Research On Key Parameters Of Statistical Energy Analysis For Interior Noise Of High Speed Train

The rapid development of high-speed railway has brought a series of noise and vibration problems,among which the interior noise problem affecting passengers’ ride comfort has attracted more and more attention.As a powerful tool to deal with the dynamics of complex structures of high frequency,statistical energy analysis(SEA)is widely used to predict the interior noise of high-speed trains,and become the main method to realize the low-noise design of high-speed trains.The aim of this thesis is to study the main parameters of the interior noise prediction model established by SEA.Specifically,the noise sources generated by high-speed train:includes the wheel-rail noise and the aerodynamic noise,and the subsystem parameters,such as the damping loss factor,the modal density of the extruded aluminium panel,and the coupling loss factors between the extruded aluminium panel and the cavity.The following work is carried out:(1)The radiation and transmission loss characteristics of panels based on the modal superposition method are studied,and the panel-cavity and the panel-panel coupling loss factors are evaluated.A simple SEA model is established concluding cavity-panel-cavity subsystems,and the influence of the resonant and non-resonant transmission of the panel is analyzed.Results show that when frequency is lower than the critical frequency,the transmission loss of the panel will be underestimate if the non-resonant transmission ignored.(2)Regarding to the wheel-rail noise prediction,a wheel-rail interaction model in high frequency at high speed based on time-domain moving Green’s functions is established,simultaneously taking into account the infinite length and periodicity of the track and the flexibility,rotation and motion of the wheelset.Effects of the wheel rotation on the wheel-rail forces with different irregularities are studied.The sound pressure radiated by the wheelset and the rail on the exterior train surface using 2.5D boundary element method is determined.Results show that the effect of the wheel rotation cannot be ignored under some specific roughness excition.The sound pressure radiated by the wheel and rail on the exterior train surface is small in the middle and large at both ends,and the radiation of the wheel is dominant above 2000 Hz.(3)Regarding to the aerodynamic noise,the flow field of the pantograph area is numerically simulated based on the three-dimensional compressible viscous fluid model.The convective and acoustic pressure are obtained by the wavenumber decomposition method,and the characteristic of the convective and acoustic pressure are analyzed.The acoustic pressure of pantograph platform is also calculated by SNGR method based on the three-dimensional incompressible viscous fluid model.The acoustic pressure obtained by SNGR method agrees well with that of the wavenumber decomposition method in high frequency.(4)As for damping loss factor,the double-exponential windowing method is developed and improved to identify the damping loss factors of extruded aluminium panels.Improvements are achieved by taking into account the effect of the finite time duration of Fourier transforms and making use of the real parts of driving point mobilities.The method is easy to implement and the improvements make the choice of the decay rate in the exponential window much flexible.Only the driving point mobility is used to minimize the influence of measurement noise.Besides,results of the proposed method are compared with those of the Poly MAX method and the decay rate method,and the reliability of the proposed method is verified.(5)As for the modal density,a method based on the 2.5D finite element method is proposed to evaluate the modal density for a waveguide structure of finite length.The dispersion curve of waveguide structure is categorized by definition of wave assurance matrix(WAC),and the group velocity and modal density of different waves are calculated.Results show that negative group velocity and modal density may appear for some specific waves.Although the group velocity is zero at cut-on frequencies,but the modal density in the frequency band will still be a finite value in a finite frequency band.(6)As for the coupling loss factors,a method based on the 2.5D finite element and boundary element method(2.5D FE-BEM)is proposed to estimate the coupling loss factors structure-cavity and structure-structure coupling loss factors.The structure-cavity coupling loss factors for different waves are calculated,and the cavity-cavity indirect coupling loss factors evaluated based on the mass law and the 2.5D FE-BEM are compared.Results show that the structure-cavity coupling loss factor is the smallest when the wavenumber of the extruded aluminium panel is higher than the acoustic wavenumber.The cavity-cavity indirect coupling loss factor evaluated by the mass law and the 2.5D FE-BEM are in good agreement when frequency is higher than cut-on frequency.The related research of this thesis can provide a reference for the acquisition of the noise sources and the subsystem parameters of the interior noise prediction model established by SEA,and then further to improve the accuracy of the prediction results.