The velocity and pressure distributions in a flow generated by a thick air jet that throttles a confined airstream have been studied analytically and experimentally. Velocity and pressure measurements were made in a duct with a rectangular cross section of 102 mm height and 19 mm depth, through which air flowed at velocities ranging from 65 to 80 m/s. The airstream was throttled by a thick air jet having velocities ranging from 130 to 150 m/s that entered the mainstream at angles ranging from 60° to 135°. The jet-mainstream contour was found to be elliptical and agreement within six per cent was obtained between the theoretically and experimentally determined maximum height of the contour. Jet spreading was found to be linear. The theory permits determination of the velocity profile in the jet and gives velocities that deviate less than ten per cent from values obtained experimentally.

An alternative analytical method is applied to interference problems of an aerofoil very close to the ground, such as a free surface, a solid wall, and a slipstream. The method of matched asymptotic expansions on the small ground clearance is used in the governing integral equation and this analysis is different from Widnall and Barrows’s analysis. They used the matched asymptotic method in the differential equation. To show that the present method is reasonable, the problem of a ground-effect aerofoil in uniform fluid flow is considered and it is found possible to obtain the same results as those of Widnall and Barrows for the case of small ground clearance. By comparing with some earlier studies, it is found that the asymptotic method used with small ground clearance is valid for a clearance less than about 0.3 of the chord length. Next, the interference of ground-effect aerofoils in the channel wall, in a free jet stream, and in a slipstream are examined, and it is found that the present method can be easily applied to these problems.

Unsteady high Mach number flows past oscillating cones and oscillating wedges are studied according to Newtonian impact theory, with and without centrifugal force corrections. It is found that, contrary to existing theories, the centrifugal force contribution in unsteady flow is significant and not negligible; for example, it contributes one-half to the dynamic stability derivatives of an oscillating wedge and one-third to that of an oscillating cone. It is also shown that in both cases the dynamic stability derivatives calculated by Newtonian impact theory with centrifugal force correction agree exactly with those obtained by gas-dynamic theory in the double limit γ → 1 and M∞ → ∞, independently. These conclusions are expected to hold for more general three-dimensional unsteady Newtonian flow when the Newtonian shock layer is attached.

This paper deals with the force coefficients on the leeward of two bluff bodies, in particular with the origin and nature of the lift force. Initially a general description of the characteristics of the force coefficients, plus the effect of parameters such as free-stream turbulence and surface roughness in the form of stranding, is given.

A resume of all the known suggestions for the origin of the lift force is listed and a series of experiments, attempting to test these suggestions using simple methods, is described. These methods consist mainly of force measurement, involving force measurement on the leeward conductor using different shapes for the leeward and windward bodies. However, static pressure tapping and flow visualisation are also employed to a small degree.

It is shown that none of the suggestions gives a satisfactory explanation for the origin of the lift force. However, several interesting points were noted; for instance the dependence of the lift profiles for the leeward body on its own characteristics in the wake and not particularly on the wake characteristics. Hence it is shown that the use of wake parameters, measured without the presence of the leeward body when attempting to assess the forces thereon, is an over-simplification of the situation as far as lift is concerned while appearing to work quite well for drag.

An analysis is presented for the post-buckling behaviour of simply-supported, laminated cross-ply rectangular plates subjected to biaxial compression. The solution to von Kármán-type large deflection equations of the plate is expressed as a double sine series for the transverse deflection and a double series of clamped-clamped beam functions for the force function. All the boundary conditions, including those involving the normal bending moment at the edges, are satisfied exactly. The series solution is found to converge rapidly. Using only the first few terms in the series, numerical results for square graphite-epoxy unsymmetric laminates under uniaxial compression is presented graphically.

A numerical method for the solution of three-dimensional supersonic flows and the associated computer programme have been devised. The present algorithm is based on a reference-plane characteristic network. The physical region between the shock wave and the body surface is replaced by a computational mesh facilitating the incorporation of different boundary conditions and arbitrary body shapes. For reasons of stability and computing time a Hartree-type grid was used where the characteristic segments are projected upstream. The base points do not coincide with nodal points and interpolation by means of Tchebycheff polynomials is required to obtain the flow properties. Numerical differentiation is carried out by means of spline functions. The programme has been used and verified under a wide variety of flows past circular and elliptic cones, tangent ogives and a delta wing body. After a series of numerical experiments it was found that accurate solutions could be found using a 7 × 7 grid, and that the essence of the solution is preserved using a very coarse mesh of 4 × 4. Furthermore, the good agreement obtained with experimental results indicates that it could be applied successfully to problems of practical interest.