Application of Statistical Energy Analysis to Noise Prediction of Co-generation System

This paper describes the application of statistical energy analysis (SEA) to predicting sound power radiated from co-generation system enclosure. To predict vibration and noise accurately by using SEA, it is important to estimate parameter called loss factors. In this study, loss factors were estimated by power injection method. Next, the noise radiated from enclosure surface was predicted by the obtained vibration and radiation efficiency of enclosure panels. As a result, the calculated sound power was relatively corresponding to measured sound power. Finally, the sound power from modified enclosure was predicted. Coupling loss factors related to a modified subsystem were estimated by ratio of the number of structure modes. By using these steps, the noise from the system was reduced.