Non-modal Method of Structure Dynamic Response Identification for a Prototype Commercial Jet Aircraft Fuselage

摘要

For a high modal density, high damping, and complex structure, various well-known modal parameter estimation techniques often fails due to the high coupling between the densely packed adjacent modes. Recent advances in data acquisition and in data processing techniques make the non-modal structure identification method feasible. These methods could be one of the keys needed to solve the problems. The proposed method uses Laser Doppler Vibrometry (LDV) techniques to acquire high spatial density mobility shapes of a commercial jet aircraft fuselage at a number of frequencies. Then a Singular Value Decomposition (SVD) technique is applied to the spatial matrices to extract the dominant singular values and the corresponding singular vectors. Data reduction and noise filtering are achieved by using the dominant singular values and corresponding vectors to reconstruct the original mobility shape. A simple linear regression is used on the singular values and vectors for the prediction of the spatial mobility over the frequency range of interest.