Modeling and Numerical Analysis of Automotive Aerodynamic Noise Generation and Transmission Considering Equivalent Nonlinear Sealing

摘要

Aerodynamic noise transmits through automotive window, causing great adverse influence on comfortability and noise-vibration-harshness (NVH) performance. However, the complicated external turbulent air flow, as well as the internal metal-rubber nonlinear sealing constraint, makes the mechanism of aerodynamic noise generation and transmission very difficult. Regarding the complex exterior aerodynamics-induced load and nonlinear metal-rubber interaction and constraint, an efficient two-step numerical prediction method is presented in order to study the mechanism of its generation and transmission. The first step uses the commercial ANSYS-Fluent computational fluid dynamics (CFD) analysis based on the shear stress transport (SST) - turbulence kinetic energy (k) - the rate of dissipation of turbulence kinetic energy ε (epsilon) model and Lighthill’s noise source theory. For low Mach number and high Reynolds number flow like the flow around a vehicle body, dipole source is regarded as the dominant contribution and can be obtained by the broadband noise source model. Exterior turbulent flow field of a full-scale automotive is established and near-field sound power distribution of automotive window has been obtained, which are both subsequently input to the acoustic model to investigate the noise generation mechanism. The second step consists of the numerical prediction of noise transmission through automotive window. Nonlinear spring-based surrogate model for seal nonlinear constraint is proposed and verified by modal experiment. Based on SAE J1400 reverberant-anechoic measurement standard, a numerical prediction model of the sound transmission loss (STL) is constructed using commercial vibro-acoustic solver Actran. New automotive window structural design by non-uniform density distribution is proposed to optimize the STL property. The present methodology of STL modeling and numerical prediction provides valuable instructions for performance optimization of automotive door under high speed driving condition.