Simulation And Calculation For Sound Quality Of Vehicle Internal Airflow Noise Based On The SEA Method

Vehicle interior noise caused by airflow have a tremendous influence on auditory perceptionof rides personnel When the vehicle is in motion, especially in driving faster today, this problem isbecoming more and more outstanding. But before the research methods of noise measure withA-weighted sound pressure level has been unable to reflect the true feelings of rides personnel, soout of the research needs of vehicle noise, sound quality concept came into being. On this basis,this paper used the SEA(statistical energy analysis) method and combining with the sound qualityobjective evaluation analysis technology simulate and forecast the sound quality for the internalair flow noise of vehicles. Complete the main work is listed below:Set up the SEA model of the vehicle airflow noise, and according to the modal similarprinciple, model is divided into32subsystems, each subsystem are simple and regular subsystems,so can use theory expression calculate their mode density, the internal loss factor and couplingloss factor.Determine the SEA model subsystem basic parameters of the model and external incentives.SEA methods including modal density, coupling loss factor and internal loss factor in three basicparameters, their correction or not relate directly to the accuracy of the establishment of the SEAmodel. Only in the situation of3person get accurate estimation, the SEA model will predict thevehicle internal airflow noise rightly. Using the classical theory method to calculate theparameters of the different types of subsystem.Using the Grid software established vehicle CFD (computational fluid dynamics) model,determine the calculation domain and mesh units, through the Fluent software determined theboundary conditions and calculated, finally use unsteady calculation method to get the externalpneumatic pressure spectrum of the monitoring points for each subsystem, take average on thepressure spectrum of the monitoring points and then get the average pneumatic pressure spectrumof the needed subsystem, namely external incentives of the model.Make the calculated three basic parameters and external airflow pulse pressure of the eachsubsystem for SEA model as input, joining into the SEA model of the airflow noise, get average pressure level response of vehicle internal airflow noise through the simulated calculation, andcompared them with the actual measured value of the eight kinds of domestic cars in foursteady-state condition speeds, the compared results showed that the vehicle internal pressure levelsimulation results matched well with the practical case of the driver’s ear side pressure value in themore than800Hz high frequency band, the relative error is quite low, and absolute error less than2dB.This indicates the established the SEA model of airflow noise is correct, and can be used inthe analysis of high frequency noise.Use subjective evaluation method (rank evaluation method) to evaluate the collected airflownoise sample subjectively, through the established SEA model of vehicle airflow noise before andthe common sound quality objective parameter mathematics calculation model calculated theobjective parameter values of sound quality of each sample, with the use of SPSS (statisticalproduct and service solution) software to do related analysis between the subjective evaluationresult and objective evaluation parameters, it was found that the loudness, sharpness androughness and subjective evaluation value has a good correlation, which can set up a sound qualitymultiple linear regression equation with loudness、sharpness and roughness as independentvariable. Use of SPSS software again solved the multiple linear regression equation, according tothe stepwise of the law, eliminating roughness, choose loudness and sharpness eventually intomodel. Their statistical significance was proved after test and analysis to sound quality objectivequantitative model of vehicle internal airflow noise. This means that the established sound qualityobjective quantitative model can be used into analysis and prediction for steady-state condition ofvehicle internal airflow noise. The prediction with this model showed that the airflow noisefidgety degrees were between6~10in the universal10cents values, accords with practice.