Fast identification of transonic buffet envelope using computational fluid dynamics

Abstract

Purpose This paper aims to present a numerical method based on computational fluid dynamics that allows investigating the buffet envelope of reference equivalent wings at the equivalent cost of several two-dimensional, unsteady, turbulent flow analyses. The method bridges the gap between semi-empirical relations, generally dominant in the early phases of aircraft design, and three-dimensional turbulent flow analyses, characterised by high costs in analysis setups and prohibitive computing times. Design/methodology/approach Accuracy in the predictions and efficiency in the solution are two key aspects. Accuracy is maintained by solving a specialised form of the Reynolds-averaged Navier–Stokes equations valid for infinite-swept wing flows. Efficiency of the solution is reached by a novel implementation of the flow solver, as well as by combining solutions of different fidelity spatially. Findings Discovering the buffet envelope of a set of reference equivalent wings is accompanied with an estimate of the uncertainties in the numerical predictions. Just over 2,000 processor hours are needed if it is admissible to deal with an uncertainty of ±1.0° in the angle of attack at which buffet onset/offset occurs. Halving the uncertainty requires significantly more computing resources, close to a factor 200 compared with the larger uncertainty case. Practical implications To permit the use of the proposed method as a practical design tool in the conceptual/preliminary aircraft design phases, the method offers the designer with the ability to gauge the sensitivity of buffet on primary design variables, such as wing sweep angle and chord to thickness ratio. Originality/value The infinite-swept wing, unsteady Reynolds-averaged Navier–Stokes equations have been successfully applied, for the first time, to identify buffeting conditions. This demonstrates the adequateness of the proposed method in the conceptual/preliminary aircraft design phases.

Transonic aeroelasticity: A new perspective from the fluid mode

“…He also found the S-A model could reproduce the buffet unsteadiness with reasonable buffet frequencies but the amplitude were underpredicted. In addition to the turbulence model, a number of authors have reported similar sensitivity studies on the numerical discretization scheme, the time steps and grid resolution [50][51][52][53][54].…”

Fast Aerodynamic Calculations Based on a Generalized Unsteady Coupling Algorithm

“…The steady and unsteady 2.5D+ flow solver was demonstrated [23] around single and multi-element wing sections in low-and high-speed of flight, using laminar and various turbulence models. The flow solver was applied to study transonic buffet for a family of wing configurations [24] and in aerodynamic shape optimisation using solvers of different fidelity [25]. Since its deployment within an industrial environment, the 2.5D+ flow solver has been used in excess of hundreds of thousands of times for production.…”