Research On Bridge Damage Detection Based On Moving Load Induced Structural Response And Wavelet Analysis

This research is sponsored by the National Key Basic Research Development Program: Project973(Grand No.2013CB036203) and the National Natural Science Foundation of China (Grand No.51078029).It is necessary to study the dynamic characteristics of damaged bridge and to identify the damage in time, since structure damage may accumulate during the service which will influence the service life and and vehicle running safety. There are various disadvantages among traditional non-destructive detecting technologies, some of which can not obtain the accurate test data or need to interrupt the traffic and so on. It has become a hot issue to find a method to identify bridge damage quickly and effectively.In this paper, the dynamic characteristics of damaged bridge are studied and the damage is identified with several methods based on Wavelet Analysis. The main work and the results are as follows:(1) Various commonly-used damage models are presented and compared. The dynamic characteristics of damaged beams such as frequencies and mode shapes are studied with the damage modeled as a rotational spring. The dynamic responses of damaged beam subjected to a moving force and a moving mass are solved separately, and the influence of parameters such as load velocity, load magnitude is analyzed. Besides, the loaded frequencies of simply-supported beam based on moving mass as well as wheel-set and sprung mass are derived.(2) On basis of the Wavelet Theory, bridge damage identification method based on structural mode and Continuous Wavelet Transform (CWT) is studied. The structure displacement mode is taken as input. With WT applied, damage location can be identified by wavelet coefficient contour plot and maximum modulus locus while damage degree can be evaluated by Lipschitz exponent. Numerical examples show that this method can identify the damage effectively. Furthermore, some influence factors are studied such as Euler and Timoshenko beam, damage extents, multiple damage, sparse measure points and test noise.(3) Using the Reciprocal Theorem of Displacement and Influence Line Theory, the singularity of bridge displacement time history when the moving load passes through the damage location is proved. On this basis, bridge damage identification from moving load induced deflection using a combined wavelet transform and Lipschitz exponent method is proposed. In this method, displacement time history of some measuring point is taken as input and the Continuous Wavelet Transform is applied. Damage location can be identified by wavelet coefficient contour plot and maximum modulus locus while the damage degree can be evaluated by Lipschitz exponent. A numerical example is taken to study the influence of different damage degrees, multiple damage, different sensor locations, load velocity and load magnitude. Besides, the feasibility of this method is verified by a model experiment.(4) Applying the Curvature Modal Difference (CMD) Method to spatial structural damage identification, the suspender damage of a tied-arch bridge is identified. In this method, the displacement mode of the nodes where the suspenders join with the deck are taken as the input, and the damage of suspenders can be located by the CMDs of relative nodes. Moreover the influence of noise is considered.(5) The Wavelet Packet Energy Curvature Difference (WPECD) Method is proposed on the basis of curvature and wavelet packet analysis. The impact vibration responses of the concerned nodes are taken as input and the Wavelet Packet Transform (WPT) is applied to obtain the WPECDs as the damage index. The influence of sparce measuring points is studied. A field test is taken to verify the effectiveness of this method, and different wavelet functions as well as decomopisition levels are also compared.