Estimation of the Aerodynamic Characteristics of Cranked Wings through Response Surface Modelling

Aerodynamic methods used in the early stages of conceptual design are necessarily simple for the sake of speed and convenience. They may also involve simplifying assumptions with a si gnificant impact on their accuracy. This may result in unrealistic levels of performance being estimated at an early stage in a design, with consequent risk later in a program. The work described here demonstrates a novel approach for estimating the aerodynamic characteristics of both simple swept-tapered and cranked wings. This approach combines analysis methods with optimal Design of Experiments to generate simple algebraic functions, or response surface models, representing the outputs of a series of more detailed analyses. Proof of concept for the response surface model technique, and subsequently, optimal Design of E xperiments is demonstrated for simple swept-tapered wings. The response surface modelling approach is subsequently demonstrated for wing planforms with up to four cranks. Application of the technique to cr anked wings has required the development of an optimal Design of Experiments method capable of handling non-linear constraints and larger sample sets. Construction of the response surface models required the automation of geometry and analysis tools to cope with large numbers of sample configurations for each problem. The resulting response surface models have been implemented in conceptual design tools and spreadsheet-based performance methods, where they have replaced ‘equivalent’ swept-tapered wing approximations, with their associated inaccuracies. This paper shows a quantitative assessment of the impact of these approximations on prediction of lift curve slope and aerodynamic centre prediction at subsonic conditions. Comparisons against experimental measurements for cranked wings are included.