Study On Dynamic Modeling And Boundary Identification Of Axial Compression Column Under Different Pressures

The reliability of major civil engineering structures has important effects on society and economy. The exact assessment of structural actual behavior is the basic premise of structural reliable work. For engineering structures, structural dynamic response measurement is easy to implement . By using modal parameters or physical parameters which can be identified with structural dynamic responses, we can establish structural models to evaluate structural performances. In the beginning of this paper, the first three natural frequencies of an axial compression column elastically restrained against rotation at both ends were identified with an improved Hilbert-Huang transformation (HHT) method based on peak-frequency bandpass signals. Then, dynamic equations of the column were established through the analytical method and finite element method. And the rotational stiffnesses of boundaries which change with the variation of pressures were determined by natural frequencies obtained from experimental modal analysis. This thesis focuses on the following aspects of research:(1) Theories of modal parameters identification based on HHT were studied in-depth. A new method of modal parameters identification was proposed in the paper, which was based on HHT method to peak-frequency bandpass signals. And needed analytical and computational programs were generated based on MATLAB. At first, power spectra obtained from Fourier transformation lay out the distributions of modal frequencies of the pulse excitation response signals. And then, bandpass response signals containing peak frequency were obtained by Butterworth bandpass filter. Finally, modal parameters were identificated through HHT method based on the bandpass response signals.(2) In experimental environment, experimental modal analysis was studied on the axial I-beam column elastically restrained against rotation at both ends. Based on pulse acceleration response signals of three points, the first three natural frequencies of the column were identified using the method provided in the paper. The experimental results demonstrated that the mode mixing phenomenon and the endpoint effection in HHT were successfully solved by the peak-frequency bandpass filtering technical and the removing endpoints method. And they improved the reliability of algorithm and the accuracy of modal parameters identification.(3) The transverse vibration frequency equation of an axial compression column elastically restrained against rotation at both ends was derived from the energy principles. And the frequency equations of two extreme boundary condition (simply supported and fixed) columns were obtained from it. According to the frequency equation, theβ?ηcurves of the frequency coefficientβand the rotational stiffness coefficientηunder different pressures were drawn. Then, the effects on natural frequencies of some factors, including boundary conditions, axial pressures, the shear deformation and the moment of inertia, were studied and some related laws were obtained. The theory of how to use natural frequencies to determine the boundary conditions of the column was described in detail. The boundary conditions of the test column were identified and verified by this theory, and finally it achieved good results.(4) ANSYS finite element models of axial columns with two extreme boundaries (simply supported and fixed) and boundaries elastically restrained against rotation were established. And boundary conditions which were identified with the analytical model were verified by the latter one.