Requirements to improve the propulsive efficiency of aircraft engines have revived interest in contra-rotating open rotor (CROR) engines. However, CRORs exhibit tonal and broadband noise emissions which are a barrier of entry into commercial service. Until recently, studies have concentrated on tonal noise emissions because they are considered to be louder than broadband emissions. However, recent work has shown that CROR broadband noise can also be significant. This project uses computational aeroacoustic methods to study the mechanisms underlying broadband rotor-wake interaction (BRWI) noise, which is a dominant CROR broadband noise source. Predictions of BRWI noise have thus far assumed the rotor blades to be flat plates, and have not accounted for the effects of the blade geometry on the noise. In this project, the effects of blade geometry on turbulence interaction noise are comprehensively studied, including the effects on the noise due to airfoil thickness, leading edge radius, angle-of-attack and camber. Airfoil thickness and leading edge radius are shown to reduce the noise at high reduced frequencies, with the noise being more sensitive to thickness than to leading edge radius. The effects of angle-of-attack and camber are found to be small for interactions between an airfoil and isotropic turbulence. The mechanisms which cause the changes to the noise are also investigated. The project concludes with a study of the turbulence interaction noise for realistic CROR blades and conditions to evaluate the error incurred when rotor blades are assumed to be flat plates. It is found that while CROR blade geometry does affect BRWI noise, the effect is sufficiently small that flat plate theory can be confidently used in most circumstances. In addition to examining the effects of airfoil geometry on the noise, this project investigates efficient methods to computationally simulate turbulence interaction noise. It is shown that the noise is sensitive to transverse turbulent disturbances, but is not sensitive to spanwise or streamwise ones. Therefore, accurate noise predictions are obtained by representing the turbulence as only transverse disturbances, which leads to significant savings in computational cost.

Vehicles moving on rough surfaces are subject to inputs which may be conveniently regarded as a combination of deterministic and random processes. Although this general problem is briefly addressed, it is the latter class of inputs which is of concern in the present work. In general, the random component of the excitation is `perceived' by the vehicle as a non-stationary random process, due either to inhomogeneity (spatial non-stationarity) in the surface roughness, or to variations in the vehicle velocity, or both. Analysis of the response of vehicles to such processes is further complicated by the multiple degree of freedom nature of the problem and by inherent non-linearity in vehicle dynamic systems, rendering exact statistical analysis of such systems analytically intractable, and thus requiring numerical simulation which is a costly alternative. A unified, analytical approach to this problem is presented here combining the techniques of linear systems theory and the approximate method known as statistical linearization, to facilitate the approximate analysis of non-linear systems excited by non-stationary random processes. The basis of the method is the use of a `shaping filter' description for the ground roughness, i.e., the height profile is represented as the output of a white noise excited, linear filter in the spatial domain (extensive justification of this assumption is presented). The key to the present work is the linking of the space domain filter with a state-space model for the vehicle dynamics by a formal change of variable, i.e., space is regarded as a function of time, related via the (variable) velocity function. This yields a time variable filter formulation for the excitation process which may then be coupled into the dynamic equations. After some further manipulations (using results from the theory of generalisation functions) differential equations may be constructed for the propagation of the mean vector and zero-lag auto-covariance matrix. For linear systems these results are exact and have been extended to the multiple input (multiple wheel) case. This extension presents no conceptual difficulty although it is computationally considerably more involved. For non-linear dynamic systems the method of statistical linearization is adopted so that the above method of obtaining means and covariances applies. This technique is essentially the replacement of the true non-linear element by a linear one such that their output differences is in some way minimised. This results in linear coefficients depending on the instantaneous mean and variance of response, just those quantities calculated by the above method and so a coupled set of non-linear, deterministic differential equations are obtained, governing the propagation of the approximate means and covariances of response. In order to validate results obtained by this method, in the absence of analytical solutions, extensive use is made of Monte-Carlo simulation. A fundamental concept arising as a result of the formulation is that of the `covariance equivalence' of two random processes. This enables the frequency-time (evolutionary) spectral analysis of random processes having a `frequency modulated' structure, a task which was hitherto not possible. This concept has also found application in the field of acoustics in the study of moving noise sources.

The great potential benefits associated with SFF (satellite formation flying) have led to considerable research in this concept around the world. As it is a new field of research and the implementation of formation flying in practice brings some inherent challenges and risks It is therefore particularly useful to develop a real time platform, where formation flying theories, technologies, algorithms and their coordination in open and close loop can be tested in scenarios close to those actually expected. In this research a low cost, real time hardware in the loop (RTHIL) test bed for low earth orbit autonomous SFF consisting of nano satellites has been developed which can integrate original hardware in the close or open loop. The proposed real time close loop test bed consists of a real time SFF model, a real time relative navigation system, guidance and control algorithm, and GPS simulator. Different SFF models already developed by researchers have been studied for their potential use in a real time test bed. Due to the limitations associated with them a novel real time SFF model and algorithm has been developed for real time hardware in the loop test bed. This real time SFF model was downloaded to a setup of two single board real time computers connected to each other through a serial port. The hardware of the relative navigation system has been designed and developed in this research. Single frequency GPS has been used as the relative navigation sensor. A model has been developed to estimate the dynamic characteristics of the GPS receiver for LEO orbits. This model simulates the relative dynamics of the GPS constellation and the LEO satellite. S3C2410 has been selected as the navigation processor. Radio transceivers are used to exchange data between the satellites. A uC/OS-II real time operating system is used in the navigation software code. The Hardware and software structure has been developed to simulate both centralized and decentralized approaches. A novel relative navigation algorithm is discussed. A Graphical Interface Software (GIS) has also been developed to initialize hardware and software prior any simulation and test run. It synchronizes the start of simulation on all real time computers, relative navigation systems and other components/simulator. It captures real time data from the deputy satellite. It also provides different debug and test options. Offline simulations were run for three models Hill’s, COEPOKE and RTSFF models by considering gravitational and atmospheric perturbation. The comparison of the results showed that the RTSFF model is able to simulate formations both in circular and elliptical orbits satisfactorily. The analysis of these models on the basis of their mathematical derivations showed that the RTSFF model gives better results than other models. The worth of this model for the real time test bed has been shown by running real time simulations onto a network of two synchronized single board real time computers and when the results were compared with offline simulation results they were found to be quite satisfactory. The hardware of the relative navigation system has been checked and tested. A basic infrastructure for navigation software has been developed. Different open loop tests have been run to verify the working of the hardware and software and these have verified the correct functioning of the system. The navigation algorithm could not be implemented in the embedded environment, Due to the unavailability of GPS simulator, no close loop simulation or test has been performed. The functionality of all the software, models, hardware and their operational control under GIS software has been checked in real time environment and found to be working correctly and are ready to be run in any open loop or closed loop simulation test.

This thesis presents the results of the research component of this EngD, entitled Use of fan rig data for the understanding and prediction of fan broadband noise and noise changes due to a variable area nozzle As suggested by the title, fan rig noise measurements form an integral part of this thesis. The analysis of a database of rig noise measurements forms the first section of this thesis, in two parts. The first part describes the analysis of a set of fan rig noise measurements, including the variation of fan broadband and tone noise in forward and rearward arcs. The second part examines a large database of fan rig noise measurements, and attempts to derive correlations of fan broadband noise and fan performance parameters. Cluster Analysis, Principle Component Analysis, and Regression Analysis are used to understand and describe the underlying physics of broadband noise generation and the relationships between these predictors. The second section of this thesis uses a cascade broadband noise model to investigate rotorstator broadband noise. Predictions of the broadband noise from this noise model are compared to rig measurements, showing good accuracy. The underlying physics of rotorstator broadband noise generation is investigated by performing two parametric studies using the broadband noise model. The first parametric study investigates the effect on broadband noise of simple flow and geometric parameters, namely number of vanes, vane chord, vane stagger angle, and rotor wake turbulence intensity, turbulent length scale, and flow Mach number onto the cascade. These results are used to derive scaling power laws for the prediction of changes in broadband noise due to changes in these parameters. The second parametric study expands upon this by investigating the effect on broadband noise of the fan design parameters shaft speed, pressure ratio, and efficiency, at approach, cutback and cruise conditions. The variation in broadband noise due to these design parameters is explained by considering the underlying flow and geometric parameters such as number of vanes and Mach number, and the scaling power laws based on these simple parameters are used to predict the change in broadband noise between different performance points. The final section of this thesis investigates the effect of varying exhaust nozzle area on total engine noise. A new method is presented that allows the transfer of changes in fan rig noise to Eugene P. Deane EngD Thesis September 2009 2 engine noise predictions, to estimate the change in fan noise due to the pressure ratio changes brought about by a variable area nozzle. Changes in engine noise are investigated for approach, cutback, and sideline conditions, and the application of the new method assessed. As the research displayed in this thesis is closely linked to industry, the foundation of work presented in several chapters is dependent on data or figures that are commercially sensitive. It has therefore been necessary to create a confidential appendix (Appendix X) to include these commercially sensitive items. These additional results and figures in Appendix X are supplementary in nature, and sufficient results are presented in the public thesis to illustrate the results of the various chapters. Where supplementary information and results are available, this is clearly indicated at the pertinent point in the published thesis, along with the section of Appendix X where the information can be found.

No description available.

Fully–resolved simulations of multiphase flow phenomena and in particular particulate flow simulations are computationally expensive and are only feasible on massively parallel computer clusters. A 3D SIMPLE type pressure correction algorithm is implemented and extensively tested and parallelized to exploit the power of massively parallel computing clusters currently available. Domain decomposition and communication schemes applicable to a general unstructured or structured multi–block CFD codes are discussed and algorithms are proposed, implemented and tested. Several high–performance linear solvers and a multi–grid strategy for the current framework are implemented and the best types of solvers are identified. A 2D CFD code is developed by the author to test several possible fixed–mesh strategies. Variations of immersed boundary (IB) and fictitious domain (FD) methods are implemented and compared. FD methods are identified to have better properties especially if other transport phenomena are also considered. Therefore an FD method is adapted by the author for the SIMPLE type flow solvers and is extended to heat transfer problems. The method is extensively tested for the simulation of flow around stationary in addition to freely moving particles and forced motion where both natural and forced convection are considered. The method is used to study the flow and heat transfer around a stationary cylinder and a new high resolution correlation is devised for the estimation of the local Nusselt number curves. Free fall problem for a single circular cylinder is considered and the effects of internal heat generation and also long term behavior of single cold particle subject to natural convection are also studied in detail. A particle collision strategy is also adapted and tested for the particle–particle collision problems. The FD algorithm is extended to the 3D framework and the flow around single stationary sphere and also free fall of a single sphere are used to validate the FD algorithm in 3D. A unique polydispersed fluid-particle turbulent modelling process is reviewed and the closure problem for this framework is studied in detail. Two methods for the closure of the non–integer moments which results from the polydispersity of the particles are proposed namely PDF reconstruction using Laguerre polynomials and a unique direct method named Direct Fractional Method of Moments (DFMM). The latter is derived using the results of the fractional calculus by writing an equation for the fractional derivatives of the moment generating function. The proposed methods are tested on a number of problems consisting of analytical, experimental and DNS simulations to asses their validity and viability which shows that both methods provide accurate results with DFMM having more desirable properties.

Evaluating the usability of In-Vehicle Information Systems (IVIS) guides engineers in understanding the interaction design limitations of current systems and assessing the potential of concept technologies. The complexity and diversity of the driving task presents a unique challenge in defining usability: user-IVIS interactions create a dual-task scenario, in which conflicts can arise between the primary driving tasks and secondary IVIS tasks. This, and the safety-critical nature of driving, must be specified in defining and evaluating IVIS usability. Work was carried out in the initial phases of this project to define usability for IVIS and to develop a framework for evaluation. One of the key findings of this work was the importance of context-of-use in defining usability, so that specific usability criteria and appropriate evaluation methods can be identified. The evaluation methods in the framework were categorised as either analytic, i.e. applicable at the earliest stages of product development to predict performance and usability; or empirical, i.e. to measure user performance under simulated or real-world conditions. Two case studies have shown that the evaluation framework is sensitive to differences between IVIS and can identify important usability issues, which can be used to inform design improvements. The later stages of the project have focussed on Multimodal Critical Path Analysis (CPA). Initially, CPA was used to predict IVIS task interaction times for a stationary vehicle. The CPA model was extended to produce fastperson and slowperson task time estimates, as well as average predictions. In order for the CPA to be of real use to designers of IVIS, it also needed to predict dual-task IVIS interaction times, i.e. time taken to perform IVIS tasks whilst driving. A hypothesis of shared glances was developed, proposing that drivers are able to monitor two visual information sources simultaneously. The CPA technique was extended for prediction of dual-task interaction times by modelling this shared glance pattern. The hypothesis has important implications for theories of visual behaviour and for the design of future IVIS.

Aircraft design is an inherently multi-disciplinary decision making process. In many design processes, the vehicle configuration is selected based on aerodynamic considerations in the concept design stage, before the structural layout is considered during preliminary design. Selection of the best vehicle configuration relies on an estimate of weight to determine the required lift of the aircraft. Structural topology optimisation is an efficient method for automatically generating a structural design layout that fits within a given design space and meets a given set of design criteria. By comparison to parametric structural optimisation approaches, topology optimisation permits a much greater design freedom. Though often difficult to manufacture using conventional methods, this design freedom can be exploited by using additive manufacture. In this thesis, an integrated concept and preliminary aerostructural design framework is proposed, which incorporates topology optimisation as a means of structural layout generation and weight estimation. The framework is utilised to optimise the wing geometry of an unmanned air vehicle, while generating a fuselage structure, intended for construction using additive manufacture, which satisfies a von Mises stress constraint. By comparison to an equivalent shell thickness optimisation study, the topology optimisation approach is shown to generate much lighter structural designs for the same aerodynamic efficiency.

An investigation of the flow physics of a diffuser equipped bluff body in ground effect has been undertaken. Situated at the rear of a racing car undertray, the diffuser is an important component and the least understood part of the vehicle. Diffuser performance can change dramatically with vehicle ride height. This includes a significant loss in performance at low ride heights which can also be a serious vehicle safety issue. An increased understanding of the diffuser behaviour in ground effect is required to assist design improvements. An accurate experimental database of the flow field is necessary both to aid this understanding and also to provide information against which the continuing development of computational simulations may be assessed. The present research is two-fold; experimental and computational. Model tests were conducted on a generic 3D bluff body equipped with a fixed angle diffuser representative of current racing car diffusers. Extensive experimental tests in wind tunnels equipped with moving belts included mean forces, surface pressures, oilflow visualisation, laser doppler anemometry and particle image velocimetry. The 3D diffuser flow field has been measured for the first time and the results are used to analyse the behaviour of the diffuser in ground effect. Complementary RANS simulations provide valuable insight into the modelling requirements. It is known that the diffuser generates down-force by accelerating air underneath the model through the channel formed by the model underside and the ground. The diffuser flow is characterised by a counter rotating vortex pair. The present research presents a new understanding of the diffuser flow field and the mechanisms causing its behaviour in ground effect. It has been found that the behaviour of the vortices alters according to the model ride height and the pressure gradient inside the diffuser. Additional down-force is generated due to the low pressure zones associated with these vortices. At relatively large ground clearances, the vortices are coherent and strong with a high axial speed core. At these heights the down-force experienced by the model increases with reducing model ride height. This behaviour is terminated at lower ground clearances by the advent of a plateau in the down-force curve and the occurrence of breakdown in the vortices inside the diffuser. The vortex breakdown results in large, diffusive and weak vortices. Maximum down-force on the model occurs at the lowest ride height of this type of flow at the end of the plateau. A sharp reduction in the down-force occurs thereafter, due to the complete breakdown of one of the vortices. The resulting asymmetric flow consists of a single coherent vortex to one side of the flow and significant flow reversal at the other side. At very low ride heights the vortices are asymmetric and weak Down-force reduction is believed to occur as a result of the steep pressure gradient inside the diffuser which advances the vortex breakdown inside the diffuser upstream as the model ride height is reduced. At the point of down-force reduction one of the vortices breaks down completely. At very low ride heights the boundary layers at the model underside and at the moving ground are believed to merge to restrict flow through the diffuser inlet. The experimental database is comprehensive and provides the necessary tool for validation of computational modelling. A computational simulation of the flow at a high ride height successfully predicts force and surface pressure coefficients and the main flow features.

In recent years, there has been growing interest in contra-rotating open rotors (CRORs) for use as the power plants on aircraft, since they are estimated to burn 20% to 30% less fuel than equivalent turbofan engines for short-haul flights. However, one of the main challenges for their introduction is their very high levels of noise emissions. There is therefore a need for schemes by which the noise from CRORs can be predicted and hence reduced. Nearly all of the research effort on the prediction and reduction of CROR noise has been so far focused on the tonal component, whereas broadband noise emissions remain poorly understood. Nevertheless, it is shown that broadband noise emissions from CRORs can be signifcant, hence the necessity of the present research. In this thesis, broadband noise emissions from uninstalled CRORs are investigated for the frst time. It is assumed that the two most signifcant sources of broadband noise in uninstalled CRORs are the broadband rotor-wake/rotor interaction noise (BRWI) and the broadband rotor trailing edge noise (BRTE). Fast semi-analytical prediction schemes are developed for these sources of broadband noise, which exhibit good agreement with noise measurements from a scaled model CROR. The relative importance of the BRWI and BRTE noise sources is investigated for a realistic CROR confguration at assumed take-off, cruise and approach-type conditions. It is predicted that both broadband noise sources are signifcant at assumed take-off, whereas only the BRTE noise contributes to the total broadband noise emissions at assumed cruise and approach. A parameter study is conducted to investigate the effects on CROR broadband noise emissions of variations in rotor-rotor separation distance, rotor speed and blade number at constant engine power, torque split and solidity. The validity of two widely used approximations for fan broadband noise predictions is also studied: Amiet's approximate BRTE model and the use of isolated airfoil theory for turbulence-cascade interaction noise. Criteria for the validity of Amiet's approximate model are established by comparing it to the general BRTE noise model developed in this work. The two models are shown to differ in the low and high frequency limits, but excellent agreement is observed for realistic rotor configurations over most of the audible frequency range. In addition, sound power predictions from isolated airfoils and blade cascades in a turbulent flow are compared. Excellent agreement is observed between an isolated airfoil model and a cascade model for frequencies higher than a critical frequency. Below this critical frequency, agreement is poor for high solidity cascades but is reasonable for low solidity cascades, typical of CRORs, thus validating the use of isolated airfoil theory in the BRWI model.