Genetic Optimizations of EC135 Noise Abatement Flight Procedures Using an Aeroacoustic Database

Abstract

This paper describes the implementation of the DLR SELENE computation chain, able to predict the noise of an arbitrary flight, in a genetic optimizer and the resulting optimization of noise abatement landing procedures for the EC135 helicopter. At each iteration of an optimization, a complete procedure noise footprint is calculated starting from way points prescribing the position and velocity of the helicopter. The noise simulation relies on the use of a database of noise directivity versus flight conditions, in the present case resulting from 2004 flight tests. The optimizer then adjusts the way points in order to minimize the noise on ground. Procedures that do not satisfy safety, comfort and flyability constraints are rejected. A series of such optimizations has been launched and exhibits some trends. The confrontation of each optimized procedure to a torque/velocity diagram shows the relevance of the already observed torque-noise correlation. It seems that the optimized procedures are quite robust to side wind and helicopter mass variations. On the contrary, a headwind component has a clear impact on the optimized procedures and on their noise footprint. Several optimized procedures were validated during flight tests in 2008. A good agreement between simulated and measured noise levels is observed. The flight tests confirm the predicted noise reduction by 10 dB SEL between 3 km and 1 km before the landing point. The optimized flight procedures are evaluated as well flyable. A modification of the helicopter fin would make them completely comfortable.