Principal Components for Structural Acoustics Analysis

The proposed innovation is a new computational and analytical structural acoustic technique for aircraft interior noise prediction. Its innovative quality resides in the unifying theory by which conventional finite element analysis (FEA) and statistical energy analysis (SEA) are related through the use of Principal Components (PC) analysis. This linking of FEA and SEA methods provides the ability to predict structural acoustic response in the so-called mid-frequency range where neither conventional FEA nor SEA approaches have given satisfactory results. FEA produces classical frequency-ordered modes that typically fail to correlate with test modes in this range. Analytical SEA fails when modal density is low. Experimental SEA is also problematic whenever poorly conditioned energy matrices must be inverted to obtain coupling and loss coefficients. The new technique involves combining FEA for finely meshed finite element models with PC analysis to derive energy-ordered modes i.e. principal components, for the coupled structural and acoustic components. These energy modes not only facilitate analysis-test correlation, but also provide a means for rigorous statistical quantification of modeling uncertainty and predictive accuracy. The new PC/SEA modeling approach is expected to significantly improve the accuracy of aircraft interior noise prediction.