This paper reviews some current problems in unsteady aerodynamics and discusses some future challenges. The emphasis is on problems associated with separated flows, such as buffeting and limit cycle oscillations. Suggestions are made to indicate how experimental facilities (such as cryogenic windtunnels) and computational fluid dynamics might be exploited to address these questions.

Shockwaves are present in a flow as soon as the Mach number becomes supersonic. Being viscous phenomena, Shockwaves are a source of drag which can be predominant when the Mach number is significantly higher than one. In supersonic air intakes, the production of entropy by shocks is felt as a loss in efficiency. At high Mach numbers, Shockwaves produce a considerable temperature rise leading to severe heating problems, complicated by real gas effects. The intersection - or interference - of two shocks gives rise to complex wave patterns containing slip-lines and associated shear layers whose impingement on a nearby surface can cause detrimental pressure and heat transfer loads. The impact of a Shockwave on a boundary layer is the origin of strong viscous interactions which remain a limiting factor in the design of transonic wings, supersonic air intakes, propulsive nozzles and compressor cascades. More effort is needed to improve prediction of these interactions and to devise new techniques to control such phenomena.

The aerodynamic performance of aerofoils performing unsteady motions is important for the design of helicopter rotors. In this respect the study of aerofoils undergoing in-plane oscillations (translation along the horizontal axis) provides useful insight into the flow physics associated with the advancing blade in forward flight. In this paper a numerical method is developed in which the unsteady thin layer Navier-Stokes equations are solved for aerofoils performing rigid body motions. The method has been applied to the calculation of the flowfield around a NACA 0012 aerofoil performing in-plane motions representative of high-speed forward flight. Comparison of computed pressure data with experimental measurements is generally found to be good. The quantitative differences observed between computations and experiment are thought to have arisen mainly as a consequence of the low aspect ratio of the model rotor employed in the windtunnel tests.

Pressure distributions over a dual cavity cascade were studied at supersonic speeds of Mach 1·5 and 2·5. The study was performed through numerical modelling and results compared with model measurements. The Reynolds-averaged Navier-Stokes equations were solved using a finite-volume algorithm in which the inviscid cell interface fluxes were estimated using Roe's approximate Riemann solver with a second-order extension, and turbulence was modelled using a two-equation k-m model with compressibility corrections. Two test configurations were selected: (1) a length-to-depth ratio L/D = 1 cavity followed by another L/D = 1 cavity, and (2) an L/D = 3 cavity followed by an L/D = 1 cavity. The prediction was compared with that of a single cavity of the same L/D. It was found that the pressure field around the L/D = 1 cavity was substantially modified by a preceding L/D = 3 cavity. Changes in the pressure and pressure drag coefficient were observed. The study clarified some earlier observations of unsteady modes over a dual cavity cascade, and confirmed model measurements of the pressure fluctuation under a number of flow and geometry conditions.

As part of a programme of research to improve mission effectiveness by studying pilot workload and task performance in mission-oriented flight tasks, a methodology has been developed in which wavelet analysis is used to extract information from records of vehicle response and of pilot control activity. By decomposing the records into discrete wavelets, components of vehicle agility and pilot workload are derived in the form of wavelet-based ‘quickness’ parameters for vehicle agility and so-called ‘attack’ parameters for pilot workload. It is shown how individual wavelet components in the records of pilot control activity, referred to as ‘worklets', can be associated with the sub-tasks of ‘guidance’ and ‘stabilisation'. It is demonstrated how these concepts can be applied to quantify changes in pilot control activity associated with increasing task difficulty or changes in aircraft handling qualities. Two examples are presented, one from a flight trial in which the task difficulty was increased by changes in a prescribed ground track and the other from a simulation trial in which an increased time delay was introduced into the response of the flight control system.