A frequency response function-based inverse substructuring approach for analyzing vehicle system NVH response.

摘要

A new multiple-coordinate frequency response function (FRF)-based inverse substructuring theory is developed to analyze the structure-borne NVH problems in vehicle systems applying measured structural-acoustic and vibration spectra. The proposed technique is used to predict vehicle system NVH responses, as well as free substructure FRFs and mount dynamic properties. Current techniques are not sufficiently adequate for modeling damped, moderately dense modal density characteristics that dominate mid-frequency range NVH problems in vehicle systems. Depending on the actual form of the structural coupling terms, the resultant formulations can be quite different. Three forms of the coupling matrix are assumed in this dissertation. The simplest one constitutes the diagonal form where the cross-coordinate dependency is completely neglected, and the other two more complex cases are the block-diagonal and non-diagonal representations. By using a finite element model of a vehicle, three forms of FRF-based inverse substructuring formulation are studied computationally. The net effect of the nature of the coupling formulation on the predicted mount, free substructure characteristics, and system response is examined in detail. Sensitivity analysis is performed to determine the effect of random and bias measurement errors on the performance of non-diagonal and approximate forms of the FRF-based inverse substructuring method by using perturbed substructures. The singular value decomposition theory is employed to improve the results of coupling procedure since the multi-coordinate FRF-based inverse substructuring approach can be sensitive to the measurement noise. The analysis also gives a better understanding of the effect of SVD in substructure dynamic coupling. To demonstrate the salient features of this approach, the measured data from a passenger car are used for the prediction of substructure FRFs, mounting stiffnesses, and system response. The results reveal an excellent correlation with the direct measured response. The total system response is then processed into key physical elements using the SRSS (square-root sum of squares) formulation to identify the primary controlling factors and contribution paths. Finally, the mount stiffness sensitivity analysis is conducted using the test data of another passenger car and a truck.