manufactured so as that several series of single-molecule fuels, and also binary fuel mixtures, could be tested to investigate the effect of fuel molecular structure on combustion and emissions. Features of fuel molecular structure that were studied include: alkyl chain length and degree of saturation, double bond position and isomerisation and the fatty acid ester alcohol moiety. The interactions between cyclic molecules and $n$-alkanes were also studied, as was the potential of carbonate esters and terpenes as future sustainable fuels; the latter produced from genetically modified micro-organisms. The engine tests were carried out at constant injection timing and they were repeated at constant ignition timing and at constant ignition delay, the latter being achieved through the addition to the various fuels of small quantities of ignition improver (2-ethylhexyl nitrate). In tests conducted at constant injection and constant ignition timing the ignition delay of the molecule was found to be the primary driver of combustion phasing, the balance between premixed and diffusion-controlled combustion and, thereby, exhaust emissions. The various features of molecular structure were found to influence the duration of ignition delay, and an effect of interactions of binary fuel mixtures was also visible. Physical properties, such as viscosity, impacted on the production of exhaust emissions, and in extreme cases also influenced combustion phasing and heat release.

Today there is a rising interest in the study of microfluidic systems. One of the main driving forces in this research area is integration of different components to have a chemical or biological analysis system on a single chip, which is called lab-on-achip system. Further development of lab-on-a-chip and micro liquid-cooling systems depend on micropumping technology. Therefore, a reliable and robust pumping technology is required for these and many other applications. Micropumps are one of the basic components of microfluidic systems and there have been several attempts to understand their underlying physics in order to improve the performance of these miniaturized devices. Most of these attempts were based on experimental work. Many different types of micropumps have been developed and studied by different research groups. One type is valveless diaphragm micropump (hereafter VDM), which in its simplest form consists of one pump chamber connected via one or more rectifying elements to an external system. The most common type of rectifying element in VDM is diffuser/nozzle element. To understand the behavior of VDMs the multi-physic nature of these devices should be investigated. An integrated approach is implemented to understand the behavior of VDMs, which consists of analytical and computational studies of the device. Analytical models (fluid dynamics and fluid-structure interaction) of different configurations of VDMs have been developed to relate physical properties of these systems to their performance characteristics. Computational studies for both fluid dynamics (CFD) and structural mechanics have been performed to establish a methodology for design and optimization of VDMs and facilitate their design process. The analytical results are compared to computational and published experimental results to investigate their correlation.

Bubble Science is a rapidly growing field, which has applications in (but not restricted to) health, climate and food engineering. Within these remits, one area of interest is related to the surface of bubbles. In this thesis specially designed and constructed microfluidic devices are combined with polymeric solutions to generate near uniform bubbles with smart coatings. Conventional bubbles possess coatings, which burst or collapse resulting in daughter bubbles or debris. The coatings developed in this study demonstrate particle precipitation from the bubbles surface. In this sense the structure is downscaling in size by a new mechanism proposed for bubble breakdown. In addition to this method, the size of resulting particles can be controlled. The in-situ particle-forming step also provides a pathway, which can retain bio- and chemical media during the formation process. Although several fields can benefit from such platforms, one such area is drug delivery, where this method is a novel approach in particle forming and delivery, when compared to conventional encapsulation routes. After initially demonstrating the process using a T-junction device was used to produce drug loaded particles with haematoxylin dye and estradiol drug using a novel bubble encapsulation system, a simple microfluidic V-junction device was used to prepare near-monodisperse polymer coated microbubbles using a hydrophobic polymer (ethylcellulose). From this coating, nanoparticles were produced continuously at a rate of ~5×106 per minute. It has been shown that these particles can be used as nanocarriers for a hydrophilic drug such an amoxicillin. During processing, varying the gas pressure has a significant effect on the nanocarrier diameter, thus we are able to control the diameter of nanocarriers (between 40 and 800 nm). Encapsulation efficiency using this process was in the range 65–88%, increasing with decreasing particle diameter and the drug-loading (Amoxicillin) was ~0.6mg. An X-junction microfluidic device was also used to prepare N2 bubbles using Tween. This method delivered a new route to generate significantly smaller bubbles (<5 m in diameter) with narrow size distribution. Moreover, Janus (two-phase) submicrometre size particles (710 nm) can also be generated using this X-junction device. The results from these devices and polymer coatings demonstrate potential applications as nanocarriers, which have greater control on size, shape and distribution. These factors are crucial for advancing healthcare science.

Demand for healthy foods has increased substantially over the recent years due to growth in the world population and unhealthy lifestyles. One approach that could help to address this problem is the development of structures that offer high volume to energy content ratio. Techniques that could offer the necessary microstructural engineering capability include: microfluidics and electrospinning. The aim of the work described in this thesis was to investigate the potential of these two techniques for application in food science. Food compatible materials were used to demonstrate the ability of these two techniques to generate bubbles and fibres. The material selection (polymer and solvent) process revolved around three fundamental principles: food biocompatibility, ease of processing and cost of material. The most suitable materials were found out to be natural polymers of low calorific value that have known to provide a pathway to manufacture food products with improved satiety index. Alginate was chosen as the primary material to generate porous foams of different sizes using a microfluidic T-junction device. This was accomplished by first generating monodisperse alginate microbubbles and then cross-linking them by calcium ions. Key parameters that affect bubble size and uniformity were also identified and the effects of bubble size on foam pore diameter were discussed. Electrospinning of ethyl cellulose and collagen to form fibres and scaffolds, respectively, constituted the bulk of the thesis with a comprehensive investigation into the process parameters (such as flow rate, applied voltage etc.). The control over fibre aspect ratio, with respect to processing parameters, was subjected to a detailed analysis consisting of optimization and reproducibility. The effect of solution concentration was also studied in detail. In addition, two new elements were introduced into the fibre generation process by customizing the electrospinning setup. Use of a glass needle and the effect of heat on fibre aspect ratio formed this part of the research. These detailed studies culminated in the development of a hybrid system that combined microfluidics and electrospinning techniques together for the first time. The result was the creation of structures using the aforementioned materials that exhibited properties specific to structures generated by each individual technique. The ability of this system to generate structures of different morphologies was discussed.

The long-term efficacy of coronary or peripheral stenting is limited by in-stent restenosis (ISR), which occurs in 15 to 30% of patients and is attributed primarily to neointimal hyperplasia. By adding a drug-eluting coating, this rate has been reduced to about 5% or less. However, recently longer-term follow-up data has highlighted problems with drug-coated stents, including late stage thrombosis. A bio-stable poly(carbonate-urea)urethane has been used for stent coating and the surface properties of the polymer have been optimised by incorporating the polyhedral oligomeric silsesquioxane molecule. These POSS polymers improve the adhesion and the growth of endothelial cells. The work described in this thesis, presents an innovative approach in self-expanding/balloon expandable coronary stent design that incorporates a NiTi/stainless steel alloy scaffold with a polyhedral oligomeric silsesquioxane- poly (carbonate-urea) urethane nanocomposite polymer (POSS-PCU) coating. Electrohydrodynamic spraying and ultrasonic atomization spraying of the non-biodegradable nanocomposite polyhedral oligomeric silsesquioxane (POSS) polymer have been investigated in detail for coating metallic stent materials and compared with dip coating. Because of the tight geometry of coronary stents, these new coating techniques have been shown to offer advantages over traditional coating techniques. These advantages include, reduced polymer consumption, precise coating thickness as low as 10 μm and a highly controllable spray which leads to consistent reproducible results. However, poor adhesion, or bond deterioration over the lifespan/ deployment of the device could reduces the efficiency and could impart even more complexity to the implant including formation of debris which can induce thrombus formation. Changing the surface physical property/chemical composition through the proposed protocol has been shown to increase the bonding strength by up to three times. This study has identified a new process and conditions which can be used in stent coating research.

Vortex-induced vibration (VIV) of a circular cylinder has been the focus of extensive research, as it can lead to fatigue damage in a wide range of industrial applications. When the forces induced by the periodic shedding of vortices from a structure in crossflow coincide with one of its natural frequencies, the structure can exhibit large amplitude vibrations. The majority of the work performed in this area has focused exclusively on transverse vibration, while relatively little is known about VIV acting in the streamwise (flow) direction, although this is known to have a strong effect on the overall response of structures with multiple degrees-of-freedom (DOFs). This work aims to characterise the behaviour of the wake and the structural response of a cylinder throughout the streamwise VIV response regime, which is crucial if the wealth of information on the transverse-only case is to be extended to the more practical and complex case of multi-DOF structures. Experiments were performed on a cylinder free to move in the streamwise direction for a range of reduced velocities in a closed-loop water tunnel. Particle-Image Velocimetry (PIV) was used to simultaneously measure the cylinder displacement and the velocity field in the wake, in the Reynolds number range 400 - 5500. The response regime was characterised by two branches, separated by a region of low amplitude vibration, as reported in the literature. Five distinct regions were identified, each of which was discussed in terms of the dominant wake mode, structural response characteristics, velocity profiles and estimates of the strength and trajectories of the shed vortices. In the first branch the wake was found to switch intermittently between the symmetric S-I mode (in which two vortices were shed simultaneously from either side of the cylinder) and the alternate A-II mode (which is similar to the von Karman vortex street observed behind stationary bodies). A criterion was developed which could determine which mode was dominant in a given instantaneous PIV field, and the effect of both modes on the cylinder response and wake characteristics was examined. Multi-modal behaviour was also observed in the second branch. At one value of reduced velocity, the wake could exhibit one of three modes; the A-II, the SA (similar to the A-II mode, with the vortices forming closer to the cylinder base) or the A-IV mode (which was characterised by the shedding of two pairs of counter-rotating vortices). Each mode was associated with a different cylinder response amplitude. The stability of the cylinder response while each mode dominated was examined using phase-portraits, which indicated that the system behaved as a hard oscillator. The forces acting on the cylinder were estimated using two methods, based on the measurements of the cylinder displacement signal and the flow field, respectively. The results found using both methods were in agreement, and the accuracy of the estimates was discussed. It was found that the amplitude of the unsteady drag force was very low between the two response branches, which was thought to be the cause of the reduction in the cylinder vibrations in this region. Finally, the effect of the various wake modes on the amplitude of the fluid forces throughout the response regime was examined. The results presented in this study provide a comprehensive description of the behaviour of the wake and the associated fluid forces throughout the streamwise response regime. The work reveals the inherent differences between the extensively studied case of transverse-only VIV and the streamwise-only case, which is crucial if the wealth of information available on transverse VIV is to be extended to the more practical multi-DOF case.

Electrospray techniques have become established in the life sciences for uses from cell encapsulation (Chang, 1964) to directed cell placement in more recent times (Jayasinghe et al., 2006a). During electrostatic encapsulation a conducting fluid in a needle connected to a high voltage power supply is charged and then drawn towards a grounded electrode by an electric field resulting in spraying. Cells and other materials can be encapsulated by suspending them in an alginate solution and electrospraying directly into a solution containing of a crosslinking agent, most commonly calcium chloride. This technique can be used to directly process and encapsulate many different types of materials (Jayasinghe and Townsend-Nicholson, 2006, Jayasinghe, 2007, Patel et al., 2008). This research adapts this technology further and progresses it by creating structures with multiple layers over an extended period with fluorescent markers contained within the layers, which are created through chemical adsorption. This allows the encoding information for the use in combinatorial stem cell biology where instead of individual experiments a large number of permutations are explored simultaneously. The research covers various parameters governing the encapsulation and layering processes as well as the biological functionality and integration as a tool for combinatorial stem cell cultures. The novel encapsulation and encoding technique presented here has a number of advantages over the currently available technology and has been filed as patent PCT/EP2010/006459.

The spray form development from a state of the art multi-hole injector for gasoline direct injection internal combustion engines is examined to attempt to determine the thermo-fluid dynamics affecting the spray development. The current state of knowledge regarding spray break-up and the interactivity of the factors on spray form are detailed. The spray under investigation was injected into purposely designed quiescent chambers to decouple the effects of the fluid mechanics on spray development from any in-engine effects. The pressure chambers, experimental apparatus and techniques used to characterise and measure the spray properties are described along with an assessment of any sources of variability in the measurement and analysis methodologies and hardware. Initial spray images of the spray produced by a range of multi-component “retail” fuels as well as single component non-oxygenated and oxygenated hydrocarbons with a range of boiling ranges and points for different injector body (and hence assumed fuel) temperatures and chamber gas pressures are presented. The experimental measurements show the strong interaction between the operational conditions in relation to the fuel properties and the physical spray form. A large amount of deviation from the nominal “ambient” spray form is observed for conditions where the fuel’s bubble point (boiling temperature at given gas pressure) is exceeded by a multiple of 10, termed spray collapse. The dependence of a multi-component fuel on the boiling characteristics of its highest volatility components suggest that it is these components which drive the fuel spray development formation, which is further illustrated by comparing different single component fluids. This suggests that higher volatility fluids are better representatives of full range, multi-component fuels for modelling or other investigative work when a single component fuel is required to be used. The onset of spray collapse was found to be gradual with no sudden “threshold” condition at which collapse occurred, also illustrated by a gradual reduction in measured spray droplet size with increasing injector body temperature and/or reducing gas pressure. The physical factors affecting spray development and break-up, and their effects are examined including the fluid flow inside a real size transparent, optically accessed nozzle, illustrating the effect of cavitation supplying nucleation sites for the subsequent vaporisation of the fuel. The scales of local air turbulence are found to affect the local vapour concentration, and hence vaporisation rate, and hence the interaction of these factors is shown to determine the spray formation.

Mixing is a key process in microfluidic systems since that samples and reagents generally need to be mixed thoroughly before chemical or biological analysis or reactions. Micromixers are designed to fulfil this critical process. In general, the development of microdevices is a competitive field that requires from researchers shorter times and lower costs in prototyping. Computational Fluid Dynamics (CFD) helps in reducing the time from concept to device design. Intuition and experience of the designer is usually behind its application on design improvement, by analyzing some physical variables to determine the effect of design parameters and to adjust them accordingly to the pursued objectives. In this thesis, a design and optimization strategy is presented and used for the analysis and design of micromixers. The method systematically integrates CFD with an optimization strategy based on the use of Design of Experiments, Surrogate Modelling and Multi-Objective Genetic Algorithm techniques. The aim is to define optimum designs that give the trade-off of the performance parameters, which in this study are the mixing index, defined on the basis of mass concentration distribution, and the pressure drop in the microchannel. Three types of micromixers have been studied and their geometric parameters have been optimized. They are the Staggered Herringbone Mixer and two novel designs, a planar micromixer with baffles in the microchannel and a 3-D T-type micromixer. A completed fabrication method was implemented as part of this thesis work and it was used to fabricate some of the micromixers. Experimental measurements and published data have been used to validate the numerical results. The outcomes of this thesis demonstrate that using advanced optimisation techniques on the basis of CFD solutions and analyses allows the design of optimum micromixers for different operation conditions, which can be set by the designer, without being necessary to use a referential design to start the method.