Acoustic optimization design of window of high-speed train under flow-induced vibration

Abstract

In this study, the sound optimization design of a high-speed train window structure under a turbulent boundary layer (TBL) pressure fluctuation is investigated. First, the Corcos model is used to characterize the cross-spectrum of the dynamic surface pressure, and analytical solutions of the window speed response, radiated sound power, and transmission loss (TL) are obtained. The effects of the cavity thickness, cavity damping, panel thicknesses, and panel damping on the TL were investigated by numerical examples. In the structural sound optimization process, the quality reduction and the sound response are essentially in conflict with each other; therefore, non-dominated sorting genetic algorithm (NSGA)-II is adopted for a dual-objective design. The numerical results show that the train speed has a significant influence on the TL of the window structure, which decreases by 4 dB when the train speed increases by 100 km/h. With the increase in the train speed, the effect of improving the TL of the window structure by increasing the thickness and damping of the window cavity gradually decreases. The TL of the window structure is more sensitive to the thickness of the window panels than that of other panels parameters; however, the thickness sensitivity decreases with the increase in the speed, whereas the improvement effect of panel damping on the TL increases with the increasing train speed. The Pareto optimal design of the window structure at 350 km/h was obtained using the NSGA-II optimization design method. The optimal design reduced the mass by 41.6 % and increased the average transmission loss (TLavg) increased by 6.2 dB compared to those of the original window structure.