Prediction of high-speed train full-spectrum interior noise using statistical vibration and acoustic energy flow

Abstract

The high-speed train (HST) interior noise in low, middle and high frequency bands is predicted in different simulations due to the limitations of current numerical methods. In this paper, a new method named statistical vibration and acoustic energy flow (SVAEF) is proposed to predict full-spectrum HST interior noise directly. In the SVAEF theory, interior noise is solved in the form of energy flow, and panel vibration responses caused by mechanical and acoustic excitations are assumed to satisfy the linear superposition principle. The acoustic energy of exterior sound sources is mainly attenuated by the carriage structure in the equivalent form of sound transmission loss, while the mechanical vibration energy flows inside carriage structure and interior acoustic cavities. The interior acoustic responses are calculated under a dynamic equilibrium of the whole energy flow. Initially, a HST carriage model is established based on SVAEF theories. And all the subsystem parameters and exterior excitations are acquired properly through the measurements, simulations and empirical formulas. Among them, inverse-fast Fourier transform (IFFT) and rigid multi-body dynamics are used for track irregularity excitations and contact force, while the wheel-rail rolling noise is extracted and verified with the sound scattering effect of carriage surfaces considered. The distributions of aerodynamic noise are calculated by the hybrid method of Reynolds-averaged Navier-stokes (RANS) and non-linear acoustic solver (NLAS) approach with the pantograph system, the bogie and windshield completely considered. In addition, the equipment noise under the carriage is measured as one of acoustic excitations and the vibration of carriage panels is measured to extract mechanical excitations. Finally, the interior noise is predicted through SVAEF and compared with the measurements and other prediction methods. The results indicate that the trend of SVAEF predictions is in better agreement with measurements in the whole frequency bands, and the deviations of overall sound pressure level (OASPL) are much smaller. In conclusion, the effectiveness and accuracy of SVAEF method have been verified in the full-spectrum prediction of HST interior noise.