Damage Detection In The Structure Subjected To Random Excitation

The Structural Health Monitoring (SHM) is a pervasive need in modern industry and life. The structural damage detection technology is the basis of SHM. Firstly, the damage detection methods are reviewed by being classified into three categories: structural modal parameters identification based method, data processing based method and non-comparative detection methods. Then it is pointed out that the method for damage detection on structures subjected to random excitation is worth to explore, and the research efforts of this dissertation mainly focus on this subject.The principles of structural damage detection are systematically described, which include the sensitivity of parameters used for damage detection, damage simulation technique, data processing method and data fusion technique for SHM. A necessary characteristic of the parameter for damage detecting is proposed as: the variation of the parameter with the extent of structural damage should be monotonous. The method to generate a stochastic excitation signal by curve fitting with random digitals is proposed.The sudden damage event detection techniques for the structure under stochastic excitation are investigated. Structural response with higher frequency may present due to the sudden decrease of local structural stiffness. Three methods are used to identify the structural response with higher frequency subjected to sudden damage occurrence: the empirical mode decomposition (EMD), high-pass filtering and response signal differentiating. The results of experiments show that the first two methods have higher detection accuracy and the third method is the simplest. A new intermittency check method in EMD method is proposed, by which the intermittency of signal is checked according to the number of points between the successive zero-crossings instead of the number of points between the successive extrema which is used in conventional intermittency check. A criterion for the precision controlling of EMD method is also proposed, which uses the ratio of the maximum mean of the upper and lower envelope to the maximum of the intrinsic mode function (IMF). This criterion is more convenient for application. A Windows application for EMD method is introduced, which can be used to process the data signal in sudden structural damage detection.The white noise excitation is generally indispensable to present structural modal parameter identification approaches. So in the damage detection for a structure based...