This paper describes rig tests made to evaluate the effectiveness of the cockpit insulation and cooling system of a high-speed aircraft.

The tests showed the dependence of cockpit internal temperature distribution and heat pick-up on the cooling air mass flow and inlet temperature. Analysis of the test data showed that there was considerable heat leakage into the cockpit; the heat leakage increased with cooling air flow and constituted two-thirds of the heat entering the cockpit when the air flow was moderately high (20 lb/min). Some of the leakage heat entered the cockpit through equipment mountings but it was evident that other leakage paths existed. One more obvious heat leakage path, at the canopy, is illustrated.

The tests also showed that the internal heat transfer coefficient increased with air flow, reaching a value of 2·5 C.H.U./hr ft2°C when the flow was 20 lb/min.

A combined experimental and theoretical study has been made of radial compressible flow without swirl between parallel discs when the fluid velocity is everywhere subsonic. It constitutes an extension of a previous analysis, wherein the flow was considered to be incompressible.

A similarity solution for the radial pressure distribution is shown to be possible only in special cases where certain terms in the equations of motion can be neglected. Approximate solutions are obtained for the laminar and the turbulent radial pressure distributions, using an integral momentum method. Both theories agree well with experiment. The critical Reynolds number for reverse transition is found to be approximately the same as that for incompressible flow in radial channels and circular pipes. The radial pressure level for compressible radial flow between parallel discs is found to be less than that for incompressible flow at the same mass-flow rate. The non-linear form, in pressure, of the compressible solution can be shown to account for this result.

An analysis is presented of the buckling of a rectangular plate under combined biaxial compression, bending and shear. The sides of the plate to which bending is applied are simply-supported, the other two sides being supported by arbitrary edge members. The solution is expressed in a form suited to automatic computation and some sample results are given.

An analysis is presented which enables the boundary-layer thickness parameters of a re-attaching shear layer to be determined when the free-stream flow upstream of the base is supersonic, the base pressure is known, and die initial boundary layer is turbulent. The application of this analysis to some experimental results, on the flow behind blunt-trailing-edge wings and over a back-step where both the base pressure and the initial boundary layer are known, would appear to indicate that the re-attached profile could be specified by one parameter, namely the transformed shape parameter, the transformation used being a turbulent analogue of the well-known Stewartson-Illingworth transformation of the laminar boundary layer and where the shape parameter is defined as the ratio of boundary-layer displacement to momentum thickness. By adopting a value of the shape parameter in advance, it is possible to use the analysis to determine the base pressure by an iterative process and so, on this basis, it is suggested that this analysis is used to replace the existing recompression criterion of the Chapman-Korst model of separated flow which aims to predict base pressures and is known to be capable of improvement.

As part of this investigation, an improvement had to be made to the existing compressible turbulent shear layer velocity profiles of Korst and others and this was achieved by means of the compressibility transformation.

A simple circuit is described which gives a direct meter reading of the intermittency factor in a turbulent flow. The operation and setting-up procedure of the instrument is discussed.

This paper presents a general procedure for calculating the natural frequencies of rectangular plates continuous over identical and equally spaced elastic beams which are simply-supported at their ends. Arbitrary boundary conditions are permissible on the other two edges of the plate. The results are compared with those obtained by using the orthotropic plate approximation for the system

The flow in a two-dimensional channel with porous walls through which fluid is uniformly injected or extracted is considered. Following Berman, a solution is obtained giving a fourth-order non-linear differential equation which depends on a suction Reynolds number R. Numerical solutions of this equation have been obtained. Series solutions of this equation for small and large Reynolds number are given and are shown to give good agreement with the numerical solutions.