The paper considers the distribution of downwash at the surface of rectangular planforms with prescribed subsonic aerodynamic loading. Three separate aspects of the problem are treated analytically, the results of each being tested against those derived from an accurate numerical procedure (Section 2). Asymptotic expressions for large and small aspect ratio are formulated and shown to apply over a wide range of aspect ratio (Section 3). Downwash routines from certain existing lifting-surface methods are studied, and their patterns of convergence are illustrated and compared (Section 4). A logarithmic singularity near the leading tip corner is identified, but it is observed that this can exist without serious detriment to the lifting-surface methods (Section 5).

The BAC subsonic lifting-surface method is applied to the problem of evaluating the downwash at the surface of wings of rectangular planform subjected to particular loadings and the results are compared with a set of datum results. Over 95 per cent of the wing span the results of the BAC method are shown to be accurate to within 0.01 per cent, irrespective of chordwise position.

An incompressible three-dimensional turbulent wall jet originating from a circular orifice located adjacent to a plane wall is studied both theoretically and experimentally. An approximate similarity analysis predicts that the two transverse length scales, l0 and L0, and the inverse of the mean velocity scale grow linearly with distance downstream x from the orifice. Experimental measurements of mean velocity and longitudinal turbulence intensity profiles were made both in air and water with hot-wire and hot-film anemometers respectively. The behaviour predicted by the similarity analysis was verified. It was found that the rate of growth of the length scale normal to the plane wall, dl0/dx, was somewhat less than that found for a two-dimensional wall jet, whereas the rate of growth of the length scale in the lateral direction, dL0/dx, was about seven times greater than dl0/dx.

The three-dimensional motion of a body towed by a flexible cable is studied. An approximate solution is developed for the non-linear equations of motion of the system to determine cable shapes. This solution is then used to compute natural frequencies and damping rates for longitudinal and lateral motions of the system. Conditions necessary for the uncoupling of the lateral and longitudinal motions are defined.

Measurements are presented of the mean flow properties of some three-dimensional turbulent boundary layers re-developing after reattachment behind short separation bubbles yawed at 26.5° to the main stream. For these measurements, Rθ11 varied from about 550 to 1450. It was found that, where the pressure gradient parameter (ν/ρu3τ1)∂p/∂s was not greater than about 0.05, the flow in the local external streamline direction conformed well with empirical laws for fully-attached two-dimensional layers with regard to the mean velocity profiles, shape parameter relationships and skin friction laws, giving support to the usual assumption that these two-dimensional relationships may be applied to the streamwise flow in three-dimensional layers, subject to the limitation on the pressure gradient parameter. The cross-flow profiles, on the other hand, were not generally fitted well by the often-used representations of Mager and Johnston. The variations of the kinetic energy dissipation coefficient and the entrainment rate were deduced for one of the layers, both quantities being found to be higher than those predicted by empirical relationships for conventionally-developing two-dimensional layers. However, the energy dissipation is in fair agreement with that in a similarly re-developing two-dimensional flow.

Measurements are presented of the base pressure coefficient, drag coefficient and Strouhal number of rectangular cylinders. The results confirm a finding in Japan that the drag coefficient rises to nearly 3 when the depth of the section is just over half the width. The flow around the sections is found to be strongly influenced by the presence of the trailing-edge corners.

The momentum analysis of the mean flow of an incompressible fluid in a straight suction duct with porous walls dealt with in a previous paper is extended to account for deflection of the inflow towards a direction parallel with the axis of the duct. Ducts with a continuous distribution of porosity and ducts with a relatively small number of rather large ports are considered and the results are applicable to regions where the velocity and pressure can be assumed to be distributed uniformly across planes normal to the mean flow inside them. Some measurements made with a one-quarter full-size scale model of the straight and more or less regularly obstructed suction duct used in the boundary layer control system of a high-lift research aircraft show that at least part of the theory permits a reasonable prediction of pressures.