Investigation And Prediction On High-speed Train Multi-Physical-field Excitations And Full-spectrum Interior Noise

Three aspects of work were carried out to accomplish the research on the CRH3 high-speed train full-spectrum interior noise,including the interior sound quality assessment,the sound insulation property evaluation of extruded aluminum panels,and the prediction of multi-physical field noise sources.First of all,to overcome the disadvantages of the existing loudness algorithms,a novel signal-adaptive Moore loudness algorithm(AMLA)was introduced.AMLA defined the signal's equivalent rectangular bandwidth(ERB)spectrum as input,which could obtain higher efficiency and accuracy.Through the assessment of the interior noise signals acquired by on-line experiment,it is found that the overall loudness varies greatly below 300 km/h or above 350 km/h,and the sound quality is relatively stable between 300-350 km/h,where the specific loudness is mainly distributed within 20 erbr.Therefore,the simulation velocity and the upper analysis frequency band are defined as 350 km/h and 2k Hz.Secondly,the fundamental research on sound transmission loss(STL)of the extruded panels and body-in-white(BIW)were conducted by structure-sound coupling method,and the influence of structure parameters such as the interconnecting angle,inner height and thickness of the plate were discussed.The results show that panels with 50°-55° angle or 80 mm inner heightare suitable for high structural strength requirements,and panels with 35°-40° angle,50 mm or 65 mm inner height for the lightweight side and top BIW area.Thickness increase of the interconnecting plate will reduce the structural stiffness,leading to the STL valley shifting to lower frequency,and it should be thicker than that of the upper or lower plate.Through sound contribution analysis of different BIW areas,the sound insulation property of flat roof panel is proved to be best,followed by the side and floor panel,and the curve roof panel possesses the worst STL performance.Thirdly,the characteristics of multi-physical-filed noise source at 350 km/h were analyzed.The pantograph and carriage surface aerodynamic noise was fluid-borne noise,and the noise distribution and propagation was predicted by large eddy simulation-high order finite element method(LES-HOFEM)and nonlinear acoustic solver(NLAS)method respectively.In the nearfield of pantograph,windshield and bogie area,the vortex shedding or the self-excited resonance frequency was prominent,leading to the generation of vortices at different scales,and the unenven aerodynamic noise source distribution on the carriage surface.Wheel/railway noise was structural radiation noise.The vertical contact force was predicted by rigid multi-body dynamics analysis using the latest China high-speed track irregularity.Then,in consideration of sound scattering effect of the carriage surface,the corresponding structure-sound coupling model was built to predict the wheel/railway radiation noise.The wheel noise reveals an evident directivity with petaloid distribution and unenven attenuation.The main excited aera locates in the bottom of the carriage bogie,and the peak frequency band is 1.25k-1.6k Hz.In the acoustic shadow aera of side and roof surface,the wheel noise is largely attenuated.The rail noise has the similar characteristics of line noise source,which mainly distributed on the bottom surface of the whole carriage,including the equipment cabin,and the peak frequency band is 800-1k Hz.Finally,the statistical vibration&acoustic energy flow(SVAEF)method was proposed to predict the full-spectrum(50-2k Hz)interior noise of a trailer carriage at 350 km/h.SVAEF method loaded the multi-physical-field noise excitations to the outer acoustic cavities,and vibration excitations to the structure subsystems.By defining the coupling parameters between different subsystems,the transfer paths of sound and vibration energy to the interior were built.With the exterior excitations and STL of composite panels achieved,the interior noise was then calculated and validated by on-line experiment.Based on the SVAEF model,Moore loudness contribution analysis was conducted to find out the main vibration and noise transfer paths into different areas.With the application of web-mouted noise shielding and dynamic vibration absorber on the original wheel and rail structure,the acoustic response at both ends of the carriage decreased evidently between 630-2k Hz,and the Moore loudness were decreased by 8.7%and 4.2%respectively.The optimization could be perceived by auditory system.