The present state of knowledge on the problem of fatigue failure due to vibrations excited by jet noise is reviewed. It is concluded that it should currently be possible to make reasonable estimates of the stress levels set up in a structure by jet noise but, in general, the resultant fatigue life of the components cannot be estimated with any confidence.

Since the introduction of aluminium-zinc-magnesium-copper alloys, investigations into the behaviour of these high strength materials, over a wide range of composition, have indicated the significance of corrosion as a factor affecting the cracking of these alloys after prolonged stressing at loads below the ultimate tensile strength of the material.

Most of the previous work designed to assess the stress-corrosion resistance of aluminium-zinc-magnesium-copper alloys (D.T.D. 683A*) was carried out on sheet and plate material and the present investigation is concerned with the examination of these materials in the form of large hand-forged slabs, tested in the short transverse direction.

The authors predict that rocket engines will fall into two categories. In both, combustion pressures will approach 200 atmospheres to attain optimum vehicle performance, balancing increased pumping losses and engine mass against increased expansion efficiency and reduced dissociation. For general purposes, engines will be intrinsically convenient: storable propellants will give specific impulses around 270 lb. sec./lb., and engine specific weights will be between 0·02 and 0·03. For exacting applications, the greatest overall convenience may necessitate using unpleasant propellants so as to obtain the highest performance. Immediate research into the use and handling of these is essential if future engine development is not to be hampered.

The paper provides a simplified picture of the processes occurring in liquid bi-propellant rocket motors. Droplet vaporisation, chemical reaction, and drag between gas and droplet are considered for a one-dimensional model. Quantitative theoretical results are presented for the variation of droplet size, droplet velocity, gas temperature and gas velocity within the chamber, particular attention being paid to the calculation of L*. The theory is applied to the German V2 rocket motor. Practical conclusions from, and extensions to, the theory are discussed.

The momentum equation is used to calculate the drag that acts near the tip of a semi-infinite swept-forward wing of constant chord that is at zero incidence to a uniform stream of incompressible, inviscid fluid. The drag is given as a function of wing section and angle of sweep and is shown to be unaffected by shaping the tip. The results may be used to calculate the drag that acts near die centre and the thrusts that act near the tips of a swept-back wing, provided that its aspect ratio is sufficiently great for the regions where the different forces act to be separated. Some results due to Neumark and to Küchemann and Weber suggest that this will be so provided that the aspect ratio is greater than about two. The results are combined with some due to these authors to estimate the spanwise extent of the forces. It is found that this extent decreases as the angle of sweep is increased and that it depends quite markedly on the wing section.

The effect of the ground on a jet-flap aerofoil in two-dimensional flow is investigated, with particular reference to the conditions under which the jet flow hits the ground. Experimental results are included for the change of lift and pitching moment for 58·1° and 31·4° jet-deflection angles, with a limited investigation into the downwash changes behind the model with a 58·1° jet-deflection angle. It is concluded that there is a definite limit to the jet coefficient, which may be used at any ground position, if the problems of a sudden change of lift and pitching moment are to be avoided. There remains the unresolved difficulty of the downwash variation at the tail arising from the presence of the ground.

A provisional assessment of the performance of a jet-flap aircraft near the ground is made, and shows that for a 58·1° jet-deflection angle there is a definite limit to the minimum flying speed for any given distance of the wing from the ground.

The fact that there appears to be a maximum pressure lift coefficient, independent of the jet-deflection angle, which may be obtained at any ground position, is investigated. An attempt is made to provide a theoretical explanation of this on the basis of the critical condition of blockage underneath the aerofoil.

The compressible subsonic flow about an oscillating two-dimensional aerofoil in a wind tunnel with porous walls is considered. The porous wall is assumed to exhibit the property that the ratio of the normal velocity at the wall to the pressure drop across the wall is constant. The aerodynamic derivatives are found for quasi-steady pitching motion of the aerofoil. The dependence of the derivatives on Mach number and wall porosity is displayed graphically.

The main advantage of the method presented in Part I of this paper is the generality with regard to the shape of the wing plan form. This is a welcome innovation in the field of supersonic wing theory, where methods of solution of the usual lifting problem are restricted to a limited number of wing shapes. Most of the lifting characteristics of wings are now known at supersonic speeds, but the range of wings with curved subsonic leading and trailing edges needs further investigations. Therefore it seems that the extension of the design problem to include these lifting problems is a useful addition to the extensive literature on the subject.

The “frozen stress” techniques of photoelasticity can give a complete knowledge of the stress, system in a solid body, but the examination of the stresses requires more time and care than in corresponding flat plate tests. In tests on tubes with transverse circular holes, sponsored by The Royal Aeronautical Society, all practicable geometrical shapes are examined and the maximum stress is measured in tension, bending and torsion. The results are comprehensive and show the inadequacy of previous results. In all cases the maximum stress occurs inside the bore of the hole. The accuracy of all the graphs of stress concentration factors is better than five per cent.

A photoelastic layer was bonded to the face of aluminium alloy lugs and the strain distribution in the lugs was determined from the photoelastic fringe patterns which were recorded in a reflection-type polariscope. The semicircularly- ended lugs were loaded through neat fitting hardened steel pins. Assuming the maximum shear stress criterion of yield, the region of yielding under different loads was determined from the fringe patterns. Yield contours are shown for each of the four ratios of (hole diameter)/(lug width) tested. The greatest shear strains in the lugs were related to the applied loads. The extent of yield across the section normal to the direction of loading was measured for different loads and it was found that the progress of yield across this section is independent of the (hole diameter)/(lug width) ratio. The residual stresses on this section were also estimated. Complete yielding across the lug was related to the properties of the material and to the fracture strength of the lugs. Local yielding due to surface irregularities remained local, showing that “bedding in” does not weaken the component under static loading conditions.

When a vibrating structure encloses a volume of fluid, the acoustic effects within this volume modify considerably the response characteristics of the structure, provided that the cylinder is vibrating in radial modes only. Measurements made of the displacement caused by a particular sound wave are of the same order as the values predicted. The calculation of the response of the cylinder to an acoustic wave also yields the sound field inside the cylinder and, again, the results are in general agreement with practical experience.

This paper is concerned with the buckling under uniform longitudinal compression of a variety of structures composed of plates whose thickness tapers linearly to zero across the section. Such structures include the angle of Fig. 1, the strut of cruciform section of Fig. 2 and the simply-supported strip of Fig. 3. For given cross-sectional area and overall dimensions (e.g. length of arm) the sections with linearly varying thickness achieve a greater buckling load (assuming that local buckling, rather than Euler buckling, is the criterion) than sections with any other smooth variation of thickness. These particular sections are therefore optimum sections and, even if they may not be used in practice, provide a convenient yardstick for purposes of comparison. The buckling loads are considerably greater than those for the corresponding “constant thickness” sections.

The equations of motion for an isolated laminar viscous vortex at moderate to large Reynolds numbers are linearised, by assuming that both the rotational velocity and the deficit of longitudinal velocity are small compared with that in the free stream. The rotational motion and the longitudinal motion may then be superimposed and solutions are readily obtained for each. If the vortex is generated by a body with profile drag it is predicted that the deficit of longitudinal velocity will be positive, which is in agreement with experimental observation. Further details of the solution and its relation to the flow in real vortices are discussed; and the theory is compared with some measurements in a turbulent vortex.

On the basis of the linearised theory, the integral relationship for the incidence distribution in terms of the velocity potential is established for wings with subsonic leading edges. Some analytical problems are analysed. A simple general numerical method is given for this design problem which compares favourably with exact linear theories. In Part II, to be published, a further numerical method is developed, for calculating the loading on any specified thin wing with subsonic leading edges, which again agrees favourably with exact linear theory. Both of these numerical techniques can be easily accommodated on desk calculating machines.

The characteristic functions for beam vibration modes are used to derive an approximate solution for the calculation of thermal stresses in rectangular isotropic flat plates subjected to arbitrary temperature distributions in the plane of the plate and constant temperatures through the plate thickness. The thermal stresses are obtained in the form of generalised Fourier expansions in terms of the characteristic functions, and their derivatives, representing normal modes of vibration of a clamped-clamped beam. Since these functions have recently been tabulated, the practical application of this new method to the thermoelastic stress analysis of plates presents no difficulty.

The paper concerns a phenomenon which occurs in rows of closely spaced blades, as in an axial flow compressor, when the flow around the blades approaches the stalling condition. Patches of stalled flow can propagate along the blade row, thus causing cyclic variations in the lift on individual blades. Under certain conditions this cyclic loading can lead to large blade stresses, and ultimately to blade failure.

Some recent theoretical work on slender pointed wings at zero lift is co-ordinated and extended. The wings considered may have any pointed plan form shape, provided that the trailing edge is straight and unswept. The root section profile and cross-section shapes are arbitrary, provided that, on any one wing, the latter are “descriptively similar” (diamond or parabolic biconvex for instance), though not necessarily geometrically similar. The chief aim of the work is to find wings with simple geometry, low wave drag and pressure distributions which are unlikely to be seriously affected by viscous effects. Wave drag and pressure distributions are calculated by slender-wing theory. General formulae, which are both simple and instructive, are given for the wave drag and the overall pressure distribution, with particular emphasis on the root pressure distribution. Results for a number of wings of special interest are presented and discussed.

This report describes the design of an annular entry for the axisymmetric flow into a circular duct, shaped so that the wall velocity gradients are favourable to the development of the boundary layer flow. As a first step the accelerating flow around a two-dimensional 90° duct bend with favourable gradients was obtained. This shape was modified and the axisymmetric flow pattern through it obtained by a combination of electrical analogue and relaxation methods and then checked by experiment. The favourable gradients were retained in the axisymmetric flow.

Previous theories on the surging of axial compressors have been based on a quasi-static theory, whereas a finite rate of growth of the circulation around each blade must occur in practice. An examination of this factor is made in the paper, studying the stability with respect to small changes in speed instead of mass flow as in the normal approach. Effects of capacity are neglected. It is concluded that an instability occurs in the last stage of the compressor, even when that stage is itself unstalled, which could give rise to surge. It has been found necessary to introduce one empirical factor into the analysis but, when this is given a value compatible with other experimental and theoretical evidence, reasonably good agreement with experimental results is obtained.

Expressions are obtained for the radial and tangential bending moments in a circular plate under the combined action of (a) a lateral load concentrated on the circumference of a circle and an end tension or compression, and (b) a uniformly distributed lateral load, having a diameter less than the diameter of the plate, and an end tension or compression. For both types of loading, solutions are obtained for plates which are simply-supported and for plates with an arbitrary end rotation.

In addition, the following limiting cases are considered: (i) concentrated lateral load with end tension or compression, and (ii) an infinite plate under the simultaneous action of an end tension and a lateral load concentrated on the circumference of a circle of finite diameter.