Core Engine Acoustics (CEA): the application of ISVR phased array methods to separate out the combustion noise from the other engine noise sources

Aero-engine noise will continue to be a significant contributor to the noise level of aircraft at the Approach certification condition and the dominant source at Take-off. In future designs of low emission aircraft engines the noise emitted from the engine core (also called core noise) becomes more significant due to the reductions being achieved in fan and jet noise. Of equal if not greater importance, there are intense environmental pressures to reduce combustion emissions without increasing core noise and also the weight and cost of the engine.

Core noise is mainly the result of the combustion process, the ‘direct’ combustion noise from within the combustor can and the ‘indirect’ noise due to interactions between the combustor-generated unsteady flow and the turbines downstream. It is already known that the low emissions ‘lean burn’ combustor design tends to be noisier and also more unstable. To enable industry to conduct the essential trade-off studies between noise and emissions requires a validated capability to predict core noise by modelling the combustion physics, thermodynamics, aero-acoustics, and turbomachinery aerodynamics, from the combustor can, through the turbines and radiation through the jet nozzle to the community in the far-field below.

The objectives of this research programme are (1) predict the far-field combustor noise by integrating existing analytical/numerical methods available at the ISVR and using as the input existing CFD data in the form of Large Eddy Simulations (LES) of the unsteady combustion flow field, which will be provided by Rolls-Royce and (2) to validate the predictions with measured engine noise data, also to be provided by Rolls-Royce. This will entail the application of ISVR phased array methods to separate out the combustion noise from the other engine noise sources.

This integration of the computational chain from the combustion chamber to far field will be challenging, but will use existing methods called LOTAN, LINEARB and GXMunt. LOTAN is a low order linear acoustical network code which computes, within a prescribed thin annular geometry, the acoustic pressure field generated by the unsteady heat release rate data from the LES input. LINEARB describes acoustic propagation through an actuator disk model of the turbines, also with a thin annular geometry. GXMunt is a modal radiation model for an unflanged duct with a simple mean flow model.

This will provide industry with a validation benchmark for core noise modelling and prediction methodology, which is a key requirement for the aero-engine industry in the UK.

If you wish to discuss any details of the project informally, please contact Dr Brian Tester, Acoustics Research group, Email: [Email Address Removed], Tel: +44 (0) 2380 59 2286.