Airborne sound source characterization for railway noise predictions - based on vibration measurements and numerical simulations

摘要

Within the European rail initiative Shift2Rail, funded by the EU, the FINE1 project aims to improve state-of-the-art noise modeling for railway systems. Current prediction tools are able to compute the sound pressure field distribution around full rail vehicles. Once the pressure field distribution of a source has been determined and the transfer functions (TF = SPL_(Spatiamstrib) - SWL_(SoundSource)) are estimated, the resulting frequency-dependent spatial pressure distribution for sources specified by sound power can directly be quantified and visualized as blocked pressure. In these TF calculations the sources are predominantly integrated as monopoles: a solution has to be found for the implementation of the directional characteristic. The necessary input for some sources, at least with direct sound pressure/intensity measurements, is difficult to obtain (due to unavailability of anechoic/reverberation rooms). One feasible solution for vibro-acoustic sources, in principle as standard available, is the determination of airborne sound power by use of vibration measurements with 'fixed/adequate' radiation. The paper starts with a simulation of an acoustic measurement for an exemplary gearbox at a defined operating point. Besides the required airborne sound power and sound directivity of the gearbox, the simulated measurement results provide also a full vibration data set for the complete outer surface. The objective of this investigation is to 'measure' the 'same' sound power and directivity with a limited, selected number of vibration sensors. The used tools, software as well as hardware, shall be available within the framework of the railway industry.