Search Results

This paper presents a validation project carried out by the Ford Motor Company and Numerical Integration Technologies (NIT) concerning the numerical prediction of engine noise radiation. The importance and the difficulty of building a structural model (predicting accurate vibration data) are discussed. A new methodology has been developed. This method consists of a modal expansion of experimental data and allows to introduce experimental vibration measurements in the numerical approach to enhance the quality of the predictions. Provided that the structural behavior is correctly assessed, the project has shown that the BEM-based acoustic predictions agree remarkably well with test data.

Global acoustic sensitivity refers to a technique allowing to define the optimal modifications to a structure in order to improve its acoustic performance. The methodology is based on a structural finite element model for providing the structural sensitivities and an acoustic boundary element model selected to evaluate the acoustic sensitivities. Both models are combined within a design optimization algorithm, leading to the prediction of the optimal changes of the design parameters. Numerical implementation aspects are discussed, showing how the methodology is designed to handle real-life models. The developed technology is applied for reducing the acoustic radiation of vehicle components such as engine oilpan or gear housing and for improving the acoustic characteristics of car passenger compartments.

The prediction of the acoustic radiation of vibrating structures is traditionally based on the acoustic boundary element method. This approach leads to good predictions, provided that the dynamic behavior of the system is well known. This paper presents the integration of experimental vibration analysis with numerical acoustic radiation prediction. The developed methodology involves several steps: (i) measure the accelerations on the surface of a vibrating body, (ii) expand these experimental data onto a numerical model, using the modal information of a finite element model, and (iii) run a boundary element analysis to determine the acoustic radiated field. This procedure presents the advantages that the dynamic behavior is as accurately as possible defined and that the boundary element approach opens wide results interpretation (contribution and sensitivity analysis…). A real-life application example (car engine) is shown to illustrate and validate the developed methodology.