Study On Damage Detection Of Space Truss Structure With Random Defects

With the advantages of light weight, reasonable structure, graceful appearance and better mechanical performance and so on, the large span truss structure is widely used in large scale public building and civilian installations. Actual structure has some differences from the ideal model because of some initial random defects which change geometric characteristics, damage evolution mechanism and bearing capacity of the ideal model. So the structural damages were emerged more easily. On the other hand, with working time increasing, structure will appear injury inevitably bearing overload, environment change and damage accumulation. According to these problems, damage detection of grid structure with random defects were analyzed as follow:1) According to modal characteristics of grid structure and advantages of wavelet translation in signal processing, a method of damage detection based on curvature mode difference and wavelet transformation was used to determine the damage location of gird structure. Through analysis, link element damages of four common gird structures including square pyramid space grid structure, orthogonal space truss structure, three way latticed grid and honeycomb triangular pyramid grid structure were simulated, damage locations can be determined more accurately by using this method.2) On the assumption that the statistical features of structural random defects obey normal distribution, sensitive parameters of structural response can be determined by PDS technology in ANSYS and modal analysis. Four common grid structures are taken for example, grid structures with random defects can be made up by MATLAB and ANSYS.3) These structures with random defects will be considered as the initial structure. Same damage testing method and damages are applied, these structures with random defects and ideal models were analyzed and contrast. The comparison shows that random defects have some affects to the damage detection of girds structures and damage locations of grid structure with random defects can be detected by this method in a certain range of variation coefficient.