FROM NUMERICAL MODELING TO ACOUSTIC VIRTUAL PROTOTYPING

Acoustic performances represent an important business aspect both referred to aeronautical product than transportation means in general. Very often, numerical modeling represent a powerful tool for the prediction of expected noise level once structural characteristics and noise source specification have been defined. These model may become very complex when strong interaction of structural and acoustic elements are involved as wells as the frequency range of interest increase. In any case, the availability of different, complementary approaches as those based on FEM, BEM, SEA, Ray Tracing , among the others, let assume that appreciable results can be achieved. In this scenario, it can be stated that one of the bigger limitation of these technique is the unavailability to investigate the subjective (more that objective) response to noise exposure that in many case in considered an important parameters when comfort or annoyance aspect need to be evaluated. That is the reason why deep interest is going to be dedicated to the “acoustic virtual prototyping” intended as the auralization process of a numerical simulation, that would guarantee the opportunity to “hear” the noise more than analyses its “technical descriptors”. It is important to underline that the not-stationary peculiarities of the phenomena, require that relative implementing tools have to be very fast in the signal computation as extremely precise in the signal reconstruction to guarantee the correspondence of “virtual “ noise to real expected one. The present work will present some results related to the implementation of a specifically designed digital FIR filters that, starting from results of “standard” numerical models (FEM/BEM or others), are able to reconstruct the time related system response and to digitalize the expected noise. First test have been performed on a geometrical simple box set-up, in relation to which both numerical model and experimental test campaign have been realized. The MATLAB®/Simulink® implemented filters have showed good results in signal reconstruction opening to further application at more complex systems and real-time implementation.