Prediction of Buffet Loads Using Artificial Neural Networks

The use of artificial neural networks (ANN) for predicting the empennage buffet pressures as a function of aircraft state has been investigated. The buffet loads prediction method which is developed depends on experimental data to train the ANN algorithm and is able to expand its knowledge base with additional data. The study confirmed that neural networks have a great potential as a method for modelling buffet data. The ability of neural networks to accurately predict magnitude and spectral content of unsteady buffet pressures was demonstrated. Based on the ANN methodology investigated, a buffet prediction system can be developed to characterise the F/A-18 vertical tail buffet environment at different flight conditions. It will allow better understanding and more efficient alleviation of the empennage buffeting problem . RELEASE LIMITATION Approved for public release Published by DSTO Aeronautical and Maritime Research Laboratory 506 Lorimer St Fishermans Bend, Victoria 3207 Australia Telephone: (03) 9626 7000 Fax: (03) 9626 7999 © Commonwealth of Australia 2001 AR-012-019 September 2001 APPROVED FOR PUBLIC RELEASE Prediction of Buffet Loads Using Artificial Neural Networks Executive Summary The F/A-18 fighter aircraft experiences random fluctuating pressures on its empennage surfaces caused by impingement of burst LEX vortices during critical high angle of attack manoeuvres. This severe buffet environment shortens structural fatigue life and causes premature structural failures. Accurate prediction of random pressure fluctuations on a vertical tail is quite difficult due to complexities in the interaction between the highly turbulent flow field behind burst LEX vortices and empennage structure in different flight regimes. Despite progress in our ability to predict the empennage buffet made during the last decade, more accurate and robust buffet prediction methods must be developed to support fleet management decisions. This work investigates the feasibility of using Artificial Neural Networks (ANN) for predicting the empennage buffet pressures as a function of flight conditions. The method is dependent on the availability of experimental data to train the ANN algorithm and is able to expand its knowledge base with additional data. Full scale F/A-18 tail buffet test in the 80ft X 120ft test section of the NASA Ames National Full-Scale Aerodynamics Complex provided the initial database for the development and assessment of an ANN-based buffet load prediction method. The study revealed that artificial neural networks have a great potential as a method for modelling the complex nonlinear relationships inherent in buffet data. Initial assessments indicated that neural networks are able to accurately predict RMS values and frequency content of unsteady buffet pressures. The ANNs have the ability to extract the essential features from many input combinations to produce an accurate output and generalise well for new conditions by detecting features of the inputs that have been learned to be significant. Based on the ANN methodology investigated, a buffet prediction system can be developed to provide detailed information about the F/A-18 vertical tail buffet environments through the use of additional experimental and flight test data. It will allow for better understanding of empennage buffeting problems and can be used in fatigue usage monitoring systems for fleet aircraft. Results of the work contribute to DSTO’s existing body of knowledge on empennage buffet and can assist in the F/A-18 International Follow On Structural Test Project (IFOSTP) fatigue test on the aft fuselage and empennage.