Thermoacoustic instabilities are a mayor problem in industrial combustors, where they can lead to catastrophic hardware damage. An industrial gas turbine combustion chamber is a very complex and expensive system. Thus, a laboratory burner has been built for research purposes, where a large number of parameters can be varied. This study is part of the Marie Curie research network LIMOUSINE, which was set up to model thermoacoustic instabilities in the combustor chamber of gas turbines. The objective of the present thesis is to theoretically model and analyze thermoacoustic instabilities in the LIMOUSINE laboratory burner. A mathematical model of the laboratory burner has been developed. A more general form of the wave equation has been derived in the time-domain, in which the mean temperature gradient was taken into account. The governing differential equation has been solved by applying the Green’s function approach, which allows separating the effects of the unexcited burner and the fluctuating heat-release. Using perturbation techniques general solutions are given for the cases when the temperature increase is either small or large. Conclusions have been drawn about the necessary complexity of thermoacoustic models by comparing increasingly complex configurations. The forcing term of the wave equation is studied by investigating the kinematics of ducted premixed flames theoretically, and a new heat-release law is derived. Instability criterion has been derived by applying the non-linear source term. The stability parameter map of the burner has been also investigated. Expressions for the limit-cycle amplitudes and frequencies were derived using weakly non-linear theory. The predictions of the mathematical model have been compared to measurements.

This project aims to develop efficient and robust tools for optimal design of wind turbine adaptive blades. In general, wind turbine adaptive blade design is an aero-structure coupled design process, in which, the evaluation of aerodynamic performance cannot be carried out precisely without structural deformation analysis of the adaptive blade. However, employing finite element analysis (FEA) based structural analysis commercial packages as part of the aerodynamic objective evaluation process has been proven time consuming and it results in inefficient and redundant design optimisation of adaptive blades caused by elastic-coupled (bend-twist or stretch-twist) iteration. In order to achieve the goal of wind turbine adaptive blade integrated design and analysis, this project is carried out from three aspects. Firstly, a general geometrically linear model for thin-walled composite beams with multi-cell, non-uniform cross-section and arbitrary lay-ups under various types of loadings is developed for implementing structural deformation analysis. After that, this model is validated by a simple box-beam, single- and multi-cell wind turbine blades. Through validation, it denotes that this thin-walled composite beam model is efficient and accurate for predicting the structural deformations compared to FEA based commercial packages (ANSYS). This developed beam model thus provides more probabilities for further investigations of dynamic performance of adaptive blades. Secondly in order to investigate the effects of aero elastic tailoring and implanting elastic coupling on aerodynamic performance of adaptive blades, auxiliary software tools with graphical interfaces are developed via MATLAB codes. Structural/material characteristics and configurations of adaptive blades (i.e. elastic coupling topology, layup configuration and material properties of blade) are defined by these auxiliary software tools. By interfacing these software tools to the structural analysers based on the developed thin-walled composite beam model to an aerodynamic performance evaluator, an integrated design environment is developed. Lastly, by using the developed thin-walled composite beam model as a search platform, the application of the decoupled design method, a method of design of smart aero-structures based on the concept of variable state design parameter, is also extended.

This report is an account of the work carried out, mainly by the author, on the development of high speed rotors for the production of hypersonic colliding beams of molecules. Initially a brief comparison is made between various standard techniques used for the production of molecular beams with translational energies in the range 1 to 20eV. Then consideration is given to the part that rotors, producing tip speeds of 2km/s, can play in producing colliding molecular beams with available reaction energies from 1 to 20eV. Following this, the theoretical and practical considerations used to produce a successful high speed rotor, using carbon fibre composite arms, are discussed. A full description of the single rotor unit is given and also of the double rotor unit, which is designed to produce colliding beams. Full circuit details of the rotor suspension system, the drive unit and a semi- digital and a fully-digital unit for phase- locking two rotors for use when producing colliding beams are given. Results of an experiment are described, using one rotor, to measure the energy- intensity characteristics of a rotor - produced beam of mercury atoms. Finally, a description of a proposed experiment to optically excite mercury atoms by collision of two rotor- produced beams is given, this to be conducted in the double unit.

A novel process in which turbine blade moulds are manufactured by rapid prototyping has been developed. Fine control over the raw materials and their processing is required. The non- Newtonian rheological behaviour of the ceramic slurries can be modified through changes in the dispersant levels. A link between dispersant concentration, particle-particle interactions and firing performance has been suggested. Alumina agglomerates that remain through mixing were identified and shown to affect the viscosity of the resin by as much as 30% leading to poor process robustness. X-ray tomography was used to identify and visualise highly orientated defect structure in the ceramic linked to the layered build and the residual stress distribution developed by the Gaussian profile of the laser. It was shown that changes to the laser settings would affect the structure of defects but did not allow for defect removal. The flexural strength of the ceramic produced was found to be by highly dependent on sample orientation and varied between 14 and 52 MPa. Methods to improve the defect structure were explored, including changes in UV cure technology, changes in process settings, removal of agglomerates from the alumina and improved firing schedules. Both a reduction in defects and improved surface finish were identified. It is apparent that moulds suitable for casting applications can be processed using this novel technique.

The main focus of this thesis was to investigate the influence of solidity on the performance, swirl characteristics, wake length and blade deflection of a Horizontal Axis Tidal Turbine (HATT) using the simulation software package Ansys. An existing laboratory scale prototype HATT was modified to improve upon previously gathered experimental data and provide further confidence of the validity of the numerical models. The solidity was varied by altering the number of blades in the numerical models. The work presented in this thesis shows that, for this blade profile, increasing the solidity increases the peak Cθ and peak Cp and reduces the λ at which these occur. Ct was found to be approximately the same at peak Cp, which was assumed to be the normal operating condition. At λ above peak Cp, near freewheeling, Ct continued to increase for the 2 bladed turbine, remained approximately constant for the 3 bladed turbine and decreased for the 4 bladed turbine, due to the change in pitch angle required to maintain optimum power. This indicates that higher solidity rotors would have to withstand lower loads in the event of a failure. In addition, the thrust per blade was shown to increase with reducing number of blades. The swirl characteristics in the wake were found to agree with swirl theory and the swirl was found to increase with solidity whilst being weak or very weak in each case. Swirl number was found to be dependent on solidity only up to distances of 10 diameters downstream. At higher turbulent intensities, the wake recovery was only influenced by solidity up to 15 diameters downstream of the HATT but at low turbulence intensities the wake length increased with solidity indicating that low solidity rotors may offer higher overall array efficiencies in areas of low turbulent intensity. Blade deflection was shown to increase with a reduction in the number of blades, due to the increased thrust per blade. The power output of the 3 bladed turbine was shown to decrease by 0.4% with a deflection of 0.12 m. However, the power output of the 2 and 4 bladed turbines was found to increase with deflections as it was subsequently found that the pitch settings found in a previous study were not fully optimised for a rigid blade. At deflections above 0.20 m the power output of the 4 bladed turbine was found to decrease. It is expected that the power output of the 2 bladed turbine would eventually decrease with further deflections but no decrease was found for the maximum deflection considered, of 0.35m. This thesis therefore shows that the optimum number of blades may vary from site to site and even from one location within an array to another. It also shows that blade deflection will alter the power output and that blades could be designed so as to reach their optimum setting at a given blade deflection.

This thesis concerns an analysis of an Integral Compressed Air Wind Turbine (ICWT), in which energy is extracted from a slow-moving renewable source through the use of compressed air. This concept is particularly applicable to large offshore wind turbines, and can be readily integrated with compressed air energy storage methods. The ICWT has a very large rotor with free pistons travelling within the rotor blades, inducting and compressing air to high pressures at each end of the stroke. The compressed air can be stored and expanded when the energy is required, solving the intermittency issue of wind energy. By gathering energy along the rotor blades, rather than at the hub, it also avoids the very high torques associated with extremely large turbines. This thesis investigates optimal control strategies for ICWTs. Firstly, an initial model of the system using coupled ordinary differential equations (ODEs) is constructed to simulate a single piston pair of an ICWT system. This framework utilises several `modes' which the system passes through in the course of each stroke, with movement between modes controlled by simple algorithms. Calculations of potential and required energy are developed to allow basic control of the valve timings. The simulation is then extended to include thermal modelling of the walls of the compression tube, using orthonormal polynomials. A long-duration instance of the model is used to identify steady-state values for the orthonormal parameters, which demonstrates that the wall temperatures would remain within 15~K of the ambient temperature. One possible solution to the high temperatures caused by the near-adiabatic conditions of the compression is added to the model; namely, the injection of water droplets to the cylinder at the start of the compression stage. A method to efficiently simulate a phase transition in MATLAB is developed and implemented, allowing an analysis of the optimum mass balance of water to be injected to reduce the exhausted air temperature. An appendix examines several of the assumptions built into the model, in particular the rigidity of the components and variations in the rotational velocity of the rotor due to Coriolis and gravitational forces. Two valve control schemes are developed and implemented into the model; firstly, a simple proportional and derivative controller, which acts according to a rule dictating a gradual reduction in the energy surplus. This option proves to be limited in the degree to which it can avoid wasting compressed air. A second scheme, involving a simple version of sliding-mode control with two controllers operating at different timescales, is instead developed and shown to be significantly more effective at improving the system efficiency. Finally, an optimisation study is carried out on the `kick' stage, in which stored compressed air is used to accelerate the piston before compression. A large dataset of simulations allows for the specification of a set of optimum parameters based on a balance between power extraction from the rotating frame and net power generation.

The present computation investigates a turbine blade with trailing-edge cutback coolant ejection designs, aiming for a comparison study of aerothermal performances such as discharge coefficient and film cooling effectiveness due to the change of trailing-edge geometries and blowing ratios. The shear-stress transport (SST) k-w turbulence model is adopted and numerical studies are carried out by two-stage investigations:- firstly, validation of an existing cutback blade model with staggered circular pin-fins array inside the cooling passage that has been extensively studied by other researchers and predicted internal passage discharge coefficient and film-cooling effectiveness along the cutback surface are compared to experimental measurements. RANS/URANS and DES are applied during this stage; secondly, further investigation of four main cases considering different key design parameters such as the ratio of lip thickness to slot height (t/H = 0.25, 0.5, 1.0 and 1.5), the design of internal features (i.e. circular pin-fin array, elliptic pin-fin array, and empty duct), the coolant ejection angle (alpha = 5 degrees, 10 degrees and 15 degrees). In addition, a trailing-edge cutback model with suction-side (SS) ─ pressure-side (PS) walls and lands is considered to create a more realistic blade design. The results show that both steady and unsteady RANS predictions are able to produce discharge coefficients in fairly good agreement with test data, but not the film-cooling effectiveness on cutback surfaces which over-predicts in far-field wake region. Further prediction improvements can be made by using unsteady DES approach. In terms of film-cooling effectiveness and shedding frequency, computational results indicate a strong dependency on those aforementioned key design parameters. This film-cooling effectiveness is strongly affected by turbulent flow structures along the cutback region, which is representing the dynamic mixing process between the mainstream flow and the ejecting coolant from the slot-exit. The use of elliptic pin-fin inside the cooling passage and thin lip thickness could improve the effectiveness of film-cooling. The increase of ejection angle yields almost near unity cooling effectiveness along the protected wall. Significant improvements on cooling performance are also achieved with higher blowing ratios. Computations of the trailing-edge cutback cooling with pressure-side (PS) and suction-side (SS) wall demonstrates that performance of the case without lands is better than that of the case with lands by discrepancy up to 18% in terms of overall-averaged film-cooling effectiveness. The blade trailing-edge design with lands causes a rapid decay of the averaged film-cooling effectiveness.

The power output from a wind turbine is proportional to rotor swept area and as a result in the past 30 years continuous effort has been made to design larger blades. In this period, the blade length has been increased about 10 times since 1980s to present time. With the longest blade currently measuring more than 100m in length, wind turbine blade designers and manufacturers face enormous challenges to encounter the effect of increased weight and other loads on fatigue durability of the blade. Wind turbine blades are mainly made from glass fibre reinforced plastic (GFRP) composite. materials. As a result, in the design of various parts of wind turbine blades such as the shear web, spar cap and the aerofoil the fatigue behaviour of F RP materials is required. The performance of these parts as well as the adhesively bonded joint under fatigue loading is crucial for structural integrity of a long lasting blade. During operation, delamination can initiate and propagate shortening blade life; hence, characterisation of failure envelope of GFRP laminates under different loading mode is necessary. In this regard in this project, quasi-static tests were carried out to find mode 1, mode 11 and mixed mode I/11 delamination fracture toughness using DCB, ENF and MMB tests and the fracture envelope was established for various mode mixity. In the next stage, the stress-lifetime (S-N) diagrams of the GFRP was studied. Fatigue-life experiments on three different types of loading, i.e. tension-tension at R=0.1, 0.5, tension- compression at R=-1 and compression-compression at R=2 and R=10 were performed. From the results of S-N diagrams, the constant life diagrams (CLD) for 90 degree and 0 degree fibre directions were constructed. CLD diagrams are useful for prediction of fatigue lifetime for loading condition that no experimental data available. The analysis of delamination crack propagation under cyclic loading was next area of the research. The onset life and propagation delamination crack grth of 0//0 interface of GR P laminate in mode I loading using DCB specimens was investigated and the Gm. from the onset life test was determined. From the fitted curve to mode I experimental propagation data the Paris’ law coefficient for the laminated GFRP in mode I was determined. The mode II fatigue crack growth in laminated 0//0 GFRP material was also investigated using ENF specimens. The fatigue behaviour in this mode is analysed based on application of Paris’ law as a function of energy release rate for mode II loading. From the fitted curve to experimental data, the Paris’ law coefficient for the laminated GFRP in mode II was determined. The effect of fatigue delamination growth on fracture surface was studied by fractography analysis of SEM images of fracture surfaces. Studying the behaviour of GFRP under cyclic loading and delamination under static and dynamic load led to full-scale testing of wind turbine blade to establish damage tolerance of the blade under cyclic loading. The sensitivity of wind turbine blade to damage has considerable interest for turbine operators and manufacturers. For full-scale fatigue testing, calibration test and modal analysis of a 45.7m blade has been done and moment-strain diagram and natural frequencies of the blade were obtained. Next, the blade sensitivity to damage under fatigue loading was investigated. The blade has been damaged intentionally by initially inserting a crack of 0.2m between the shear web and spar cap and later it was extended to 1m. The effect of these damages on the modal shape, natural frequencies and strains at various locations of the blade were investigated. The damaged blade fatigue tested, the structural integrity and growth of damage were monitored, and the results were discussed. Finally for the improvement of delamination resistance for joints between spar beam and aero-shell stitching method was used. T-beam and box beam joint were chosen as the platform for testing the stitching effect on the delamination. Various pattern of stitching was applied and the optimum pattern was determined.

Small scale distributed compressed air energy storage (D-CAES) has been recognized as promising technology which can play major role in enhancing the use of renewable energy. Due to the transient behavior of the compressed air during the discharging phase, there are significant variations in air pressure, temperature and mass flow rate resulting in low turbine efficiency. This research aims to improve the expansion process of the small scale D-CAES system through optimization of a small scale axial turbine. A small scale axial air turbine has been developed using 1D Meanline approach and CFD simulation using ANSYS CFX 16.2. For improving the turbine efficiency, different optimization approaches like single and multi-operating point optimization have been performed. The turbine blade profiles for both stator and rotor have been optimized for minimum losses and maximum power output based on 3D CFD modelling and Multi Objective Genetic Algorithm (MOGA) optimization for single and multi-operating points. Using multi-operating point optimization, the maximum turbine efficiency of 82.767 % was achieved at the design point and this approach improved the overall efficiency of D-CAES system by 8.07% for a range of inlet mass flow rate indicating the potential of this optimization approach in turbine design development.

A novel process in which individual turbine blades are produced by the high rate solidification method has been developed. The technology, with an optimised radiation baffle, gave a thermal gradient of 11.64 x 10\( 3\) K.m\( {-1}\), whilst maintaining a flat solidification front, calculated using a specially created and validated process model. The corresponding primary dendrite arm spacing was reduced to 165 x 10\( {-6}\) m and the calculated freckle potential was below the critical threshold limit identified by Beckermann et al. (2000), even in highly freckle prone alloys. Low angle grain boundaries formed when misorientation accumulates in growing dendrite envelopes which subsequently converge were studied experimentally. While extensive dendrite branching or steady state growth was found to lead to an average primary dendrite misorientation of 2.3 ° that was random in nature, enhanced growth kinetics accompanying non steady state growth, found in platforms, produced a monotonic increase in accumulated misorientation of up to 10 °. It was concluded that the latter is due to mechanical moments arising from extensive growth of unsupported tertiary dendrite branches growing laterally across the platform normal to the direction of gravity. The degree of misorientation is therefore dependent on local geometry and mushy zone shape.