Some experiments are described in which an axisymmetric free shear layer attaches to a spherical surface, the exterior flow Mach number ranging from 0-75 to 207. The Chapman-Korst reattachment model is shown to correlate the results very well. The relation between the quantities measured and the numerical predictions of the model is discussed in detail and it is shown that the pressure changes which occur just downstream of reattachment are significantly different from those observed on flat walls.

Flight tests on the Handley Page suction wing showed that turbulence at the wing root can propagate along the leading edge and cause the whole flow to be turbulent. The flow on the attachment line of a swept wing was studied in a low speed wind tunnel with particular reference to this problem of turbulent contamination.

The critical Reynolds number, RθL, of the attachment-line boundary layer for the spanwise spread of turbulence was found to be about 100 for sweep angles in the range 40°–60°. A device was developed to act as a barrier to the turbulent root flow so that a clean laminar flow could exist outboard. This device was shown to be effective up to an Rθ of at least 170, so that experiments were possible on a laminar boundary layer at Reynolds numbers above the lower critical value. A spark was used to introduce spots of turbulence into the attachment-line boundary layer and the propagation speeds of the leading and trailing edges were measured. The spots expanded, the leading edge moving faster than the trailing edge, at high Reynolds numbers, and contracted at low values.

The behaviour of Tollmien-Schlichting waves was also investigated by exciting the flow with sound emanating from a small hole on the attachment line. Measurements of the perturbation phase and amplitude were made downstream of the source and, although accurate values of wave length and propagation speed could be found, difficulties were experienced in evaluating the amplification ratio. Nevertheless, all small disturbances decayed at a sufficient distance from the source hole up to the highest available Reynolds number of 170.

At the Aerodynamische Versuchsanstalt Gottingen (AVA), base pressure measurements were made on five elliptic cones. The ratios of the axes of the ellipses were: 1:12, 1:3, 1:1, 3:1, and 12:1. All the cones had the same volume and the same length. The investigations were carried out for Mach numbers M∞=1·50, 1·73, and 1·98 at angles of incidence between about —2 degrees and about 8 degrees. Reynolds numbers, based on a mean length, lm, varied from 2·5×106 to 3·0×106; the boundary layer approaching the base was turbulent. The base pressure coefficients are given as a function of geometry. Some of the results were compared with the base-flow theory of Korst and reasonable agreement was found.

In recent papers Messiter and Hida have proposed a first-order correction to simple Newtonian theory for the pressure distributions on the lower surfaces of lifting conical bodies with detached shocks. The method involves the solution of an integral equation which Messiter solved numerically for thin delta wings, while Hida gave an approximate solution for thick wings with diamond and bi-convex cross-sections. It is shown in the present paper that Hida’s approximate solutions give poor agreement with experiment, and a series of more precise numerical solutions of the equation are given for wings with diamond cross-sections. The pressures, and shock shapes, obtained from these solutions are in very good agreement with experiment at Mach numbers as low as 4·0.

The method has also been extended to Nonweiler wings at off-design when the shock wave is detached from the leading edges. Again the agreement with experiment is good provided the integral equation is solved numerically.