A method is presented for determining the potential flow around bodies of revolution, at incidence, in uniform, incompressible, flow. This method utilises polynomial distributions of both sources and doublets, in the manner of Fuhrmann, and gives considerable advantage over existing techniques. In particular, when compared with the method of Nielsen, in a typical case, the computer storage requirement is reduced by a factor of eleven. Compared with panel methods the saving is very much greater.

Unlike the method of Landweber, both body surface and flow field conditions are obtainable and it would appear that the new method is ideally suited to the ‘store trajectory’ application in place of the existing discrete source method.

Measurements are described of the drags of various forms of three-dimensional excrescence mounted on balances installed in the walls of the working section of the RAE 8 ft x 8 ft wind tunnel. The tests cover a range of Mach numbers between 0·2 and 2·8 and a range of Reynolds number.

The excrescences tested include circular cylinders and wings mounted normal to the surface, fairing shapes and mushroom shaped rivet heads. For excrescences which were small compared with the boundary-layer thickness the scale effects were correlated in terms of the wall variables of the turbulent boundary layer. For circular cylinders and wings which protruded well into or beyond the boundary layer the drag may be determined by using a drag defect relationship. The fairings tested comprised half-bodies of revolution with pointed or rounded ends and bodies of rectangular, including square, section with pointed ends.

For a 15° boat-tailed heat shield or payload shroud, the unsteady, compressible Navier-Stokes equations are numerically solved. For a Reynolds number of the order of one million and for four different freestream Mach numbers, the time integration is performed for 64 x 30 grid computational domain. The entire flow-field is analysed to delineate the gross dominant characteristics. It is observed that the flow in the boat-tail region is completely attached for Mach number 2·47 and beyond whereas it is separated for lower Mach numbers. A comparison between the numerical solutions and the available experimental results is provided and, for incipient flow separation conditions, an evaluation of unsteady surface pressure data is presented.