Skip to main content
European Commission logo print header

Buffet Control of Transonic Wings

Final Report Summary - BUCOLIC (Buffet Control of Transonic Wings)

Executive Summary:
Buffet is a flow instability which can cause structural vibration (“buffeting”) on civil and military aircraft. Transonic wing buffet is important to civil aircraft designers, because it defines a limit to cruise flight conditions which impacts on airframe cost and efficiency. As a result there is an increasing interest in predicting buffet offset early in the design process, and in delaying or alleviating the effects of buffet by active or passive means. The practical difficulties of modelling buffet have been known for many years and much research has been undertaken with ever more sophisticated modelling tools to analyse the phenomenon, with a focus on two-dimensional aerofoil flows. More recently attempts have been made to reduce the effect of buffet, but these have been hampered by a lack of understanding of the physics of transonic buffet on three-dimensional swept wings.

The aims of the BUCOLIC programme were to significantly improve our understanding of the underlying flow physics and to develop techniques to predict buffet in a practical timescale for industrial aircraft design. These aims were addressed using a combination of experimental investigations of buffet on a representative transport aircraft wing in the Aircraft Research Association’s industrial transonic wind tunnel, and numerical simulations at the University of Liverpool.

Project Context and Objectives:
The objectives of the BUCOLIC project were to advance the state of the art in buffet prediction and control, specifically to:
1) Increase understanding of the flow physics of 3D transonic wing buffet,
2) Increase understanding of the parameters that affect 3D buffet,
3) Develop an industrialised Computational Fluid Dynamics (CFD) method able to predict buffet for routine use in the design of buffet control devices, and
4) Improve understanding of buffet control devices to improve performance of next generation wings.

Project Results:
Wind tunnel testing at the Aircraft Research Association:

Transonic wind tunnel testing was undertaken on a large half-model of a representative transport aircraft wing/body configuration. The model was tested at typical transport aircraft cruise conditions, for a ‘clean’ configuration and two configurations with passive flow control concepts applied. The model was heavily instrumented with conventional force and pressure transducers to provide baseline buffeting data consistent with previous work, but the most important aspect of the work was the use of Dynamic Pressure Sensitive Paint (DPSP) to measure unsteady surface pressures. Although DPSP has been previously applied to buffet testing at laboratory scales, this was the first ‘production’ use of DPSP on a complete model in an industrial tunnel. The ability to capture instantaneous pressures at high spatial and time resolution over the entire surface of a wing in buffeting conditions proved of critical value in the programme, enabling flow features to be observed that would have been impossible with conventional discrete transducers. For example, previously unreported inboard and outboard-running waves superimposed on the main shock structure were identified.

CFD studies at the University of Liverpool:

Simulations of the flow over the wind tunnel model were performed using the unstructured finite volume solver TAU, developed by DLR and widely used in the European aerospace sector. Steady-state Reynolds-Averaged Navier-Stokes (RANS) solutions were used to optimise turbulence models, and to identify the onset of significant flow unsteadiness. Unsteady flows were mostly simulated suing the Unsteady RANS (URANS) approach, with additional computations using Delayed Direct Eddy Simulation (DDES). Both methods were able to reproduce the wave systems seen in the experimental data. The more computationally-expensive DDES simulations generally gave a better match to experiment, but the success of the cheaper and faster URANS simulations has positive implications for development of an industrialised buffet prediction methodology.

Overall results from BUCOLIC:

The BUCOLIC programme has made significant contributions to understanding the flow physics of 3D wing buffet, the parameters that affect it, and means to control it.

The majority of previous theoretical and experimental work on transonic buffet has focused on 2D aerofoil flows, where the buffet mechanism is generally agreed to be an interaction between shock and boundary-layer, with a chordwise feedback loop leading to large-amplitude shock oscillations at a relatively high frequency. The assumption has been that (as long as the aspect ratio is high enough) this mechanism also drives buffet on 3D swept wings, leading to buffet alleviation concepts based on control of shock-induced separations – for example extensive vortex generator (VG) arrays. However, the experimental and computational results from BUCOLIC have shown that the 3D case is much more complex, being dominated at buffet onset by lateral rather than longitudinal instabilities, with superimposed wave systems running inboard and outboard. The associated frequencies are much lower than the 2D case, and show significant intermittency. It is clear that additional cross-flow feedback loops are present in the 3D case.

Much work remains to be done in understanding fully a flow field which is much more complex than anticipated at the start of the project – nevertheless two important observations can be made.

Firstly, comparison of integrated unsteady pressure loads with measured model responses has demonstrated the importance of the interaction between aerodynamic forcing and structural response modes. The intermittent nature of lateral shock oscillations at buffet onset (the point of most interest to an aircraft designer) gives very different characteristics from the 2D case, so that the initial appearance of local flow unsteadiness/flow separation normally taken as a buffet indicator is not necessarily well correlated with the onset of airframe vibration.

Secondly, the dominance of 3D instabilities at buffet onset suggests that conventional flow separation control concepts may not be appropriate. This is borne out by the experimental test programme, where a ‘sparse’ VG array was as effective at controlling buffet onset as a standard array, despite there being far too few devices to control large-scale shock-induced separation directly. Preliminary analysis of unsteady DPSP results suggests that the VGs are acting in a similar manner to fences, with the device wakes constraining lateral oscillations of the wing shock. Simply reducing the device count in a conventional flow control array will have an immediate benefit in a lower drag penalty; however, a greater benefit will be likely to accrue from looking at devices that control local cross-flow instabilities rather than downstream flow separations.

A further objective of BUCOLIC was to develop an industrialised CFD method able to predict buffer for routine application in wing design. The unexpectedly complex nature of transonic 3D wing flows on representative transport aircraft wings has precluded the achievement of this objective, but very considerable progress has been made. On the experimental side, the extensive instrumentation on the wing/body model tested has provided significant additional data to support conventional methods already in use in industry, typically based on identifying ‘breaks’ in static aerodynamic characteristics. On the computational side, the intention was to develop a reduced-order methodology (ROM) based on steady RANS data, in a similar manner to previous work on 2D aerofoil buffet. RANS and URANS computations within the project have demonstrated that the relevant flow physics can be captured in this method, but the extension to 3D wings has proven a significant numerical challenge. Computation costs for standard methods for extraction of eigenvalues of the fluid Jacobian matrix become prohibitive for large 3D cases, so an alternative approach is needed. A new iterative solver has been implemented for the TAU code, but had not progressed beyond prototype stage at the end of the BUCOLIC project; subsequent work has been very encouraging and development is continuing at the University of Liverpool.

Potential Impact:
The BuCOLIC project has succeeded in increasing understanding of the flow physics of 3D wing buffet, and of the effectiveness of flow control devices. The groundwork has also been laid for an industrialised ROM method for buffet prediction, which continues to be developed.

Although much work remains to be done, it has been demonstrated that previous work on 2D aerofoil buffet is not generally applicable, and that attention needs to be focused on the 3D nature of transport aircraft wing buffet. This focus will be likely to lead to new innovative concepts for control of wing buffet and in the longer term for design of buffet-free wings that will avoid the flight performance penalties associated with current concepts. This will permit aircraft designers to more effectively optimise wing size for cruise, resulting in lighter, cheaper aircraft that burn less fuel.

Results from both experimental and computational studies have been presented at a wide range of technical conferences and meetings. Interest from industrial stakeholders has been high and discussions are continuing on specific applications of some of the results obtained. A large experimental and computational database has been generated, much of which remains to be fully analysed. Preliminary discussions have been held with interested universities and RTOs, and it is anticipated that significant additional value will be gained from collaborative work, specifically on understanding of the lateral shock instabilities and the effectiveness of passive flow control.

List of Websites:
http://www.ara.co.uk

Simon Lawson
slawson@ara.co.uk
Aircraft Research Association
Manton Lane
MK41 7PF