The glancing interaction of a Prandtl-Meyer expansion fan with a supersonic wake is examined. The particular supersonic wakes of interest are the combustion wake generated by a vertical strut injected scramjet, and the side wall boundary layer of a supersonic wind tunnel. A physical structure for the interaction is defined, and an approximate analysis of the downstream behaviour of the wake is presented. The phenomenon of wake over-turning by the glancing expansion fan is introduced, based on this analysis. This over-turning produces a high pressure zone on the expansion generating surface, which for the vertical strut injected scramjet, is part of the thrust nozzle. Combustion wake over-turning is suggested to be a major thrust producing mechanism in this form of scramjet.

Pressure measurements were taken on the expansion generating surface of a glancing interaction involving the side wall boundary layer of a small supersonic wind tunnel. Results from the analysis presented were found to be consistent with these measurements.

A study of the thrust augmentation arising from the application of ejectors to high pressure reaction control systems typical of VSTOL aircraft is presented. A numerical analysis was undertaken to identify parameters critical to the attainment of high performance. A novel technique for improving mixing in confined ejector flows was investigated experimentally.

Theoretical work considered configurations with both sonic and supersonic primary flows. The studies show that thrust augmentation ratios ⋍ 1·3 are possible at pressure ratios around 15 using an ejector area ratio of 35. Little benefit accrues from diffusion at high pressure and performance is sensitive to mixing efficiency in terms of the mixing duct exit flow uniformity and primary nozzle losses.

Jets issuing from underexpanded nozzles fitted with castellations around their exit were compared experimentally with those leaving plain nozzles. Significant improvements in jet entrainment rates arise from the castellations. Strong correlation is observed between the free jet mixing rate and ejector secondary thrust. The level of thrust augmentation is low in conventional terms but constitutes a significant improvement over plain nozzle ejectors and non-augmented convergent nozzles.

The turbulent flow around a delta wing at incidence is simulated numerically using a finite-volume Navier-Stokes method. The numerical simulation makes use of a simple algebraic turbulence model. The influence of grid refinement is investigated. The more realistic wing-body configuration with round leading edge is used to discuss the influence of the position of the transition line and to compare experimental and numerical data to validate the numerical method. The topological structure of the flow is discussed. An explanation is given for the low particle density area close to the primary vortex as it is visualised by laser light-sheet technique in the experiment. Results for a close coupled delta-wing-canard configuration are discussed.

Safety regulation usually has some direct impact on airline costs. All operators aim to operate safely - some more than others - and one might ask, should safety be left to manufacturers and operators, all of whom have a real vested interest in high standards, including the added spur of potential litigation? However, there is, I believe, a clear wish in the minds of the public, and hence governments, for independent and positive regulation. Major fatal accidents frequently cause a clamour of demands for higher safety, and governments, through their appointed bodies, are expected to act. Airlines broadly accept such pressures but wish to see common and consistent standards applied so that the economic burden is spread equally - preferably internationally.

ICAO requires all aircraft involved in international navigation to conform to ‘a detailed and comprehensive code’, but the detailed provisions and their interpretation may vary from country to country.

This paper presents the results of an interferometric study of the shock/boundary-layer interaction at a compression corner. The boundary layer upstream of the corner is a non-equilibrium state having been disturbed by injection through a porous surface. As well as being relevant to the design of transpiration-cooled surfaces, the results show many similarities to the interaction in an adverse pressure gradient. The changes in the nature of the interaction as the approaching profile is distorted is fully explored and the results should provide a useful test case for computational fluid mechanics. The results are also used to evaluate the accuracy of interferometric techniques.

The paper discusses recent work in the Theoretical Division at the Aircraft Research Association Limited (ARA) on the modelling of complex configurations, rotors, propulsion and viscous flows. The general features of the ARA multiblock system, which provides a grid generation and Euler flow solution capability for complex configurations, are discussed. Although very successful, the system has various limitations, in particular the tendency for the grid quality to depend on configuration complexity and type. A new multiblock system is being developed which allows regions of unstructured grid to be included, giving increased flexibility in dealing with complex geometries and other improvements. Work on rotors has included the development of a new Euler code for propellers, as a replacement for the code in current use. Improvements in flow solution offered by the new code are illustrated. In the propulsion field, the extension of a viscous-coupled Euler code for afterbody flows to a complete cowl unit is described briefly. The code is particularly relevant to the new generation of large civil turbofans. Work on Navier-Stokes methods for afterbody/nozzle flows is also discussed. Other work on viscous flows includes a 2-D investigation into viscous modelling with hybrid grids, and the further development of an established transonic flow aerofoil code.

A knowledge based expert system for the design and analysis of composite laminated struts is described. This system is part of a Composite Design EXpert (CODEX) system that not only performs analysis and design of composite laminated plates and struts, but also incorporates the modelling of uncertainties of information, definition and decision when assessing competing designs. The analysis part of the system is based on a minimum weight optimisation that satisfies the failure criteria of local buckling, overall buckling and maximum strain or maximum stress. Available in the system for analysis and assessment are seven types of strut shapes, six composite materials and aluminium alloy. Examples of strut design are carried out and assessed using this system.

An expansion tube with a free piston driver has been used to generate quasi-steady hypersonic flows in argon and air at flow velocities in excess of 9 km/s. Irregular test flow unsteadiness has limited the performance of previous expansion tubes, and it has been found that this can be avoided by attention to the interaction between the test gas accelerating expansion and the contact surface in the primary shock tube. Test section measurements of pitot pressure, static pressure and flat plate heat transfer are reported. An approximate analytical theory has been developed for predicting the velocities achieved in the unsteady expansion of the ionising or dissociating test gas.

A computational method to predict the viscous transonic flow development around two-dimensional aerofoil sections is described. The Reynolds-averaged Navier-Stokes equations applicable to turbulent flow are discretised in space using a cell-centred finite-volume formulation. A multi-stage, explicit, time-marching scheme is used to advance the unsteady flow equations in time to a steady-state solution. Turbulence closure is achieved using either the Baldwin-Lomax algebraic model, or a one-equation model based on the turbulent kinetic energy equation. This latter equation is solved using essentially identical procedures to those for the mean-flow equations. Results are presented for the RAE 2822, RAE 5225, CAST 7 and MBB-A3 transonic aerofoil sections. The relative performance and limitations of the two turbulence models are discussed.

It is customary to begin lectures of this kind with reference to the person after whom the lecture is named. I do so with particular relish on this occasion.

This is because I spent most of my working life at The City University, formerly The Northampton Polytechnic, with which Sir Frederick Handley Page was closely associated throughout his working life.

The results of a little archival work reveal that in June 1909 a class in aeronautics was arranged for the following session. As it turned out, the momentous event of 1910 was the appointment of a Mr Frederick Handley Page to be lecturer in aeronautics to attend for two evenings per week at £75 for the session.

Not surprisingly the subject was a great success at the Northampton, and 81 years later at The City University, it still is.

A symmetric Mach 12·76 hypersonic flow in a 90° corner, formed by 30° swept back intersecting 8° wedges, was investigated in detail through numerical simulation using locally conical Navier-Stokes equations. Three different numerical schemes for spatial discretisation, MacCormack central differencing, van Leer’s flux vector splitting and Osher’s flux difference splitting, were studied to compare their capabilities to capture both strong shock waves and thin shear layers. Comparison with experimental data was made to validate the simulation. The numerical simulation provided further insight into the flowfield and a pair of counter-rotating vortices were discovered near the junction of the corner.

Wind tunnel data have been obtained from a flapped tanker wing and receiver aircraft model at varying vertical separation and the results compared with theory. In the aerodynamic model, the tanker wing is represented by a pair of horseshoe vortices from the wing and flap tips and an allowance is made for the self-induced vertical displacement of the trailing vortices at the receiver aircraft position. The aerodynamic loads on the receiver are determined by the vortex-lattice method and lifting-line theory, although an approximate method is used to determine the side force on the fin. Data were obtained from open and closed test sections in order to estimate the wind tunnel boundary interference effect. In the longitudinal case, significant differences were obtained between theory and experiment. Two reasons for the differences are the assumption that the trailing vortices are fully rolled up and the neglect of viscous decay of the vortices. The lateral aerodynamic interference was determined by banking the tanker wing and displacing it sideways and by yawing the receiver aircraft model. In the case of the rolling moment due to sideways displacement, which is the most significant lateral aerodynamic interference term, the wind tunnel boundary interference is highly significant and the difference between theory and experiment is large due to incomplete roll up of the trailing vortices. The theoretical and experimental trends are similar although the theory overpredicts the rolling moment due to bank and sideways displacement while the corresponding side force and yawing moment are underpredicted. As in previous tests using an unflapped tanker wing, the effect of the sidewash due to the tanker wing on the receiver in yaw is found to be relatively insignificant.

Measurement methods using holography are non-invasive, whole-field and can be applied in real-time. They are particularly appropriate for experimental validation of current CFD (computational fluid dynamics) codes which have 3-dimensional, transient, irregular geometry capabilities.

In holographic interferometry, the fringes formed by refractive index changes represent lines of constant Mach number in isentropic compressible flows and isotherms in convection. Examples are given of two-dimensional inter-ferograms in these areas, and their quantitative interpretation.

Automatic fringe and data processing is a necessity, and the method is extendable to three dimensions using tomographical reconstruction. These issues are discussed, together with the general question of comparison with flow predictions.

Fluid fields may also be treated on a three-dimensional time-dependent basis, using HCV (holocinematographic velocimetry), a holographic extension of PIV (particle image velocimetry). It is proposeds to run an experiment to measure both fluid and thermal fields, and surface temperatures simultaneously, using HCV or PIV, holographic interferometry and liquid crystal methods respectively.

Free vibration analysis of laminated composite beams, plates and stiffened plates using the finite element method has been presented. To give due importance to the shear deformation in composite materials, a high-order element which considers the quadratic variation of shear strain along the thickness of the element has been developed. The element has been developed using the quadratic isoparametric shape functions. Therefore, it can accommodate curved boundaries. Moreover, the stiffener element can have an arbitrary planform and it need not pass through any nodal point. The stiffener element has been developed in such a fashion to make the mesh division free from the location of the stiffener. This is extremely advantageous for optimisation of the path of the stiffener in a stiffened plated structure, while it is not necessary to modify the mesh division whenever the stiffener path is modified. Two different shape functions — the eight-node Serendipity and the nine-node Lagrangian — have been employed and a comparison between their performances has been presented. Two schemes for the generation of mass matrix, namely the lumped mass scheme and the consistent mass scheme, have been developed. The elements have been implemented in the ASKA element library and they can be used along with any standard ASKA element.

Vibration of a flexible wing strut of 30% thickness/chord ratio occurred during icing trials of a commuter aircraft. This was caused by the formation of ice on the leading-edge of the strut producing a ‘double-horned’ ice shape. The two horns formed one above the other, having very sharp leading-edges with radii less than 0·6% of the strut chord of 0·23 m.

Wind tunnel testing demonstrated that, over a limited range of incidence angle, the flow around the flexible strut was dominated by a long separation bubble that periodically ‘burst’ and re-attached. This occurred at a reduced frequency of nominally 0·06, based upon the chord of the profile and the free stream velocity. The periodic flow was also spatially highly coherent, producing relatively large unsteady forces compared with the buffet forces generated by incoherent turbulent separated flows.

A high-speed film was made using smoke visualisation of the flow around a 40% rigid scale model of the strut. This was used to correlate the unsteady pressure field, measured at 18 points around the profile, with the dynamic behaviour of the separation bubble.

The periodicity was found to be inherent in the flows over both the flexible and rigid struts for a limited range of incidence, occurring even when structural response was restrained to a level below that at which lock-in occurred between the flow and the response motion. The peak value of the unsteady pressure was measured on the top surface of the upper ice horn. The root mean square value of 18% of the free stream dynamic pressure was measured, with response restrained to below the lock-in threshold.

A mathematical model of unsteady aerodynamic effects for use in lateral aircraft dynamics has been proposed based on Weissinger’s arrangement of trailing vortex system. The suggested model is shown to be suitable for use in parameter extraction algorithm. Simulated flight data of an example airplane has been analysed through maximum likelihood parameter estimation algorithm in frequency-domain to show the effect of inclusion and omission of unsteady aerodynamics on estimated parameters.

Any review of the development of maritime satellite communications is essentially the history of Inmarsat and its predecessor system, Marisat. With perhaps two qualifications, one might extend the definition to mobile satellite communications generally. In the land mobile satellite communications domain two companies, Geostar and Qualcomm, have also made their mark. Geostar was recently declared bankrupt but Qualcomm continues to go from strength to strength.

We are frequently asked ‘what is Inmarsat?'. The term Inmarsat is generally used, and used herein, to mean the permanent staff (the Directorate), Land Earth Station (LES) operators, manufacturers and value-added service providers who offer mobile satellite communications products or services via the satellites and other infrastructure which makes up the Inmarsat system. It is, essentially, a partnership but within it there is also internal competition. LES operators compete in the provision of telecommunications services (quality is consistently high, but they each decide what services to offer and at what tariffs). Manufacturers compete to secure market dominance and profitabilty for their products. Not all will be equally successful. More than anything, Inmarsat is a global partnership dedicated to meeting mankind's need to communicate while on the move. It aims to do so cost-effectively and using satellites.

In this paper a brief account of the optimal tracking problem (O.T.P.) is given, to show that the response of the output vector of a linear observable system can be made to be close to some desired dynamic response. This optimal tracking problem has been applied to linear mathematical models of the F-100 jet engine for various operating conditions, and the results obtained prove that it is possible to achieve the desired thrust response at each operating condition without violating the engine's operational requirements. An attempt was made to establish whether a single control law, based on the solution of the optimal tracking problem at one condition, could be used with the engine throughout its operating range.

Heat transfer rates on a straight biconic, an attractive configuration for an aero-assisted orbital transfer vehicle, were made in the ANU T3 shock tunnel at various angles of attack and a stagnation enthalpy of 26 MJ/kg, equivalent to a flight speed of 7-2 km/s. The heat transfer database at this higher enthalpy has been compared with the data obtained on the same model in hypersonic and hypervelocity facilities elsewhere. The non-equilibrium flow effects on the model have been discussed in terms of the Hornung reaction rate parameter Ω.

The feasibility of using a single effective on-line control law to regulate the dynamic response of a jet engine throughout its flight envelope is of considerable importance. In this paper, a study was made to find such a single effective control law, which was able to regulate the dynamic responses of the linear, and ultimately the non-linear, model, throughout the engine’s entire operating range. This effective control law was obtained from solving a Linear Quadratic Problem (LQP) to obtain an optimal control law, a linear state feedback control.