Vibro-acoustic analysis of a vehicle integrated with design of experiments methodology using three performance criteria

Abstract

The interior noise inside the passenger cabin of automobiles can be classified as structure-borne or airborne. In this study, we investigate the structure-borne noise, which is mainly caused by the vibrating panels enclosing the vehicle. Excitation coming from the engine causes the panels to vibrate at their resonance frequencies. These vibrating panels cause a change in the sound pressure level within the passenger cabin, and consequently generating an undesirable booming noise. It is critical to understand the dynamics of the vehicle, and more importantly, how it interacts with the air inside the cabin. Two methodologies were used by coupling them to predict the sound pressure level inside the passenger cabin of a commercial vehicle. The Finite Element Method (FEM) was used for the structural analysis of the vehicle, and the Boundary Element Method (BEM) was integrated with the results obtained from FEM for the acoustic analysis of the cabin. The adopted FEM-BEM approach can be utilized to predict the sound pressure level inside the passenger cabin, and also to determine the contribution of each radiating panel to the interior noise level. The design parameters of the most influential radiating panels (i.e., thickness) can then be investigated to reduce the interior noise based on the three performance metrics. The performance metrics selected for this study are "Percentage over 80dBA", "Max Amplitude", and "Idealized Performance Error". Design of experiments (DOE) technique was employed to understand the relationship between the design parameters and the performance metrics. The components that have the highest contribution to the sound pressure levels inside the cabin are identified. For each run, the vibro-acoustic analysis of the system is performed, the sound pressure levels are calculated as a function of engine speed and then the performance metrics are calculated. The highest contributors (design parameters) to each performance metric are identified and regression models are built. These regression models can be used in future studies to employ optimization runs to find the optimum configuration of the panel thicknesses to improve the sound pressure level inside the cabin.