In many engineering activities aeroacoustic noise is a unwanted by-product. Well-known examples are, for instance, the noise produced by a turbulent flame or the noise produced by a jet engine of an airliner. In many cases the noise is harmless but in other cases it can be a severe nuisance or even unhealthy for people who are in the neighbourhood of the acoustic source. Noise can also cause damage to equipment, mainly by causing mechanical vibrations or it can disturb considerably the primary process for which the equipment has been designed. An example of the latter case is the acoustic flame instability in a furnace. Therefore, a predictive tool for aeroacoustic noise generation will be very useful for many applications, ranging from well-known engineering problems, like the aforementioned turbulent flame, to prediction of sound produced by high speed vehicles like trains.
Aeroacoustic noise in practice is primarily linked to turbulence. Present
engineering turbulence models, like the 
or Reynolds stress
models, are not able to provide the necessary space-time information which
is needed to predict the acoustic source. Therefore, at present noise
computations are either not performed at all or they are done on an
empirical basis. In practice this frequently means that expensive measures
have to be taken when the noise turns out the be larger than expected.
A step forward in the computation of aeroacoustical noise will be the use of modern simulation techniques like Large Eddy Simulation (LES) or Direct Numerical Simulation (DNS). These simulations provide in principle all the necessary information needed for an acoustic prediction. However, the DNS technique will never become an engineering tool due to its high computational cost. LES is computationally much cheaper (nowadays a LES can be performed on a powerful PC) and has potential for engineering applications. However, the LES technique still has to be validated in the context of noise predictions. Such a validation is one of the outcomes of the present proposal.
Once the LES methodology is validated for simple-well defined flows like the jet flow of this proposal, we are more confident to use LES for prediction of aeroacoustic sound in engineering flows and show potential users the possibilities of the technique. At the moment we have the following test cases in mind: i) the noise produced by a turbulent jet diffusion flame ii) the noise produced by the wake behind a moving object (a model for a high speed vehicle). However, if requested by potential users, we could perform other test cases.