An SEA Modeling for Wind Noise Using Wavenumber-Frequency Spectrum

Abstract

The wind noise typically covers broad band frequency range, for example, from 100Hz to 5kHz in the case of automobile. For the high frequency noise prediction, Statistical Energy Analysis (SEA) is the most suitable method considering the number of modes at the frequency of interest. In SEA development progress, the excitation model for Turbulent Boundary Layer (TBL) has been developed based on the Corcos type models. In low mach number flow, however, due to the difficulty in describing the wavenumber component distribution of the fluctuated surface pressure caused by turbulence using the TBL excitation model, especially in describing the acoustic wavenumber component which is sensitive to transmission to passenger compartment, SEA has not been applied to the analysis of the wind noise of automobile. In this paper, the turbulence generated spartial distributed fluctuated surface pressure computed by Computational Fluid Dynamics (CFD) is decomposed to the wavenumber-frequency spectrum for describing the wavenumber component in detail. The SEA modeling for predicting the wind nosie using the wavenumber-frequency spectrum is proposed considering the main path to the interior cavity. Then, the validation of the proposed method is done by comparing with the measurement and shows good agreement.