Despite the common title of this series of papers, the specific area into which we now move originated on a more global scale than the title suggests. Thus in the spirit of the plan set down at the outset for this review, viz that major ideas from abroad must be included, contributions from such countries as France, Germany and the United States here warrant greater prominence.

In Part 3 of this review we surveyed the development within Britain of the stress-skinned streamlined monoplane. Since the 1920s increasing numbers within Britain’s government research establishments, the universities and industry had been actively promoting this advance. Nonetheless, the practical hardware’s eventual emergence by the mid-1930s was largely driven by industry’s need for greater aerodynamic efficiency and a desire to excel. Thus record-breaking attempts, events such as the Schneider Trophy and MacRobertson races, all provided the impetus for this. So, too, did the commercial pressure of a new generation of American airliners and, for those astute enough to interpret them as such, the alarum calls of an impending Second World War. As a result of the appreciable decrease in drag coefficient achieved, coupled to a massive improvement in piston engine power, military aircraft, in particular, proved capable of flying not only further and higher but also dramatically faster. A further consequence, however, was that certain wartime high-speed fighter aircraft began to encounter serious, even fatal, compressibility effects as their enhanced speeds in dives reached subsonic critical Mach number conditions.

Modern fighter aircraft have been associated with lateral self-excited limit cycle oscillation known as ‘wing rock’. Simulations of wing rock have been encouraged to develop a complete understanding of the fluid mechanism that triggers and drives the oscillation, as well as for prediction purposes. Previous simulations of wing rock in wind/water tunnels were almost exclusively limited to a single degree-of-freedom in roll, due to the difficulty encountered in mounting the model to freely oscillate in more than one degree-of-freedom. Numerical simulations, utilising computational fluid dynamics, were also limited to roll-only degree-of-freedom. The loss of simulation accuracy due to the reduction of the actual wing rock degrees-of-freedom to roll-only has not as yet been fully investigated. In this study wing rock is numerically simulated in three degrees-of-freedom: roll, sideslip, and vertical motion for a generic fighter model. The unsteady Euler equations are coupled with the rigid-body dynamic equations through an innovative sub-iteration algorithm to simultaneously solve the coupled equations. The effect of including the sideslip and vertical degrees-of-freedom was found to delay the onset angle-of-attack of wing rock by 5° and reduce the limit cycle amplitude by about 50% with the frequency remained almost unchanged.

The method of matched asymptotic expansions has been applied to analyse the asymptotic structure of the incompressible turbulent attachment-line boundary layer at large Reynolds numbers. Inner and outer solutions were obtained that give the ‘wall-unit’ and ‘velocity-defect’ represention of both the spanwise and chordwise velocity profiles. The predictions from the theory agree with experimental and DNS data in the range of fully developed turbulence.