In the design of compression structures, the effect of lateral loading is usually ignored. This assumption has been justified where the structure has represented the wing skin of a conventional aircraft, in which the compressive end load due to bending is high, but the lateral load due to aerodynamic pressure is comparatively low.

However, in some cases the approximation is too crude, such as where the effect of the aerodynamic pressure is added to by pressurised fuel tanks together with the possibility of the accelerated weight of fuel. Such situations arise in supersonic transport aircraft where a considerable weight of fuel is carried in the wing and aerodynamic heating causes high compressive stresses in the skin.

A new aerodynamic open-circuit test rig for studying boundary layer ingestion (BLI) propulsion has been developed by National Research Council of Canada. The purpose is to demonstrate the advantages of BLI in reducing the power required for a given thrust and to validate the performance of BLI fan concepts. The rig consists of a boundary layer generator to simulate boundary layer development over an aircraft fuselage. The boundary layer generator can be used to create a natural boundary layer due to skin friction but also comprises an array of perforated plates through which pressurised air can be blown to manipulate the boundary layer thickness. The size of the boundary layer thickness can be controlled upstream of the fan blades. Parametric studies of boundary layer thickness were then feasible. The test calibration was conducted to validate the concept.

Before we begin, let me say that this paper is not a catalogue of engines, their bores and strokes, their output and weight. These aspects have been well covered before and are well recorded.

However, engines will figure prominently because that was the business of de Havilland Engines and our fortunes or misfortunes followed their pattern. However, instead of rigorously following the historic trail I prefer to pick out some of the landmarks and to outline some of the problems we met on the way. There may be a few anecdotes and something about the people who took part.

To begin at the beginning, we must start with Geoffrey de Havilland, who wanted to fly.

A centrifugal compressor rotor is quite robust, but the blading of an axial compressor looks horrifyingly fragile and very susceptible to damage by objects sucked into the engine. The introduction of axial compressor engines into airline service brings this again to the attention of designers and service engineers. The object of this paper is to outline the problem, and to describe one experimental approach to assessing the relative merits of different blade materials from the damage point of view.

A design procedure has been developed to generate aerofoil shapes for prescribed pressure distributions in an incompressible viscous attached flow. It is based on the method of singularities, originally proposed by Chen and later modified by Kennedy and Marsden, for inviscid flows. The classical approach of adding the displacement thickness of the boundary layer and wake to the aerofoil contour is used to account for viscous effects. Several numerical examples are worked out and are compared with the inviscid flow results. Significant changes in aerofoil contours due to viscous effects are observed and these are discussed.

During the past hundred years much data on the fatigue behaviour of a wide range of materials has been accumulated and it is now well known, as a result of service failures and laboratory tests, that stress concentrators whether mechanically formed or inherent in the material, can cause serious reductions in endurance.

The work described in this note forms part of a project being conducted at the Aeronautical Research Laboratories, Melbourne, Australia on the effect of stress concentrations on the fatigue properties of materials. The primary aim of the project is to establish a series of fatigue curves for different materials under a wide range of theoretical stress concentration factors and to attempt to find a basis on which to assess the notch sensitivity of different materials. The first investigation was carried out on 24S-T aluminium alloy and tests on two additional aluminium alloys, annealed copper and an alloy steel are in progress.

The subsonic doublet lattice method (DLM) has been the industry standard method for the calculation of unsteady air loads for the past three decades. A recent comparison between the DLM and the subsonic constant pressure panel code ZONA6 suggested that the DLM lacked robustness. However, the DLM code used in the comparison was flawed and not representative of the DLM as such. Results from a contemporary DLM code are presented for the same test cases that were used in the misleading comparison. These new results show that the DLM can produce valid results for all the test cases considered. Where feasible, a comparison is made between results from the present DLM code and the published ZONA6 results.

An aspect of structural design which is becoming increasingly more important is the behaviour of a structure when it is subjected to random pressure loading. Random pressure fields can be encountered in several different ways, two of the principal sources of such fields being the noise emitted by a jet, and the fluctuating pressures in a turbulent boundary layer. Many jet-propelled aircraft have suffered secondary structure fatigue failures in those parts of their structures near to the jet stream, and it is not unlikely that certain failures of panelling in engine intakes have arisen because of vibrations set up by the turbulent pressure fluctuations in the boundary layers at the walls.

The effect of multigrid acceleration implemented within an upwind-biased Euler method for hovering rotor flows is presented. Previous work has considered multigrid convergence for structured single block rotor solutions. However, for forward flight simulation a multiblock approach is essential and, hence, the flow-solver has been extended to include multigrid acceleration within a multiblock solver. The requirement to capture the vortical wake development over several turns means a long numerical integration time is required for hovering rotors, and the solution (wake) away from the blade is significant. Hence, the solution evolution and convergence is different to a fixed wing case where convergence depends primarily on propagating errors away from the surface as quickly as possible, and multigrid acceleration is shown here to be less effective for hovering rotor flows. Previous single block simulations demonstrated that a simple multigrid V-cycle was the most effective, smoothing in the decreasing mesh density direction only, with a relaxed trilinear prolongation operator. This is also shown to be the case for multiblock simulations. Results are presented for multigrid computations with 2, 3, and 4, mesh levels, and a CPU reduction of approximately 80% is demonstrated for 4 mesh levels.

A private enquiry about the initial application of Dimensional Analysis to experimental measurements in aerodynamics has been made of the writer. The consequent literature search has revealed what is now to be described. If due credit is to be given correctly the matter needs checking particularly by those who had direct contact with the early aerodynamicists throughout the world: initiation of that check is the reason for this paper. Dimensional Analysis originally came from ideas of similarity which were considered by such masters as Kepler, Newton and Galileo. The idea of equality of dimensions in an equation was used by Fourier. Dimensional Analysis was, after Galileo, used by biologists in the last century but in the general area of physics the initial impetus came from Rayleigh. Rayleigh first advocated the value of Dimensional Analysis in 1885 and though he usually referred to it as a method of similitude he also, despite what has often been written to the contrary, used the description of ‘method of dimensions’.

The natural frequencies and modes of vibration of a structure may be obtained more or less directly from experiments. The generalised masses and stiffnesses, however, are usually obtained indirectly from experimental investigations of the dynamic behaviour of the system. One of the techniques widely used for this purpose in the aircraft industry is known as the method of displaced frequenciesm.Stated briefly, the method consists of observing the changes in the natural frequencies caused by the addition of one or more small masses to the vibrating system.

The problem of bending of thin elastic plates has interested mathematicians and engineers for many years. Various methods have been put forward for determining the approximate solutions among which the variational method of Ritz is most common. This procedure leads to a system of linear simultaneous equations, the solution of which is often quite tedious, although generally by taking a few rows and columns quite accurate results are obtained. The purpose of this note is to show that by using bar eigen-functions for plates which are either clamped, or when some edges are freely supported, certain simplifications can be made in the process and the difficulty of solving simultaneous equations can be avoided; and that the deflection surface of the plate can be derived directly in a series form.

Since it is too complicated to obtain analytically the stress distribution in a plate containing two reinforced circular holes under uniaxial tension, the stress distribution along the inner edge of a reinforcing ring will be determined experimentally in this report.

The relationship between σθ/p at the inner edge of a ring, the dimensions of a ring and the distance between the centres of two holes will be investigated. Experiments are done in the cases when the uniform load is applied in the perpendicular or parallel direction to the centre line (see Figs. 1 and 2).