Preliminary Study On Steel Truss Bridge Health Monitoring

Structural health monitoring is an invariable issue in the field of structural engineering. For damage diagnosis, the task is to determine the possible damage locations and to estimate the degree of damage. Methods of signal processing such as adding window function, digital filtering, decreasing noise, FFT spectrum analysis, power spectrum analysis and transfer function analysis are playing an important role in structural health monitoring.According to the theory in structural dynamics and advanced technique of tests, structural damage cause variety of structural behavior, at the same time, the model parameters make differences. The foundational task is analysis and processing the data of experiment research. Furthermore it could obtain the evaluated index of the health monitoring.In this paper, a survey is given for structural health monitoring. Several key issues related to signal processing is summarized. The concepts, theories and scientific application are presented . Computer program is done from MATLAB to carry out signal processing. Experimental study for damage detection of a steel truss model is the central part in this paper. Cracking depth is different and the location is changeless. It is classified according to four cases. The process involves the observation of featured parameter that can reliably provide the states of the damaged degree, the method to efficiently assess the health state. After cooking the data, it comes to the conclusion that natural frequency can be served as the evaluated index, the analysis of transfer function is an more efficient method that compared with FFT spectrum analysis and the time-history method, it is easy to identify the damage of symmetrical structure.