NVH – Structure-borne sound correlation of a Gasoline High Pressure Pump. A Front-Loading digital NVH method to help the design of components

Either from a legislative point of view or because of OEM demands, the automotive industry is increasingly facing of reducing vibration & noise in the vehicle. This vibration and acoustic troubles from separated elements of the full vehicle assembly, leads the group Powertrain Engine System of Continental Automotive to design all new hydraulic fuel components and systems with challenging NVH targets. This approach becomes more and more a must, due to the limited time to cost development to stay competitive with a high value to the final customer through stringent low noise targets.
The prediction of acoustic noise of WET components in the audible frequency band is a multi-disciplinary task, which combine a lot of fields and skills. Reaching an accurate digital prediction of the acoustic level for such a real WET component, means to master several steps. At first, the structure-borne sound of the total assembly with its located sources of vibration and their energy propagation through transfer paths inside the whole structural mechanics assembly. At second, the coupling influence of fluid cavities to structural parts, which may change considerably the structure-borne sound in WET condition. At last, the sound emission, which consist of radiating the structural mechanics in the free field with all obstacles and to hear the final emitted acoustic noise.
The proposed paper will be more particularly focused on the structure-borne sound applied to a gasoline high pressure pump. The identification of mode shapes of the complete assembly is essential to get an accurate knowledge of the main structure-borne sound contributors. The target being to establish a structural mechanics FEM model, the experimental analysis is then very soon connected to the model definition. Although it’s a low sizing component, the NDOF can increase drastically to several millions. Therefore, the interaction between the digital model and the experiments is at each stage very present. The most significant stages of the entire correlation will be shown, the results will be evaluated and some typical use cases in interaction with the component designers exposed.