In this thesis we present theoretical predications of the error relative to sampling area when estimating turbulent parameters, such as the mean and rms value, form Particle Image Velocimetry (PIV) measurements. These results emulate similar time-domain analyses for Laser Doppler Velocimetry (LDA). Homogeneous and shear turbulent flow regimes are considered. The key analysis parameter, the mean square error, leads to an expression for the normalised standard error, The correlation coefficient is a prominent factor in the result and the error is evaluated for three distinct forms of this in one and two dimensions. Furthermore the results are simplified to yield general expressions for the error relative to sampling area given the integral lengthscale and rms turbulence level of the flow. The results are applicable, not only to PIV, but to any measurement method that produces results on a spatial domain. Results from PIV measurements of homogeneous isotropic grid turbulence in air and water are presented. The usefulness of the technique is discussed in terms of; the shape of the curve of the correlation coefficient, the resolution obtained relative to the turbulent lengthscales and coherent structures in the flow, and are, whenever possible, compared to LDA measurements. One of the measurements have been performed as a time series, and a number of vector and velocity plots show both the temporal and spatial development of the flow. In order to verify the theoretical predictions, several measurements of the same flow has been collected and a range of subareas of the total accumulated area has been defined. The relevant parameters and normalized standard error have been calculated for each of the sets of subareas. Given the required factors can be provided, we find the experimental and theoretical results in good agreement, proving this to be a useful result.

Today gas turbines are a crucial part of the global power generation and aviation industries. Small improvements to the efficiencies of individual components within the gas flow path can, over time, lead to dramatic cost savings for the operator and at the same time improve on the amount of carbon dioxide gas emissions to the environment. One such technology is the reduction of secondary flow losses in individual blade rows within the compressor or turbine section of the gas turbine through the use of non-axisymmetric endwall contouring. By introducing subtle geometrical features onto the endwall it has been shown to be possible to improve the efficiency of individual blade rows by between 1 and 2%. Few studies of these non-axisymmetric endwalls have been performed outside of the two dimensional cascade and computational domain, in addition these endwalls have been designed and tested to improve the performance of blade rows at a single design point with the off design performance having been ignored. The work presented here is aimed at investigating the use of such endwalls in a rotating blade row both at design and off-design conditions and in the presence of an upstream blade row. To this end a 1½ stage, low speed, turbine test rig has been refurbished and a new set of blades was designed to accommodate the profile of the Durham cascade at the hub. The Durham cascade is a de facto industry test case for non-axisymmetric endwall applications and therefore a generic, cascade proven, endwall design is available from the literature. The design of this new blade set is unique in that it is openly available. The results include steady-state 5-hole pressure probe measurements between blade rows and computational fluid dynamics solutions to provide detailed analysis of the flow quality found within the turbine. These results are reproduced for a turbine with annular or reference endwalls and one with the generic P2 endwall design obtained from the Durham cascade. Experimentally a 1.5% improvement in mixed out stage efficiency at the design condition has been found with a positive trend with increasing load. Additionally the rotor exit flows are show to be generally more uniform in the presence of profiled endwalls. The rotor torque is however reduced by as much as 3.5% and the improved flow uniformity does not always translate into a improved performance in the downstream row. Insight into the overall performance and fluid mechanics of the generic non-axisymmetric endwall at a variety of load conditions has been gained and an analysis of the parameters commonly used in optimising these endwalls is discussed with Cske being clearly shown to be the superior parameter in this case. CFD evidence suggests that while the cross passage pressure gradient is reduced by endwall profiling the extent of the effect of the change in hub endwalls reaches as far as the tip. The mechanism by which the overall loss is reduced appears to be a through a change in the relative strengths of the suction and pressure side horseshoe vortices and through the delayed migration of the passage cross flow, this change the relationship of these two vortex structures; dispersing the vortex structures as they leave the row and reducing the potential for mixing losses downstream.

This thesis focuses on practical condition monitoring of wind turbines. With offshore wind playing an increasing part in UK electricity generation, prompt fault detection leading to preventative maintenance is gaining in importance. This work describes the development of a condition monitoring test rig and the innovation and application of signal processing techniques for the detection of faults in non-stationary signals. Work is supported throughout by information from wind turbine operators and their experiences of variable speed, variable load wind turbines in the field. Experimental work is carried out on a condition monitoring test rig comprising a wound rotor induction generator, gearbox and DC driving motor. The test rig operates at variable speed and allows the implementation of a number of fault-like conditions including rotor electrical asymmetry, shaft mass unbalance and gear tooth failure. Test rig instrumentation was significantly developed during this research and both electrical and mechanical condition signals are monitored. A signal processing algorithm was developed based on experience with analysis techniques and their relationship with the characteristics of a wind turbine. The algorithm is based on Fourier analysis and allows the analysis of fault-related speed-dependent frequencies within non-stationary signals such as those encountered on a wind turbine. The detection of different faults is discussed and conclusions drawn on the applicability of frequency tracking algorithms. The newly developed algorithm is compared with a published method to establish its advantages and limitations.

Novel results on the optical binding of optically-trapped micro-particles are presented. A sophisticated Mie scattering model is developed, capable of performing time-evolution simulations of a multi-particle system. This is used to analyse and interpret experimental results in evanescent and Gaussian beam traps, and to develop simple, intuitive explanations for the observed phenomena. Novel trapped states are reported, that do not conform to the symmetry of the underlying trap. A common theme throughout this thesis is the "emergent'' phenomena that occur when multiple particles are trapped together, which cannot easily be predicted by considering each particle in isolation.

Flows with a limited optical access for PIV measurement can be probed using endoscopic PIV techniques. Conventional endoscopic PIV utilises two separate probes, one to relay a light sheet and the other to provide imaging optics to capture a two-dimensional image of the flow. This research aims to validate the velocity measurements taken from the novel single-stem endoscopic PIV system. The specific objectives are to determine the accuracy of the single-stem endoscopic PIV result, identify any shortcomings in the technique, and identify improvements for future single-stem endoscopic PIV systems. The single-stem endoscopic PIV system is applied to an atomised spray flow and velocity measurements are compared with conventional PIV and Pitot-static data. The endoscopic PIV system provides localised velocity maps that are comparable with the measurements from the conventional PIV system. This comparison is based on the spray ensemble mean flow field and its fluctuating velocity component statistics. A detailed analysis on the hardware setup, image capture, calibration and pre/post processing techniques is carried out to identify possible sources of systematic error in the measurement and how the measurement uncertainty accumulates. The mean velocity vector map, recorded from the single-stem endoscopic PIV system was used to estimate the spray mass flow rate and its entrainment characteristics, the centreline velocity decay, and the spreading rate similar to the corresponding estimates from conventional PIV. Furthermore estimates of the localised Strouhal number and of the spray fluctuation are also compared. By considering the measurement uncertainty as an accumulation of a series of component uncertainties, this study has identified and quantified the uncertainty contribution from each component. The largest sources of uncertainty are primarily due to two components. The first is the optical aberration, which leads to image defocusing and reduction in particle identification. The second component is the larger uncertainty source which is the uneven illumination of the measurement plane.

This study deals with the leakage flow over a shrouded turbine stage, its interaction with the main passage flow, and the associated losses. The study addressed these topics by providing an analytical correlation loss model and detailed CFD simulations. An analytical model of leakage flow loss over a shrouded turbine stage has been developed. The analytical model uses directly measurable flow quantities to predict the effect of some of the over-shroud design parameters on stage performance. The model displays good predictive ability for the mass leakage fraction and the mixing losses. The model resolves the negative incidence angle induced by mixing the leakage flow with the main stream and predicts the increment in the total mixing loss coefficient at increasing leakage jet injection angles. The main contributions of this model to the leakage jet models documented in the open literature are the effect of the leakage jet injection angle on the mixing loss and the accounting of the effect of the number of fins on the leakage mass fraction in an explicit way. The present model exhibits a good qualitative and quantitative agreement with comparative benchmark data. An in-house three-dimensional turbomachinery CFD code was developed and validated against six test cases, showing its ability to capture the salient flow features in each test case. This work makes an innovative use of Detached Eddy Simulation as an advanced Reynolds Averaged Navier-Stokes (RANS) model. A detailed leakage flow structure over the rotor shroud and its interaction with the main passage flow were modeled for seven test cases to investigate the effect of the number of fins, the clearance gap ratio, and the leakage jet injection angle on the flow. The results showed that reducing the injection 90° to 30° leads to a reduction in entropy mixing loss coefficient by up to 24.7% and gives a 0.2% increase in the rotor static to static efficiency and highlighted that reducing the leakage jet injection angle is a promising concept to control most of the adverse effects of the leakage flow.

Aerators had always been widely used for vanous purposes. The scale of the applications ranged from household fish tanks to water treatment plants. It was also applied to re-circulate large water body reservoirs, which this thesis aimed to address. Despite the wide application, the hydrodynamic theories of the flows created as a result of .air bubble plume application were still not totally validated. Therefore experimental results from past researches and from the present studies were used to validate the mathematical model of Goossens et ai's [1979,1980,1982], Fannelop et ai's [1980, 1991], in which the accuracy of the models were found to be less accurate than claimed in the publications. Through the experiments conducted, valuable results were obtained. These results were not only applied for numerical model validations mentioned above, but also used to clarify the physical characteristic of the circulations. Through these experiments, it was found that the secondary circulation do contribute to the overall mixing process, but in a slower process. The dye injected to trace the physical circulations was detected using the dye concentration detector known as SCUFA, in which the data were utilized to determine the concentration dispersion coefficient caused by the aeration. A commercial CFD software known as FLUENT was applied to model the case of aeration in both the homogenous and stratified conditions. Investigations were carried out systematically to determine suitable settings for the simulation. Experimental results obtained from the present study and from other published researches were used to verify and validate the CFD model. The results from the CFD model were compared with the experimental results through the vertical velocity profile, horizontal velocity profile and the overall circulation pattern. A verified and validated CFD model of the bubble aeration in the homogenous was obtained, in which the vertical velocity profile within the bubble plume, the horizontal outflow velocity profile and the overall circulation pattern can be accurately simulated. However, the simulation of the stratified conditions reqUIres a more complex procedure, in which the surface boundary condition has resulted in various inaccurate simulations. A 'Used Defined Function' was subsequently created to deal with the problem. A working FLUENT model for the stratified condition was successfully created in the end and the simulated results were validated with experimental results.

This work presents a new extension to a generalised nonlinear Maxwell model for the theoty of viscoelastic material flow. Nonlinear terms within this constitutive model are used to replicate many experimental phenomena such as shear-thickening/thinning, shear banding and dynamic stress responses found in complex materials such as polymers, Micelles, colloidal dispersions and even granular media. Numerical simulations of the stress tensor under spatially homogeneous plane Couette flow reveal a range of solutions from steady state to chaotic, chosen in part by the strength of nonlinear terms. Bifurcation and stability analysis reveal the onset of chaotic flow and are used to study the various transitions to chaos. A detailed phase space diagram is produced to categorise different dynamical regimes by determining the Lyapunov exponent under variation of two main model parameters. The route to chaos is identified primarily as a Hopf bifurcation.

No description available.

No description available.