Parameter Identification And Damage Diagnosis Research On Reinforced Concrete Bending Members

The purpose of this dissertation is to solve the problems associated with dynamic performance of the system and its application. This study can be summarized as two parts:system simulation and system tests. The former part is mainly concerned with theoretical research, such as the dynamic model of the system, the establishment and the solution of the equation of motion. The latter part is mainly concerned with the experimentation, such as the test of structural dynamic characteristics, the system identification of structures. Two aspects are all discussed in the dissertation. Such as: the principle of digital signal processing (DSP) and the methods of parameters identification. Dynamic tests were carried out on reinforced concrete (RC) flexural members. According to the dynamic parameters of the members obtained from measured vibration responses, two aspects of the laws, as dynamic testing technique and structural dynamic performance, were summed up. The main works of the dissertation are presented as follows:The loading-unloading performances were tested in detail on RC flexural member. The curves were gotten that described the changed trend of flexural stiffness with increasing load. The main features of flexural stiffness function were speculated as basic facts to support theoretical research and practical application. Firstly, as long as the current load is not greater than the maximum value of load that the member has been subjected to, load-deflection relationship of bending member will be in line with linear characteristics. While to identify the structural dynamic parameters based on dynamic measured data, it should be aware that non-linear of the load-deflection curve is very weak, the other factors will be considered in the case of conflict between the assumed structure model and the test data. Secondly, the area where the stiffness reduces due to cracking cross-section of member also tend to be smaller.The discrete Fourier transform (DFT) was deduced from the Fourier series expression. And then the physical significance of DFT was explained clearly and completely. When the signal spectrum is calculated using the type of DFT expression, the spectral coefficients and the frequency domain coordinates need be adjusted correspondingly only. The symmetrical property of the spectrum was described in principle. Key spectral lines would not be missed generally in according to the general number N/2be considered as effective spectral lines. But the understanding of the related issues is likely to be affected. Such as the quantity relations might be confused in the process of calculation in frequency domain. Base on the discussion of DFT, the improved method that would visually extract the frequency as better peak of the spectrum, zoom spectrum analysis, was further discussed here. While it was discovered that a single frequency harmonic zoom spectrum and a rectangle window spectrum was very similar. Based on this law, the digital signal included several close frequency harmonics can be identified. Furthermore, Harmonic parameters can be extracted from the most close frequency digital signal.Based on the identification of frequency and the initial phase using zoom spectral analysis and the important properties of the DFT. Discrete convolution and deconvolution principle, a new method was proposed for identifying the damping from free decay response (FDR) without dependency on a prior knowledge of damping model. The time sequence can be identified by discrete deconvolution technique which can be employed to reflect the changes of harmonic amplitude. The method is an excellent tool for the first estimation attempts. However, frequency resolution, leakage, coherence dropout at resonance, and the other factors make the peak amplitude susceptible to error and do not calculate the most accurate data. In that case, in according to the first results the appropriate parameter range and step size were selected for loop calculation. Next, in turn, these parameters would be substituted into the specified theoretical signal expression, the optimal results were determined on the principle to make spectrum curve of theoretical signal fit best with that of the experimental signal. As the original expectation, it is implemented that system parameters are identified using output data only.Dynamic tests on RC frame, RC simply supported beam and composite beam have been performed to identify the system parameters using these measurements. The influences of test conditions were discussed. And the laws of dynamic testing techniques and the structural dynamic performance were summed up. When a structure is done by a damage diagnosis with the modal method, it can be listed below that Modal parameters change the rules:the frequencies will reduce while the damage of structure increases, meanwhile exponential decay coefficients is on the trend of increasing. The two parameters all have a larger trend in the high frequency harmonic. The frequency and damping are likely to be useful for detecting damage in RC structure. However modal shapes only can be used to determine the repetitive mode.