The Key Modeling Technology Of Hybrid FE-SEA And Its Application In The Optimization Design Of Acoustic Package For Commercial Vehicle Cab

According to the overall requirements of the "Made in China 2025" development plan,automobile design gradually develops towards energy conservation,safety,comfort and other directions.NVH performance is an important index to measure the "sense of quality" of automobile.In recent years,NVH performance has gradually become the focus of attention of automobile manufacturers.Traditional NVH performance research focuses on the low frequency band 0-200 Hz.With the deepening of vehicle NVH performance research,the problem of medium and high frequency vibration noise has been widely concerned.This project relies on "National Key R&D Program Project","Chongqing Science and Technology Innovation Fund Project" and "SAIC Seed Fund Plan".Focusing on the key problems of hybrid FE-SEA method in the design of acoustic package for commercial vehicle cab,in-depth research was carried out on the identification of hybrid FE-SEA modeling parameters,the connection and modification of hybrid FE-SEA modeling,the construction and modification of smooth domain of hybrid FE-SEA model,and the identification of acoustic parameters of porous materials of hybrid FE-SEA model.It lays a theoretical foundation for the accurate modeling and the accurate prediction of intermediate frequency of the acoustic package in the cab of commercial vehicle.The main research contents and results of this paper are as follows:Aiming at the parameter identification problem of hybrid FE-SEA model,this paper establishes the effective coupling loss factor parameter model of hybrid FE-SEA model by using the input power method and the inverse solution method of coupling matrix.In this paper,the parameter identification problem of mixed model under the coupling condition of multiple subsystems is studied,the coupling matrix solution method is proposed,and the displacement response of deterministic subsystem and the energy response of stochastic subsystem are derived.On this basis,the energy matrix is normalized,and an improved matrix iteration method for solving the inverse matrix of energy ratio density is proposed by introducing intermediate variables and error transfer on the basis of principal element weighted iteration algorithm.The results show that the parameter identification method has a high fitting degree with the measured values in the middle and low frequency range,which verifies the effectiveness of the identification method for effective coupling loss factor of hybrid FE-SEA model.For mixed FE-SEA model of "mixed-point" connection modeling method and correction problem,using polar coordinate system under the condition of plane wave superposition principle,the establishment of a "mixed-point" connection plane outside the tensile and shear wave motion equation,the cartesian coordinate system is studied under the condition of "mixed-point" relay with the wave number problem.According to the wave propagation characteristics of thin plate structure,a method of "hybrid-point" connecting radiation radius correction factor was proposed.The results show that the correction method of radiation radius has high precision in the low frequency range.Aiming at the modeling method and correction problem of the "mix-line" connection of hybrid FE-SEA model,the triangular waveform function model of the "mix-line" connection is established by using the linear difference method.In this paper,the direct field stiffness matrix of "mix-line" connection under the condition of nodal coordinate system is studied,and according to the shape function of "mix-line" connection under the condition of wave-number space,the shape function "mix-line" correction method of the mixed model is proposed.The results show that the system response can be modified effectively by the "mix-line" shape function at the modified joint.Aiming at the problem of smoothing domain correction of the deterministic subsystem of hybrid FE-SEA model,the stiffness matrix models of two-dimensional and three-dimensional smooth domain are established by using the method of nodal sound pressure and interpolation function.In this paper,the smooth domain correction problem of two-dimensional finite element model is studied and the method of deterministic subsystem correction in mixed model is proposed.The stiffness matrix of the smooth domain of the deterministic subsystem is derived.On this basis,the optimized smooth domain is constructed for the two-dimensional finite element model.The global optimal smooth domain model is constructed by reducing the degree of freedom of two-dimensional smooth domain and discretizing the unit size.The calculation error caused by the stiffness of the system is effectively solved.The results show that the FE-SEA model optimized by smoothing domain is more accurate than the existing smoothing domain processing method,which verifies the effectiveness of the modified method for the smooth domain of the deterministic subsystem in the mixed model.In mixed FE-SEA acoustic parameter identification model of porous materials,aiming at the existing mixed models ignore porous error which is caused by the coupling effect of the acoustic parameters of problem,according to "theory of particle swarm algorithm hybrid FE-SEA model is established in porous materials acoustic parameter optimization model,is proposed based on hybrid FE-SEA acoustics parameters of reverse recognition algorithm of the model.Through the physical relationship between the acoustic parameters in the Johnson-Champoux-Allard theoretical model,such as flow resistance rate,porosity,bending degree,viscous characteristic length and thermal characteristic length,and the acoustic envelope absorption coefficient in the mixed FE-SEA model,the difference between the acoustic material's comparative absorption coefficient and the measured value of "impedance tube method" was solved.The inverse identification of acoustic parameters in the hybrid FE-SEA model is realized by introducing the parameter error correction method to iterate the global optimal solution of the acoustic coefficient.The results show that,compared with the existing technology,the reverse recognition algorithm fully considers the coupling effect of hybrid FE-SEA model and acoustic parameter multi-system,with high computational efficiency and small parameter identification error.Finally,according to the proposed hybrid FE-SEA model modeling and parameter identification method for heavy commercial vehicle cab acoustic performance of acquiring parameters and model verification,the accuracy of the models,the build agent model acoustic package parts to commercial vehicle cab system optimization,and vehicle air tightness is optimized,and the optimization effect has carried on the real vehicle testing.