Search

A growing desire exists to develop Micro Air Vehicles (MA Vs) that fall within a 15cm span. Their small scale and low operating flight speeds encourage a low Reynolds Number (Re) regime, in the order of Re - 104 - 105 . Wings under these conditions are highly susceptible to separated flows, posing a significant challenge for the MA V. Natural flyers are able to confront these issues through flapping flight, which has inspired an entire research field on the aerodynamics of oscillating wings. While the number of parameters that govern the problem is exhaustive, studies are required to explain the contribution of each and any phenomena that may ensue. This lends itself to a canonical approach. This thesis presents an experimental study on various wing geometries, undergoing a small amplitude oscillation in the form of a pure plunge. The focus lies on understanding the three-dimensional effects of oscillating a finite wing with a positive geometric angle of attack, to encourage greater lift than that achieved from an unforced wing. This expands on the current research which predominantly focuses on the thrust generating capabilities of a 'flapping' airfoil. Force measurements, hot-film measurements, Particle Image Velocimetry (PIV) and volumetric velocimetry, are used to examine the performance and flow topology that ensues from actuating the various wings. The study presents time-averaged force measurements as a function of Strouhal number (non-dimensionalised plunge frequency) for the various low aspect ratio wings. It is shown that while the finite nature of these wings suppresses lift, significant improvements are nonetheless possible. For example, a semi Aspect Ratio = 2 NACA0012 rectangular wing, is able to achieve 180% more lift than the unforced wing. A phenomenon arises in which peaks are observed in the time-averaged lift curve, for various rectangular and delta wing planforms. This suggests optimal lift conditions at particular Strouhal numbers. In a similar manner to a 2D airfoil the oscillating wing stimulates the formation of both leading and trailing-edge vortices. The trajectory and timing of these vortices, in relation to the plunge cycle, appear to be significantly affected by Strouhal number. At particular frequencies, the vortices interact in such a way that their induced flow generates a significant region of low velocity, recirculating flow near the wing. The size of the recirculating region closely correlates with the shape of the time-averaged lift curve, agreeing well with points of troughs and peaks when this region is maximised and minimised, respectively. It is thought that these Wing/vortex and vortex/vortex interactions contribute to the selection of optimal frequencies, and therefore determine optimal lift for the oscillating wing.

Numerical analysis conducted using Callisto, which is Airbus’s three-dimensional momentum integral boundary layer code coupled with Green’s lag-entrainment method has shown that there might be a small but worthwhile form drag reduction through attachment line control, up to about 0:4-0.6 counts for an aircraft. However, in order to overcome numerical issues in the modelling a few approximations have been made in the method while calculating the flow very near the leading ledge. The detail of the leading edge flow needs to be verified if the drag results are to be trusted. Therefore, an experiment carried out, aiming to capture the velocity profiles starting from the attachment line and up to about 3% chord downstream. In order to design the experimental model, a systematic approach was used based on previous semi-empirical work on the attachment line flow. The model was designed so that the attachment line boundary layer is turbulent due to contamination from the turbulent boundary layer from the wall (floor) of the wind tunnel and thick enough to give a sensible experimental domain size including sufficient chord wise extent for hot-wire measurement. The velocity profiles were captured by means of hot wire anemometry using a micro displacement traverse designed and manufactured in-house.

This thesis investigates three elements necessary for the prediction of the broadband noise from a turbofan engine due to the interaction between the turbulent rotor wakes with the Outlet Guide Vanes (OGVs). These are (i) the sound radiation from a cascade of closely spaced blades interacting with rotor wake turbulence, (ii) an analysis of the behaviour of hotwire velocity data from a Large Scale Fan Rig (LSFR), (iii) the development of a scheme for the prediction of the blockage due to the transmission of multi mode sound across the rotor necessary for the prediction of noise in the forward-arc. (i) Cascade noise model A noise model is presented for the prediction of rotor wake turbulence with a cascade of OGVs. Similar to other approaches of this kind, computation time becomes excessive at high frequencies as the number of modes required increases. This thesis shows that at sufficiently high frequencies, when at least two modes are cut-on between adjacent blades, the acoustic blade coupling is weak and the cascade sound radiation closely approximates to that of an isolated aerofoil whose radiation can be computed efficiently using single airfoil theory, thereby greatly reducing computation time. (ii) Characteristics of rotor wake turbulence One factor currently limiting accurate fan broadband noise predictions is an understanding of rotor wake turbulence at the OGV leading edge. This thesis analyses in detail recent hotwire velocity data measured in the interstage of an LSFR. The focus here is on assessing the extent of self-preservation in the rotor wake, whereby the mean and turbulent wake characteristics can be deduced at any position downstream of the rotor and at any operating condition from a limited number of measurements. Unlike as previously assumed, this analysis demonstrates insufficient self-preserving behaviour to justify further pursuit of this approach. Rotor wake turbulence must therefore be measured or predicted at each operating condition separately. An analysis procedure is developed by which the characteristics of individual wakes, necessary for broadband noise predictions, may be inferred from rotor wake velocity measurements in situations in which there is significant overlap between adjacent wakes. (iii) Multi mode rotor blockage Noise generated by the OGV propagates to the forward arc by passing upstream through the spinning rotor. This thesis presents a model for the sound power transmission loss associated with crossing the rotor that includes modal frequency scattering effects. It is shown that the results obtained using exact cascade scattering closely agree at low and high frequencies with the results from a relatively simple prediction scheme that assumes that only plane waves propagate through the cascade, thereby ignoring modal scattering effects. The advantage of making this approximation is that the computation is considerably more efficient than a full cascade calculation. At low frequencies, where only plane waves propagate in the gap, exact agreement is obtained between the exact and plane wave models. Close agreement is also observed in the high frequency limit where a large number of cascade modes are cut-on, most of which are well cut-on and hence whose behaviour tends that of the plane wave mode. The three components of the prediction procedure outlined above are combined to perform a prediction of the rear-arc and forward-arc broadband noise from an LSFR. Comparison of the measured and predicted noise spectra are in reasonable agreement with variations with working line and fan speed being reasonably well captured.

The growing global concern over environmental emissions such as nitrogen oxides and noise sets challenging problems for aero-propulsion engineers. Acoustic waves generated by unsteady combustion not only contribute towards the overall noise transmission, but may also cause thermoacoustic instability in combustors, particularly those designed for low NOx emissions. Combustion noise is generated by unsteady combustion – either by the direct generation of acoustic waves or indirectly by the creation of entropy waves. Entropy waves by themselves are silent, but when accelerated, such as through the combustor exit, they create further acoustic waves known as entropy noise. This thesis aims to study transmitted and reflected combustion noise. Current predictions for noise transmission often assume that the wavelengths of the flow perturbations are large compared to the combustor length, known as the compact assumption. We will develop predictions for finite-length combustors accurate to first-order in frequency. The effect of the interaction between an oscillating shock wave with combustion noise is also studied analytically. The predictions agree with data from numerical simulations. Combustion acoustics reflected at the combustor exit may go on to interfere with the combustion process, setting up a feedback mechanism that may lead to thermoacoustic instability. A modified combustor model is presented to study the effect of dissipation and dispersion of entropy waves on the instability, and it was found that the extent of dissipation or dispersion not only plays a significant role on whether instability occurs, but also determines the dominant frequency of oscillations. Furthermore, analytical and numerical investigations suggest that entropy waves are convected with the flow undissipated, and that modelling improvements may be made to take entropy dispersion into account. The findings in this work provide better tools to understand indirect combustion acoustics and to analyse their importance in both transmitted combustion noise and the thermoacoustic instability experienced by low NOx combustors.

The Unsteady Vortex-Lattice Method provides a medium-fidelity tool for the prediction of non-stationary aerodynamic loads in low-speed, but high-Reynolds-number, attached flow. Despite a proven track record in applications where free-wake modelling is critical, other models based on potential-flow theory, such as the Doublet Lattice and thin-aerofoil approximation, have been favoured in fixed-wing aircraft aeroelasticity and flight dynamics. This dissertation presents how the Unsteady Vortex-Lattice Method can be re-engineered as an enhanced alternative to those techniques for diverse situations that arise in flexible-aircraft dynamics. A historical review of the methodology is included, with latest developments and practical applications. Different formulations of the aerodynamic equations are outlined, and they are integrated with a nonlinear beam model for the full description of the dynamics of a free-flying flexible vehicle, which furnishes a geometrically-nonlinear description of both structure and aerodynamics. Nonlinear time-marching captures large wing excursions and wake roll-up, and the linearisation of the equations lends itself to a seamless, monolithic state-space assembly, particularly convenient for stability analysis. The aerodynamic model and the unified framework for the simulation of high-aspect-ratio planes are exhaustively verified by comparing them to lower- and higher-fidelity approaches. Numerical studies emphasising scenarios where the Unsteady Vortex-Lattice Method can provide an advantage over other state-of-the-art tools are presented. Examples of this comprise unsteady aerodynamics in vehicles with coupled aeroelasticity and flight dynamics, and in lifting surfaces undergoing complex kinematics, large deformations, or in-plane motions. Geometric nonlinearities are shown to play an instrumental, and often counter-intuitive, role in the aircraft dynamics. The Unsteady Vortex-Lattice Method is unveiled as a remarkable tool that can successfully incorporate them in the unsteady aerodynamics modelling.

Owing to the problems caused by propellers, research has turned to the biological world for inspiration for non-propeller propulsion. Rays were chosen for further study and it was found that a key feature of their swimming is the asymmetric-in-time movements of their pectoral fins. The main goal was to determine whether asymmetric-in-time oscillations produced a larger resultant force. Two flexible fins were used (NACA and biomimetic stiffness profile “BIO”). Asymmetry was defined by the proportion of the time period taken to effect one half stroke. The experiments showed that at low frequencies, asymmetric oscillation produced greater resultant force and that this force was at an angle to the chord of the fin at rest. At high frequencies, the BIO fin produced lower resultant force when oscillating asymmetrically and the angle of the resultant force was the same as for the symmetric oscillations. There was no difference between the resultant force magnitude or direction produced by the NACA fin at high frequencies. More power was used when oscillating asymmetrically but the force efficiency, the resultant force per watt, was often the same for symmetric and asymmetric oscillations. The trailing edge kinematics of the fins were analysed. Some of the kinematics variables correlated with the resultant force magnitude independently of fin type. The wake structures behind the fins oscillating at two different frequencies were examined. The wakes were geometrically asymmetric behind both fins oscillating asymmetrically at low frequency. At the higher frequency, the wakes behind the asymmetrically oscillating fins were no different to their symmetric counterparts. Asymmetric-in-time oscillation is suggested as a method of effecting turning manoeuvres and forward propulsion with only one motor. Despite the increased energetic cost, asymmetric-in-time oscillations can produce larger steady-state forces than symmetric oscillations do.

The continual need to better aircraft performance is increasingly prompting designers towards the realisation of morphing structures. One such enabling morphing technology are "multistable composites", and interest in them stems from the fact that they are able to sustain significant changes in shape without the need for a continuous power supply. This research benefits from the tow-steering technique to develop laminates with curved fibre-paths or Variable Angle Tows (VAT) in a ply (i.e. exploiting the anisotropy of composites), to introduce residual thermal stresses capable of imparting bistability. The principle idea is that VAT laminates can impart bistability with the new feature of ensuring fibre continuity within the wider structure facilitating structural integration (i.e. blending of lay-ups across components). Additional structural strength is shown to be imparted due to the load path continuity achieved via the tow-steering technique. The thermally induced bistable behaviour of VAT laminates is investigated through Finite Element (FE) modelling and experimental studies. The effect on the stable shapes for variations in the resin layers, fibre volume fraction (Vf), resin thermal expansion and laminate thickness are reported and found to be influential (from high to low) in the same quoted order. An approach, aiding the development of well- defined finite element models that are capable of predicting the bistable behaviour of manufactured laminates, is presented. It is shown that despite the inherent variabilities in laminates, a shell model capturing sufficient detail (i.e. resin layer[s), ply thicknesses and the Vf of a ply) is successful in predicting the cured shape(s). An optimisation based design framework aimed at realising the morphing potential of VAT laminates is presented. A morphing alternative to the plain flap is achieved, and the two potential solutions: one aiming to maximise the deployed flap-angle, and the other to attain maximum flap-angle increment, are discussed. Moreover, a good agreement exists between the FE predictions and the experimental observations for the manufactured laminates, demonstrating the feasibility of the morphing flap concept.

The main aim of this study was to investigate noise reduction techniques for bluff body noise. Three methods were investigated, using a splitter plate on a fairing- strut configuration, applying flow control to the surface of a fairing and fitting a splitter plate behind a isolated bluff body. Aerodynamic tests were performed in wind tunnel facilities using particle image velocimetry (PIV), hotwire anemometry, pressure sensors and a force balance. Acoustic tests using a microphone array, on-surface microphones and freefield mi- crophones were performed to investigate the noise generated by the models. The splitter plate fitted to the fairing-strut configuration was found to be dominated by large scale vortex shedding. The addition of the splitter plate blocked the interaction between the two opposing shear layers aft of the shell's trailing edge thereby reducing their interaction with the downstream strut. Broadband noise reductions were observed as well as reduction in the noise levels of the peaks asso- ciated with the shedding. Applying flow control showed noise reductions for both cases when suction and blowing were applied. These reductions were observed at the lower tested Reynolds numbers (ReDshell = 1.75 x 105), at higher Reynolds numbers (ReDshell = 3.5 x 105) the noise reductions decreased when compared to the baseline case. The splitter plate fitted behind an isolated bluff body modified the wake, decreasing shedding frequency and drag with an increase in the splitter plate length. Broadband noise reductions were observed with all three splitter plate lengths and the tonal peak of the vortex shedding noise was suppressed. The study shed light on the possibility of achieving noise reductions using the three methods. However more research is required to apply these findings on a landing gear.

This research intends to provide new insight into the aerodynamics of wings in ground effect under dynamic motion. This work represents a new step forward in the field of race car aerodynamics, in which steady aerodynamics are well understood. As the first comprehensive study on oscillating wings in ground effect, several modes of oscillation were studied numerically, including heaving, pitching and combined motion of an airfoil and heaving of a wing fitted with endplates. A wide range of reduced frequencies were tested for the simulations at different ride heights, which showed appreciable differences with respect to a stationary wing. The flowfield around the airfoil was obtained by solving the Reynolds-Averaged Navier- Stokes equations, while Detached Eddy Simulation was used for the wing. A dynamic mesh model was implemented to adapt the grid to the wing motion. The results showed other aerodynamic mechanisms in addition to the ground effect, namely the effective incidence and added mass. Stall can be postponed to lower ride heights by increasing the frequency of heaving, while a pitching airfoil can stall below the static stall incidence when placed close to the ground. A stability analysis showed that flutter can occur at low frequencies in heaving motion but increasing the frequency always stabilises the motion. The behaviour of the vortex formed on the inboard face of the endplate is altered by the heaving motion and has an important effect on the downforce generation. Vortex breakdown can be induced or suppressed depending on the frequency and effective incidence. At high frequencies, these vortices interact with counter-rotating trailing edge vortices to form vortex loops that transform into omega vortices in the wake. Additional experiments for a stationary wing serve to qualitatively validate and complement the reference cases.

This thesis is directed at developing and assessing a multifidelity model-based methodology for the flight performance analysis and multidisciplinary optimisation of flexible wing aerofoils. Such a strategy is pursued because of the high computational cost involved of solving such optimisation problems via high-fidelity simulations only. The methodology is applied to the preliminary design of a small flexible winged Unmanned Air Vehicle (UAV), the likes of which are particularly susceptible to wind gusts. The strategy adopted is directed at optimising both the passively adaptive structure and the shape of the flexible UAV wing for aerodynamic performance (i.e., drag reduction), weight reduction and gust response alleviation, formulated as an unsteady coupled Fluid- Structure Interaction (FSI) problem. A metamodel of the high-fidelity model response, based on a tuned low-fidelity one, is built in order to verify and validate the approach for aeroelastic problems. Both models are based on solutions of the aeroelastic equations for the wing Typical Section and the low-fidelity response tuned accordingly as prescribed by suitable Design of Experiments (DOEs). Several levels of complexity and computational cost are employed for modelling aerodynamics and structural dynamics. The role of aerodynamic damping, structural nonlinearities and turbulent Computational Fluid Dynamics (CFD) is investigated. Good agreement between the high-fidelity results and corrected low-fidelity ones shows that the methodology is suitable for use m aeroelastic performance optimisation problems. Using the multifidelity strategy developed, the flexible wing of a small UAV is optimised for best flight efficiency under aero-structural constraints. The wing structure is assumed fully flexible and a semi-analytical model for the aeroelastic analysis and gust response of a flexible Typical Section developed. Having tuned the low-fidelity response, a Genetic Algorithm (GA) is employed to fmd the global optimum, showing that a flexible wing aerofoil is characterised by a higher aerodynamic efficiency than its rigid counterpart.

This study investigates the aerodynamics of nil inverted wing in ground effect, a race car wheel and the interaction between the two components, using numerical and experimental methods. The wheels were located behind the wing at flU overlap and gap of 20mm, and the wing ride height. iu the vertical direction was the primary variable. Models of 50% scale were used , giving a Reynolds number of 5.8 x 105 based on the wing chord . The Detached-Eddy Simulation model was validated against wind tunnel measurements including PIV, surface pressures and forces , where it was found to outperform a Reynolds averaged Navier-Stokes approach which used the Spalart-Allmaras turbulence model. It accurately predicted the wing vortex breakdown at low ride heights, which is of the bubble type with a spiralling tail, and the wake of the wheel. A mesh sensitivity study revealed that a finer mesh increased the amount of structures captured with the DES model, improving its accuracy.

In this work physics-based modelling of bluff-body noise was performed with application to landing gear noise production. The landing gear is a primary contributor to airframe noise during approach. Noise is primarily generated from the unsteady pressures resulting from the turbulent flow around various components. The research was initiated in response to the need for an improved understanding of landing gear noise prediction tools. A computational approach was adopted so that the noise generating physics of the problem could be captured. Governing laws were solved numerically to predict the noise source characteristics and the resulting acoustic far-field. Three-dimensional compressible Navier-Stokes simulations were performed to solve the unsteady turbulent near-field flow and the acoustic analogy was used to predict the resulting far-field acoustic pressure. The flow solver included a high-order computational aeroacoustics code adopting large-eddy simulation, whilst a Ffowcs Williams and Hawkings solver was used for the acoustic prediction. Circular cylinders in various configurations were selected to represent basic landing gear struts and results were used to form a modelling database. Initially, cylinders at various Reynolds numbers were investigated in cross-flow to determine the noise characteristics of a simple model strut. The work was extended to investigate the effect of strut alignment to the flow by simulating cylinders in yaw. The effect of yaw was shown to modify the peak level and frequency of far-field noise spectra. Component interaction effects were then investigated by simulating cylinders in tandem arrangements. The resulting aerodynamic and far-field noise characteristics were shown to be complex and extremely sensitive to the separation distance between the cylinders. Finally, a prediction model was developed and validated by comparing predictions against theory and measurements of the noise radiated by a simple two-wheel landing gear model. The results demonstrated the capability of the model to accurately predict correct spectral and directivity characteristics.

Flight flutter testing is a mandatory test performed to demonstrate that prototype aircraft are free from flutter - a violent destructive vibration. Flight flutter testing techniques have advanced a great deal over the last 50 years but the process is still risky, time consuming and costly. Therefore, the demand for reduction in time and cost of the flight test procedure and development of efficient online flutter prediction tools among other related issues is ever-growing. This research has dealt with two of the key issues and the methods developed have been validated successfully on simulated aeroelastic data sets corrupted with various levels and types of noise.

An experimental and computational study was made of a single circular jet impinging onto a flat ground board. A 1/2' nozzle running at a fixed nozzle pressure ratio of 1.05 was used in the experimental phase (giving an nozzle exit Reynolds number of 90xlO'), the nozzle to ground plane separation being varied between 2 and 10 nozzle diameters. Measurements were performed in the free and wall jets using single and cross-wire hot-wire anemometry techniques and pitot pressure probes in order to detemine mean velocity and normal and shear stress distributions. Some analysis is also presentedo f earlier measurementso n high pressurer atio impinging jets. Nozzle height was found to effect the initial thickness of the wall jet leaving the impingement region, increasing nozzle to ground plane separation increasing the wall jet thickness, although this separation distance did not seem to affect the rate at which the wall jet grew. Nozzle height was also found to have a large effect on the peak level of turbulence found in the wall jet up to a radial distan ce from the jet axial centre line of 4.5 nozzle diameters, after which the profiles become self-similar. Lowering the nozzle tended to increase the peak level measured in all the turbulent stresses within this development region. The production of turbulent kinetic energy in the wall jet, which is an indication of the amount of work done against the mean flow by the turbulent flow was found to increase dramatically with decreasing nozzle height. This was attributed to greater shearing of the flow at lower nozzle heights due to a thinner wall jet leaving the impingement region. A moving impingement surface was found to cause separation of the wall jet inner boundary layer on the 'approach' side leading to very rapid decay of peak velocity. The point of separation was found to occur at radial positions in the region of 7.0 to 8.0 nozzle diameters, this reducing slightly for lower nozzle heights. A parametric investigation was performed using the k-e turbulence model and the PHOENICS CFD code. It was found that due to inadequacies in the model, it failed to predict accurately the growth of the wall jet, both in terms of its initial thickness and the rate of growth. It did, however, predict an increase in wall jet thickness with both increasing nozzle height and exit turbulence intensity and decreasing nozzle pressure ratio. Modifications were made to the constants in the model to try and improve the predictions,w ith a limited degreeo f successT. he low Reynoldsn umber k-F-t urbulence model was shown to give a slightly improved non-dimensional wall jet profile, although this did not improve the predicted rate of growth of the wall jet.

The computation of high-lift flows poses considerable challenges to the aerodynamicist. The work presented in this thesis describes the development of an efficient unstructured grid generation method for high lift flows and the application of Differential Reynolds Stress Models (DRSM). The method described in this thesis, separates the grid generation process into two steps: generation of the anisotropic regions and isotropic grid generation. All surfaces are triangulated as part of the anisotropic grid generation process. The user can also specify the wake regions. Using surfaces or wakes, the algorithm computes the normals to the surfaces and distributes new points along those directions. This initial grid is connected to the outer boundaries. The cells connecting the anisotropic grid with the outer boundaries are continuously refined using an edge-splitting algorithm that creates new computational cells, grid points and cell connectivities simultaneously. The second part of this research involves the assessment of two DRSM against a k - e model. The validation study included the Launder-Rodi-Reece (LRR) DRSM and the Speziale-Sarkar-Gatski (SSG) DRSM. The Reynoldsaverage Navier-Stokes (RANS), are solved explicitly using a cell centred finitevolume scheme. The focus of the validation study, are high-lift configurations tested as part of the U.K. National High-Lift Programme. The aerofoils considered include a 3, 4 and 5-element configuration, designated NHLP L1/T2, NHLP L1/T7 and NHLP L1/T8 respectively. The results for the L1/T2 and L1/T7 aerofoils show only marginal differences on the predicted aerodynamic loads. The results for the L1/T8 show significant differences between both types of models, with the LRR exhibiting the closest agreement with experimental data. Results indicate that only in these extreme cases, the advantages of a Differential Reynolds Stress Model influence the aerodynamic loads results. However in most cases the DRSMs produced a better representation of the flow physics.

Noise is an environmental concern and due to the increasing interest in helicopters as an alternative inter-city transportation, research for more environment friendly helicopters is continuously growing. Building on this demand, this study aims at finding an efficient and accurate noise prediction tool for rotating sources. This study therefore investigates the modelling of noise from rotating sources such as helicopter rotors by addressing noise propagation in both subsonic and transonic/ supersonic regimes. The aim of this research is to explore the field of aeroacoustics prediction for rotor generated noise and to develop a noise prediction tool for sources moving in subsonic or supersonic flow regimes. The aeroacoustics predictions presented have been obtained using a hybrid approach. With such an approach the near field noise generation process is simulated by means of an aerodynamics prediction tool while the noise propagation to the near field is computed by mean of the Ffowcs Williams-Hawkings equation in time domain. For the near-field aerodynamic calculations di erent CFD tools have been exploited. More precisely, three test cases have been analysed. For the first test case of 2D aerofoil-vortex interaction, reproducing the experimental campaign of Lee et al., the near-field is computed via the commercial software Fluent. The unsteady implicit Euler solver with second order discretisation both in space and in time is exploited. This uses the ROE FDS scheme for the fluxes calculation. The same solver is used in the near-field simulations of the third test case, where the analysis of a non-lifting hovering rotor is carried out in delocalised conditions, reproducing the experiments of Purcell on the UH1H model rotor. The second test case analysed is based on the HELISHAPE experimental campaign for the ONERA model rotor in BVI conditions. Two comprehensive codes, from Agusta-Westland and Roma Tre, are used to simulate the complex aeromechanics of the rotor in low speed descent. The noise propagation phase has been performed via the new noise prediction tool developed during this study, named HelicA (for Helic-opter A-coustics). This tool is based on the Emission Surface formulation and exploits a novel root finder and Emission Surface construction algorithms. It can use control surfaces which are in subsonic or transonic/supersonic conditions. Verification and validation processes have been performed on the noise prediction tool before using this code in the aforementioned test cases. These processes are based on the comparison of the tool’s predictions with available analytical and numerical results. The verification and validation cases include sources moving at Mach numbers ranging from MT = 0 to MT > 1. The noise prediction tool is applied to the three aforementioned test cases and the results are in very good agreement with the measurements even for the strong shock delocalisation cases.

Flapping-wing micro air vehicles, based on insect-like apping, could potentially ll a niche in the current market by o ering the ability to gather information from within buildings. The aerodynamics of insect-like apping are dominated by a large, lift-enhancing leading-edge vortex (LEV). Historically, the cause and structure of this vortex have been the subject of controversy. This thesis is primarily intended to provide insight into the LEV, using computational uid dynamics coupled with validating experiments. The problem is simpli ed by breaking down the complex kinematics involved in insect-like apping and examining only a part of these kinematics; rstly in 2D, before progressing to 3D sweeping wing motions. The thesis includes discussion of published literature in the eld, highlighting gaps and inconsistencies in the current knowledge. Among the contributions of this thesis are: descriptions of the e ects of changing Reynolds number and angle of attack for 2D and 3D ows; clari cation of terminology and phenomenology, particular in the context of 2D ows; and detailed descriptions of the development and structure of the LEV in both 2D and 3D cases, including discussion of Kelvin-Helmholtz instability. The issues of Strouhal number, delayed leading-edge separation, dynamic stall and the Wagner e ect are also considered. Generally, the LEV is shown to be unstable in 2D cases. However, in 3D cases the LEV is seen to be stable, even if Reynolds number is increased. The stability of the LEV is found to be critically dependent on wing aspect ratio.

Insect-inspired micro air vehicles (MAVs) have the capacity for higher lift forces and greater manoeuvrability at low flight speeds compared to conventional flight platforms, making them suitable for novel indoor flight applications. This thesis presents development studies of an actuated flapping mechanism for an insect-inspired MAV. An original theoretical understanding has shown that the kinematical constraint of a flapping mechanism fundamentally determines its complexity and performance. An under-constrained mechanism is optimal but almost always requires a linear input. A power optimisation study has demonstrated that the only technologically mature actuation devices with viable power densities for flight are rotary. Consequently, previous airborne flapping MAVs utilised constrained rotary-input mechanisms which require conventional control surfaces that significantly reduce flight manoeuvrability.

Although significant progress has been made in the study of dynamic loading of aircraft structures, several areas have been identified that require further research. In particular, attention is drawn to problems involving transient, dynamic fluid-structure interaction, where fluids play an important role, heavily influencing the response of the structure to the applied dynamic load. In this work the use of existing numerical modelling techniques for the evaluation of such problems is investigated.

The noise produced by low speed flows over cavities with and without plasma actuators were studied using experimental techniques. The acoustic and flow environments of open cavities with length-depth ratios of 0.5, 1.0, 2.0 and 3.0 have been investigated. Measurements using oil flow, particle imaging velocimetry, boundary layer surveys and surface mounted microphone measurements were taken. Experiments were conducted in a low speed wind tunnel running at freestream velocities between 10 m/s and 20 m/s corresponding to Reynolds numbers, based on the depth of the cavity, of 3.6 x 104 to 7.1 x 104. An array of paraelectric plasma actuators were located on the approaching surface to the cavity, aligned with the direction of the oncoming flow. Results show that the paraelectric plasma actuators can produce a significant attenuation of the dominant cavity mode. The particle imaging velocimetry surveys around the electrode elements reveal vortical structures produced by the plasma actuators. It is shown that these structures convect downstream with the mean flow and produce disturbances similar to that of vortex generators. This affects the convection of spanwise discrete vortices in the cavity shear layer, disrupting the mechanisms that allow the cavity to produce tones. Results also show that for any given geometry and freestream flow speed the attenuation of the dominant mode reduced with increasing voltage to the actuators.

The effect that wavy and flat surfaces have on the aerodynamics of aerofoils operating in Ground Effect (IGE) has been investigated. Experiments were conducted in the Southampton University rolling road wind tunnel and the Circulating Water Channel (CWC) at QinetiQ Haslar. Two aerofoils were tested a DHMTU 12-35.3-10.2-80.12.2 and a NACA 0012. The DHMTU aerofoil was the primary section of interest to this research and the NACA 0012 was used as a control. Force measurements, flow visualisation and pressure measurements were used to assess the DHMTU aerodynamic characteristics in ground effect. The operating Reynolds Number was 830,000 and 550,000 in the wind tunnel and CWC respectively. Wings operating at these low Reynolds Numbers are applicable to a novel small Unmanned Wing In Ground Effect Vehicle (UWIGV) describe in this thesis. Analysis of data obtained in the rolling road wind tunnel showed that the overall drag of both the DHMTU and NACA 0012 aerofoils is greater in ground effect than out of ground effect. A significant reduction in the vortex drag of both aerofoils is produced as a result of flying in ground effect. Unfortunately the significant reduction in vortex drag coupled with the increase in the zero lift drag is not sufficient enough to reduce the overall drag of the trail sections in ground effect. As the altitude of both the aerofoils decreases the lift increases. This showed that the increase in aerodynamic efficiency as a result of flying IGE is due to the increase in lift. Analysis of simulated flight over wavy surfaces in the CWC shows that as a wing flies over a wavy surface the lift varies from a maximum to a minimum. The amplitude of the lift force oscillation increases with decreasing altitude. A general result is that as the wavelength of the surface decreases the maximum lift experienced by the wing decreases and the minimum lift increases. The drag on the wing over a wavy surface in ground effect varies with altitude and as a function of wing position over the wavelength. Like lift it varies between a minimum and maximum value. The positions over the wavy surfaces where minimum and maximum drag was recorded are identical to the extremes of lift.

This thesis details a feasibility study into the use of active control for the attentuation of the low frequency components of the 'buzz-saw' sound field radiated from an aircraft engine on take off. Passive liner sections are currently used to attenuate the sound field, however, they have insufficient thickness to attenuate sufficiently the very low frequency components of the sound field. An active control system positioned after the liner section could be used to further attenuate the sound field. This thesis shows that due to the simple modal structure of the buzz-saw noise, effective control of the sound field can be achieved by minimising the sum of the square pressures at a single ring of error sensors using a single ring of control actuators. In the absence of noise and extraneous modes, the level of control performance is fundamentally limited by the level of the evanescent modes at the error sensor array. An analytic model is developed which predicts the sound power reduction. Reflections are shown to be important in the control mechanism with standing wave regions identified between the secondary source array and the exhaust termination and between the inlet termination and the error sensor array. The thesis concludes with the results from an experimental rig. A real-time FXLMS algorithm is used to control a synthesised low order spinning mode by minimising the sum of the squared pressures measured at a single ring of 7 sensors using a simple ring of 7 control actuators. The relationship between sound pressure reduction at the error sensors and the transmitted sound power is investigated. Sound power reductions of up to 14.5dB are achieved. The control mechanism is identified from computer simulation.

This thesis presents the development and validation of numerical methods for the study of Blade-Vortex Interaction. Aspects addressed in this work include the aerodynamics and aeroacoustics of the interaction between a vortex and an aerofoil. The phenomenon of Blade-Vortex Interaction (BVI) is central to the study of the aerodynamics of rotors as well as to the calculation of the acoustic field radiated by rotor craft. The simulation of BVI is challenging since the solution scheme tends to alter the characteristics of the vortex which must be preserved until the interaction. The basis of the present thesis has been the code developed at the University of Glasgow. Some numerical improvements have been carried out to allow the simulation of BVI. First, the numerical developments concerned the time discretisation with the comparison of two different implicit schemes for their robustness and reliability. The implementation of an implicit unfactored method allowed better results in terms of convergence. Secondly, the Compressible Vorticity Confinement Method (CVCM) has been implemented into the solver to allow the preservation of vortical flows. The CVCM has been tested and validated on a benchmark problem for the case of vortex convection. The use of the CVCM was found to be capable of preserving the vortex characteristics assuming the optimum confinement parameter was chosen for a given grid. Hence the use of the CVCM made the simulation of BVI possible. The capabilities of the CVCM were assessed with the simulation of head-on and miss-distance BVI cases. Results were compared against experimental surface pressure measurements and flow visualisation data. Good agreement was obtained. It appears that the CVCM is useful for preserving the characteristics of vortices on coarse grids.

Non-linearities within structures often present difficulties when developing algorithms to analyse their dynamic properties. Developing a combined aerodynamic and structural- aeroelastic - code is an example where non-linearities can induce particular characteristics as the presence of aerodynamic non-linearities can compound the complexity of the analysis. Furthermore, when non-linearities occur within actuation devices the impact of coupling control systems with the aeroelastic algorithms - aeroservoelastic - must also be considered. In this work, new methods of analysing aero(servo)elastic systems containing various structural non-linearities are studied. The first technique is used to analyse piecewise linear systems. In this method, aeroelastic equations are recast in a form where the independent variable is the time at which the system reaches a discontinuity, sets of these equations are then combined to form an algebraic set of equations describing a Limit-Cycle Oscillation (LCO). The second technique is applied to more general non-linearities by approximating any discrete non-linearities with trigonometric functions, creating a set of continuous Ordinary Differential Equations (ODEs). For both methods, computational efficiency is achieved by applying numerical continuation to track solution branches. The models analysed in this work are two and three degree-of-freedom aerofoil sections containing non-linearities in their heave, pitch and/or flap freedoms. Four different aerodynamic representations are used, two incompressible codes establish the accuracy of the new methods. The other codes are used to study transonic flows and show good agreement with work based on aeroelastic systems with both linear and non-linear structures. Three different control laws - fixed gain, optimal and adaptive - are also investigated to assess their ability to delay flutter onset and suppress LCOs. Optimal control showed the best overall ability to achieve these aims, although it was found that care must be taken not to destabilise areas below the flutter boundary. Finally, a method of analysing fatigue due to structural non-linearities is investigated. The analysis combines the numerical continuation techniques with the Rainflow method to quantify damage due to simple acceleration-deceleration profiles.

The work presented in this thesis attempts to provide a deeper understanding of the physical phenomena associated with the dynamic stall process on finite wing planforms. The work involves the analysis of data from the Glasgow University unsteady aerodynamics database that has been built up over a number of years through contributions from a range of researchers. Analysis focuses on two finite wing models; one a rectangular wing of aspect ratio three and the other with the same overall dimensions but with 60o swept tips. However, as most research to date has focused on nominally two-dimensional data, the results are referenced to measurements made on a nominally two-dimensional NACA 0015 aerofoil model. This is appropriate as this aerofoil was used as the wing section of both of the three-dimensional wing models. Flow visualisation images collected in a previous study also provide valuable information to supplement the pressure analysis. It is shown that, although the flow at the mid span sections of the finite wings exhibit many of the features of the two-dimensional case, there are some significant differences. In particular, the three-dimensional flow is dominated by the downwash from the wing tips. This causes the normal force response during pitching to lag the static normal force curve. This is in complete contrast to the two-dimensional case where the shed vorticity induces the opposite effect. The downwash also influences the incidence of lift stall but it does so in a manner that is dependent on the reduced pitch rate. Despite these effects, it is established that the flow behaviour in the mid-span region is almost two-dimensional prior to vortex inception. This provides an opportunity to examine the relationship between the generation of vorticity, or vorticity flux, in the leading edge region and the origins of the dynamic stall vortex at specific span locations in location. The vorticity flux distributions around the leading edges of the nominally two-dimensional NACA 0015 aerofoil and the two finite wings are then examined for pitching cases. On this basis a link is established between the peak vorticity flux and the dynamic stall vortex formation. This is confirmed by comparison of the vorticity flux measurements with a previous dynamic stall vortex detection method. The two methods are shown to five almost identical results in situations where the flow may be considered nominally two-dimensional. This suggests that monitoring vorticity flux may provide a practical method of dynamic stall vortex detection. In regions of the finite wings that exhibit strong three-dimensional flow effects, i.e. away from the mid-span, the peak vorticity flux is achieved after the dynamic stall vortex forms. This suggests that vortex formation is triggered by interference from adjacent sections of the wing. To examine this possibility, the vorticity flux is compared to a criterion used to detect the initial instability of the boundary layer at the leading edge. It is shown that the relationship between this criterion and the peak vorticity flux is the same along the span of the wing. This is a significant result as it demonstrates that, although the leading edge response determines the incidence of vortex onset near the mid-span, the formation of the vortex on sections of the wing closer to the tips occurs before the leading edge becomes critical. The implications of this for dynamic stall modelling of two-dimensional dynamic stall predictors with lifting line formulations will not capture this effect.

MEMS actuators and their effect on boundary layers is investigated using numerical simulation. The thesis is specifically focussed on jet actuators and their application to the targeted control of turbulent boundary layers. A complete numerical model of jet-type actuators, including the popular synthetic-jet actuator, is developed. The assumed input is the voltage signal to the piezocermic driver and the calculated output is the exit jet velocity. Thorough validation of the numerical code is presented and simulations performed to highlight the key issues in MEMS-actuator design. The three-dimensional boundary-layer disturbance created by the MEMS actuator is modelled using a velocity-vorticity formulation of the Navier-Stokes equations. The parallel code is rigorously validated against results from linear stability theory and transitional-streak measurements. The boundary-layer code is used to determine a performance criterion for MEMS jets; it is shown that the net mass flow from a jet best determines its effectiveness. The code is also used to demonstrate the macro-scale capabilities of MEMS-scale actuators; a grid-scaling method is described and employed to facilitate this calculation. A method is presented that enables high- and low-speed streaks to be modelled economically in otherwise undisturbed mean flows. Using this model, the fundamental principles of targeted control using MEMS actuation are explored. The MEMS-actuator and boundary-layer models are coupled, and an investigation into the interactive effects of the two systems is described. Using the coupled code, disturbances in the boundary layer are shown to induce velocities in inactive devices. One special case occurs when an oscillating pressure field creates Helmhotz resonance within the cavity of a MEMS actuator, thus causing large mass flow rates in and out of the device. It is also suggested that the MEMS device could strongly interact with the random fluctuations of a turbulent boundary layer, leading to highly unpredictable actuator responses.

Coherent flow structures and turbulent spot formation, induced by wall perturbations in a Blasius boundary layer, were studied experimentally. The study has specific engineering relevance in the area of turbulent boundary layer control to achieve drag reduction and in the wider context of improved understanding of the fundamental mechanics of turbulent generation and dynamics. Experiments were carried out in a re-circulating water channel, and the measurements and observations were made in a zero pressure gradient, laminar boundary layer, formed along the horizontal upper surface of a smooth glass plate, suspended in the channel flow. Controlled 3-D disturbances were introduced into the boundary layer flow by fixed and transitory wall perturbations. The fixed perturbations were in the form of a low-profile spherical cap. Transient perturbations were generated by the transitory inflation of a circular rubber diaphragm, 30mm in diameter, mounted flush with the wall. The diaphragm is deformed into a low profile spherical cap using a computer controlled inflation system. This allows flexible variation of the maximum inflation height (amplitude) and the height/time function shape through the variable period of the perturbation cycle. Coherent vortical structures, induced by both perturbation types are compared and vortex shedding and formation characteristics are discussed. A three-component laser Doppler anemometer has been used for the quantitative measurement of the three-dimensional velocity and vorticity field downstream of the perturbation. Special data-processing techniques have been developed to derive spatial mappings of the corresponding 3-D vorticity field. Hydrogen bubble and dye tracers were employed for flow visualization of the hairpin vortices analyzed from digital recording of high-speed video. A general parametric study of the influence of Reynolds number, perturbation diameter, maximum amplitude, height/time function shape and period on the downstream position of the induced turbulent spot is presented. The results of detailed investigations on the effect that these parameters have on the formation and intensification of the hairpin vortices shed or formed downstream from the protrusion are presented. These observations are linked to the observed subsequent chaotic breakdown of the flow to turbulence and the global behaviour of the turbulent spot formation. Experimental results are compared with previous observations obtained by analytical and numerical simulations.

Passenger cars are bluff bodies and are prone to unsteady phenomena with scales comparable to the scale of the vehicle itself. This type of large-scale, self-excited unsteadiness is the subject of the present work. Aerodynamic unsteadiness can be important for two reasons. It can cause unsteady pressures and forces on the car and it can impact the time-averaged flow through the generation of Reynolds stresses. A range of parametric two-dimensional bodies have been used in the development of novel experimental techniques and analyses and for CFD validation. Detailed investigations have been undertaken on the Ahmed model and on models of a Rover 200 passenger car in wind tunnels at Durham and at MIRA at scales of up to 40%. A method was developed which makes it possible to visualise periodic flow structures from measurements made sequentially in the wake or on the model surface. Unsteady flows for fastback passenger cars were found to be much less periodic than for two-dimensional vortex shedding cases. Pressure fluctuations were significantly lower on the model surface than in the wake resulting in limited unsteady forces. Unsteady flow structures, Strouhal numbers and levels of unsteadiness were similar for the Rover 200 model with and without a backlight spoiler and for the Ahmed model, indicating that sharp corners do not have a dominant effect on unsteadiness. Two principal unsteady structures were observed in the wake of the fastback shapes. A structure was observed at Strouhal numbers around 0.1 involving the alternate strengthening of the two c-pillar vortices in an antisymmetric mode. At Strouhal numbers in the range 0.3 to 0.6 an unsteady structure was observed consisting of the oscillation of the strength of the two c-pillar vortices in a symmetric mode. At the same time the location of the vortices oscillates in the vertical direction.

The results of an experimental investigation into the control of vortex shedding in the wakes of bluff bodies are presented. The main aim of this investigation was to gain further insight into how the wake structure of nominally 2D bodies such as circular cylinders could be modified by the introduction of some spanwise geometric disturbances. Types of models studied were those with a fixed separation point, such as square sections and thin plates and also those with moving separation points such as circular cylinders. Experiments were conducted in both air and water over a large range of Reynolds numbers (101 - 3.4x 105) using four wind tunnels and two water tank facilities. Quantitative measurements have shown a significant drag reduction on modified bodies of up to 47% in comparison with the unmodified form. Results from particle image velocimetry (PIV) and flow visualisation have shown that even a small amount of three-dimensionality can produce large changes in the wake width and structure along the span. It is shown that a common feature of these perturbed bluff body shapes is that they set up significant spanwise velocity components that prevent the wake, from rolling up into the conventional 2D wake of von Karaman vortex shedding. This in turn leads to a steady near symmetrical wake with none of the adverse effects associated with vortex shedding. Applications of the results of this research to practical engineering structures are also discussed. A 3D cylindrical body is developed with elliptically shaped surface protrusions that shows an absence of vortex shedding in the wake regardless of flow direction, achieving a 25% drag reduction over a conventional 2D circular cylinder. Results are also presented which indicate that these bodies may be useful in the attenuation of vortex-induced vibrations as the amplitude of oscillations is significantly reduced when sufficient spanwise three-dimensionality is present.

This thesis presents the results from a series of numerical experiments, designed to study the fundamental wake vortex dynamics that cause the suppression of the Karman shedding and lead to a drag reduction of the flow past a bluff body with a wavy stagnation face. The simulations have been performed using the spectral/hp element method over a range of Reynolds numbers from 10 to 500 past wavy square and circular cross-section cylinders. Starting from fully developed shedding past a straight cylinder at a Reynolds number of 100, a sufficiently high waviness is impulsively introduced resulting in the stabilisation of the near wake to a time-independent state. It is shown that the spanwise waviness sets up a cross-flow within the growing boundary layer on the leading edge surface thereby generating streamwise and vertical components of vorticity. This redistribution of vorticity and the subsequent development of the three-dimensional shear layers lead to the breakdown of the unsteady and staggered Karman vortex wake into a steady and symmetric near wake structure. The steady nature of the near wake is associated with a reduction in total drag of about 16% compared with the straight, non-wavy cylinder. Further increases in the amplitude of the waviness lead to the emergence of hairpin vortices from the near wake region. This wake topology has similarities to the wake of a sphere at low Reynolds numbers. It is found that wavy circular cylinders need to be deformed more than wavy square-section cylinders in order to suppress the Kârmân vortex shedding. At higher Reynolds numbers, the drag reduction is substantially increased. For example, at Re = 200 and 500, the drag reduction is 20% and 34%, respectively. Two methods based on three-dimensional forms of surface bleed for suppressing the Kârmân shedding past straight, non-wavy cylinders are also investigated.

High-speed unsteady spiked body flows were simulated and analysed by using Computational Fluid Dynamics (CFD). The model geometry was a forward facing cylinder equipped with a protruding spike. Various spike lengths were considered between L/D=1.00 and 2.40 in both flxed spike and moving spike configurations. Axisym- metric, laminar flow conditions were considered at two freestreams of Mach 2.21 and Mach 6.00 and Reynolds' numbers based on the blunt body diameter of 0.12 x 10.

A series of experiments were undertaken at the Defence Evaluation Research Agency (DERA) Farnborough, within the Aero Physics group into the phenomenon called Jet Interaction (JI). Jet Interaction (JI) is produced by the interaction of a jet with the external flow around a vehicle. This study focused on investigating the effects of a divert thruster employed to provide a vehicle with a rapid divert capability on the external flow-field and on the induced forces and moments exerted on the vehicle by the jet. The research was based on studying the effects on a hypersonic interceptor sometimes referred to as a KKV (Kinetic Kill Vehicle). There are many important parameters in JI. One of these is the jet Amplification Factor (AF). This is caused by the deflection of the free-stream around and over the jet and results in a pressure increase on the vehicle surface which adds to the divert thrust force. The experiments were carried out in the intermittent hypersonic gun tunnel, at a free-steam Mach number of 12.1. This produced a Reynolds number (based on diameter) Red of 300,000 with a one-tenth scale vehicle; these conditions correspond to a full-scale vehicle flying at an altitude of 41 km. To simulate the divertthruster, nitrogen was supplied to the model through a purpose-made force balance. Measured forces ineluded normal, axial and side as well as pitch and yawing moments. The experimental results have been compared with that of a full size vehicle featuring a 2kN divert thruster. The results have then been matched to the effective altitude as a function of the thrust coefficient (Cr). This allows the experimental data to be interpreted for a full-scale vehicle in order to answer design questions important to system engineers. The results from this investigation show that the effectiveness of a divert jet is influenced by the vehicle's altitude, achieving a negligible increase in AF with a Cr > 2.5 at an altitude > 50 km. The seeker will suffer from jet induced problems at low Cr levels for a. = 10°. An increase in Cr causes the separation region in front of the jet to extend to the nose of the tested configuration for M1 = 12.1 while complete separation is achieved at Cr > 1.2. Injection Mach number (MJ) has a small influence on AF. However it does not influence the separation region. Penetration height (h) of the jet is increased for higher Mach number injection. Both AF and the separation region are influenced by nozzle geometry. A series of different nozzle geometries were tested. These had the effect of reducing the measured amplification factor to a maximum of 1, except for the dual circular orifice combination, which doubled the measured AF achieved for a single circular orifice. The influence of nozzle geometry reduced the Cr levels required to produce a negligible increase in AF and the corresponding altitude. The angle of attack (œ) has a strong influence on AF at low Cr levels, however it becomes negligible when complete separation is achieved.

Two-dimensional (2-D) cavity flow physics and oscillation control were investigated through solutions of the Reynolds-Averaged Navier-Stokes equations coupled with a two-equation k-ω turbulence model. Effects of the leading edge modifications including compression ramps, expansion surfaces and mass injection on supersonic cavity flow oscillation were investigated. Different flow mechanisms were observed at Mach 1.5 and 2.5. The study proposed a different explanation of the flow control mechanism when mass injection was used. An optimal mass injection rate was identified. A further improvement on the 2-D model was made by considering the effect of the turbulent viscous sub-layer using the Integration-to-the-wall boundary condition. The results confirmed that the shear layer instability reduced gradually from Mach 1.5 to 3.5. An improvement in SPL prediction was achieved. Dominant modes were also correctly predicted. The capability of the model was extended for the investigation of the 3-D compressible unsteady turbulent flow physics. It was validated against a hypersonic symmetric corner flow. The turbulent effect was modelled by a two-equation k-ω turbulence model. A laminar cavity flow oscillation at Mach 1.5 was predicted. The result showed a self-sustained pressure oscillation. The predicted pressure oscillation was dominated by the second mode and its frequency was 5702H z which was close to the measured value of 5900H z. The SPL discrepancies with the measurements were within 2.3dB. A secondary symmetric flow pattern inside the cavity displayed a 3-D effect and showed the effect of the side wall within a spanwise distance of 2D. Further simulation of the turbulent cavity flow at Mach 1.5 showed a weak pressure oscillation, which indicated the RANS is sensitive to the choice of turbulence model. However, the surface flow pattern and surface pressure distribution were consistent with the measurements. The strongest surface pressure oscillation was observed near the rear corner on the centre plane. A secondary symmetric flow pattern also existed. A spanwise wave was found in the cavity. At the trailing plate, a flow separation was formed in the laminar boundary layer but was not observed in the turbulent boundary layer.

There are problems specific to the helicopter/ship dynamic interface, which limit helicopter operations. Amongst these is the difficulty associated with landing on a moving platform. The ship airwake, which includes large velocity gradients and areas of turbulence, is considered a crucial factor in limiting these operations. For this reason knowledge of the air flow around the ship and through the helicopter's rotors is necessary to understand the problems the helicopter encounters as it lands and takes off. A CFD model of a hovering helicopter main rotor is developed to examine airflow in the presence of ship structures and side winds. The rotor is modelled by modifying the governing Navier-Stokes equations in the region of the disc. The extra terms added to the governing equations apply a downforce to the fluid; these forces are independent of the flow around the rotor and equal to the helicopter weight. The boundaries of the computational domain are also modified in order to generate a physically correct solution. Flow solutions in both two and three dimensions are achieved using the commercial flow solver CFX. The flow solutions exhibit very good correlation with established momentum and power principles. The rotor model is also flown in steady horizontal flight. The resultant flow solutions agree with theoretical flow fields thus proving the validity of the rotor model. An extensive sensitivity study of CFD grid and solver parameters is also presented. This ensures that all flow solutions achieved are of the highest fidelity but are reached in a computationally efficient manner. The turbulence models are adjusted to produce solutions which agree with wind tunnel data. CFD flow solutions are presented which correspond to full-scale version of experimental studies on bluff bodies in wind tunnels. The results show that qualitative features of the wind tunnel flow regimes are recognised and resolved by the computational solution. The CFD also agrees with the quantitative data where available. Finally the helicopter rotor model and the ship model are combined to yield one flow solution, which cannot be achieved by superposition. The resultant flow yields valuable data about the induced velocities at the rotor which ultimately determine the control pitch and power required to maintain the hover in a given location.

An aircraft in low-level flight experiences a large increase in lift and a marked reduction in drag, compared with flight at altitude. This phenomenon is termed the 'wing-in-ground' effect. In these circumstances a region of high pressure is created beneath the aerofoil, and a pressure difference is set up between its upper and lower surfaces. A pressure difference is not permitted at the trailing edge and therefore a mechanism must exist which allows the pressures above and below to adjust themselves to produce a continuous pressure field in the wake. It is the study of this mechanism and its role in the aerodynamics of low-level flight that forms the basis of our investigation. We begin in Chapter 2 by considering the flow past a thin aero-foil moving at moderate distances from the ground, the typical ground clearance a being of order unity. The aforementioned mechanism is introduced and described in detail in the context of this inviscid problem. Chapter 3 considers the same flow for large and small ground clearances and in the later case shows that the flow solution beneath the aerofoil takes on a particularly simple form. In this case the lift is shown to increase as a-1. In Chapter 4 we focus on the flow past the trailing edge of an aerofoil moving even nearer the ground, with the ground just outside the boundary layer. We show that in this case our asymptotic theory for small a is consistent with a 'triple-deck' approach to the problem which incorporates ground effects via a new pressure-displacement law. The triple-deck ground-interference problem is stated and solved. In Chapter 5 we investigate the case where the aerofoil is so near the ground that the ground is inside the boundary layer. Here the moving ground interacts with the aerofoil in a fully viscous way and the non-linear boundary layer equations hold along the entire length of the aerofoil. Again a pressure difference at the trailing edge is not permitted and this produces upstream adjustment back to the leading edge. Regions of reversed flow can occur and their effects, with regard to downforce production and racing car undertray design, are considered. In Chapter 6 we consider 'wing-in-tunnel' effects.

In this work, a code has been developed that solves the Navier-Stokes equations using the finite volume method with unstructured triangular grids. A cell-centred, finite volume method is used and the pressure-velocity coupling is treated using both the SMTLE and the MAC algorithms. The major advantage of using triangular grids is their applicability to complex geometry. A special treatment is developed to ensure good quality triangular elements around the boundaries. The numerical simulation of incompressible flow at low Reynolds number is studied in this thesis. A code for generating triangular grids using the tri-tree algorithm has been written and an adaptive finite volume method developed for calculating laminar fluid flow. The grid is locally adapted at each time step, with grid refinement and derefinement dependent on the vorticity magnitude. The resulting grids have fine local resolution and are economical in reducing the numerical simulation time. The discretised equations are solved by using an iterative point by point Gauss-Seidel solver. For calculating the values of velocity and pressure at vertices of triangular grids, special interpolation schemes (averaged linear-interpolation and scattered interpolation) are used to increase the accuracy. To avoid the well known checkerboard error problems, i. e., the oscillations occurring in the pressure field, third derivative terms in pressure, first introduced by Rhie-chow (1983), are added to the mass flux velocity. Convective terms are approximated using a QUICK (Quadratic Upstream Interpolation for Convective Kinematics) differencing scheme which has been developed here in for unstructured grids. Three cases of two-dimensional viscous incompressible fluid flow have been investigated: the first is channel flow, in which the numerical results are compared with the analytical solution; the second case is the backward-facing step flow; and the third case is flow past circular cylinders at low Reynolds number (Re). The numerical results obtained for the last two cases are compared with published data. The evolution of vortex shedding is presented for the case of unidirectional flow past a circular cylinder at Re=200. In addition, drag and lift force coefficients are calculated and compared for single and multiple cylinders in unidirectional flow.