Large Structure Vibration Test Based On Wireless Sensor Networks And Applications

The service life of large-scale civil structures, such as bridges, high-rise buildings, spatial structures, offshore platforms etc, are being affected by operational loadings and surrounding environment. The coupled effects of material aging, long-term load effect, fatigue effect, mutation effect, environment erosion, lead to the damage accumulation and the resistance degradation. The structural capacity of resistance to natural disasters such as typhoons, earthquakes, and heavy rain will be reduced. Some of them in service come crashing down, inducing significant personnel and property loss. The structural damage changes the structural dynamic characteristics and modal parameters reflect the dynamic characteristics of the structure. Therefore, it has been a concern in civil engineering field to monitor the structural condition by observing the change of modal parameters, through measuring the structural responses subject to ambient excitation, and extracting out of the modal frequencies, mode shapes, damping etc.This dissertation takes the New Headquarters of Shenzhen Stock Exchange which has a huge cantilevered floating platform like a skirt as a testbed to conduct the ambient vibration test. Due to wiring difficulty, high cost and hard to protect, the traditional wiring system is not easy to be implemented in vibration tests of the large-scale structure. To meet the actual needs of the project, a large-scale integrated vibration test system based on wireless sensor networks (WSN) has been used as wireless nodes designed by Yang Wang and Kincho H.Law who work in the Stanford University.First, the dissertation introduces the integrated vibration test system based on WSN. The composed modules in the test system, including sensor system module, data acquisition module, wireless communication module, data display and analysis module are described. Second, two modal analysis methods, peak picking (PP) and frequency domain decomposition (FDD) are programmed by using MATLAB software. With NI PX11052 test system, trial tests in the laboratory with 8 sensors installed on a simple-supported beam model are carried out. The validity of the two algorithms for modal parameters identification is verified and modal frequencies and mode shapes are obtained. Then, simultaneous tests with multiple sensors are conducted at the testbed under study, and the common problems occurred in the field test are analyzed. Additionally, the vibration signal preprocessing methods commonly used in engineering, including the sample data calibration transformation method, data central processing method, and trend-term elimination method are introduced. The data acquired from the field test are analyzed. Finally, the modal parameters of the structure are extracted out using the modal parameter identification method, and the results are compared with the theoretical values. Test results show that the integrated vibration test system based WSN is effective for the ambient vibration test of large-scale structures, especially for the in-construction stage monitoring.The dissertation is ended by a summary of the work, and the recommendation for future works is also included.