Musser, C. and Rodrigues, A., "Sound Package Optimization for Fully-Trimmed Vehicle Using Statistical Energy Analysis Technique," SAE Technical Paper 2007-01-2873, 2007, doi:10.4271/2007-01-2873.
Predictions for vehicle interior noise and vibration levels can be made analytically using Statistical Energy Analysis (SEA), particularly for the middle and high frequency ranges. A SEA model can be effectively used together with some minimal baseline measurements to identify and predict changes to the dominant airborne and structureborne paths and to predict the effects that changes to the sound package or structure will have on these paths. Especially for relative changes in noise or vibration level, good accuracy is expected for acoustic or vibration response points such as driver's ear. An SEA model that has been validated with some baseline and various data, which may even come from a previous generation vehicle or component-level testing, can predict if a change to a sound package component will achieve a transmission target or if a proposed change will not be effective. This information is very important to guide to which part of the sound package the cost and design effort should be directed. Component-level measured data from a baseline sound package and a proposed sound package design change is presented. Analytical comparisons to this measured transmission loss and absorption data are shown and the uses of this validated component-level model are discussed. An extension to the fully-trimmed vehicle using this information in the context of a full-vehicle SEA model is shown. Predicted change in driver's ear SPL using a full vehicle SEA model with these sound package configurations is presented, and points about the use of these models and the energy flow results are made. General conclusions about this methodology and recommendations for future use are given.