Abstract

Info

Author(s):: WU Duo; LIU Lai-jun; ZHANG Xiao-yu; LIU Hong-yan; National Engineering Laboratory for Bridge Structure Safety Technology, Chang’an University

Keywords:: bridge engineering; damage identification; curvaturemode curve; data fitting; neural network

Abstract:: Aiming at the problem of traditional damage identification method which could only determine the damage position, and the damage degree recognition was poor, a damage identification method based on the change of curvature-mode was proposed in the light of the distortion of curvature-mode curve caused by bridge damage. Based on the parameter index of curvature mode, the change of curvature-mode curve of bridge before and after damage was studied. Based on polynomial fitting and neural network fitting technology, according to the size of curvature-mode curve distortion area after bridge damage, the location and degree of damage were reversely fitted. Taking a simply supported bridge as an example, the single and multi location damage conditions were studied and analyzed, and the location of structural damage was determined according to the location of curvature-mode curve distortion. The damage degree was determined by the size of curvature-mode curve distortion. The results show that the proposed method is more effective in identifying the small damage condition which is commonly seen in the actual engineering, and it can be used in the health monitoring of engineering structure.