A study on constraint condition setting for spatial matrices identification of flat plate structures

Abstract

Elastic vibrations have become a considerable problem for recent light-weighted and structurally simplified machines to increase their performance and value. In order to study vibration reduction, numerical investigations are required, and the finite element (FE) analysis is used widely for those purposes. However, a detailed FE model for elastic vibration analysis is not easy to build, particularly in cases where the target structure size is large, such as railway vehicle carbodies. Therefore, there are demands to identify spatial matrices (mass, stiffness, and damping matrices) from experimental data and use them to investigate structural modification to reduce elastic vibrations. This study aims to identify spatial matrices under the condition that the target structure consists of a lattice-like structure (each lattice point represents a vibration measurement point) and has physically reasonable constraint relations. Using the stationary excitation test data for a flat plate structure (a 1:10 scale model of the underframe of a railway vehicle), this paper applies the FRF (frequency response function) direct method to identify the spatial matrices. The four constraint conditions are introduced in the identification procedure, and the identification results are tested by measured FRFs and modal characteristics. It has been demonstrated that considering the connection with the next and two ahead lattice points leads to a good identification result.