No abstract

The paper presents the principle and construction of computer program on stability analysis of multi-bay and multi-story plane frame by finite element method. Research on influence of the parameters: ratio of bending stiffness of beams and columns, ratio of length of beams and columns, and number of element per every column on the critical load parameters corresponding to symmetric buckling modes and antisymmetric buckling modes. The results show two-finite-element model per every column is sufficiently accurate for the stability analysis of multi-bay and multi-story plane frame.

Modal identification of structures using vibration measurement data has attracted considerable attention. Current identification methods can be divided into several groups: single-degree-of-freedom and multi-degrees-of-freedom methods; time domain identification, frequency domain identification. In which, the simultaneous determination of many natural frequencies of the structure from a single measurement data remains a complicated problem, in particular in the presence of noise. The article develops the rational fraction polynomial method using combined measurements in the frequency domain to identify the natural frequencies of structures. The effectiveness of the proposed method is validated by numerical simulations and experimental tests of steel beam.

This article presents identification methods of the natural frequency of structures from single or multiple measurements in the frequency domain and its application to identify the natural frequencies of the offshore truss structure. The simulated measurements of structural vibrations in the time domain are determined by analyzing the finite element model of the offshore truss structure subjected to random dynamic loads using SAP2000 software. The frequency domain measurements are obtained by the fast Fourier transform of time domain measurements using MATLAB software. The identification method from the multiple combined frequency response function measurements can estimate more natural frequencies than using a single measurement. A study of the considered frequency range was also performed showing the accuracy of the identification methods compared with the finite element model. By subdividing the considered frequency range, it is possible to improve the accuracy of the obtained natural frequencies.