Location And Quantitative Identification Method For Bolt Looseness Disease Of Steel Truss Bridge

The development of heavy-haul and high-speed railways has put forward higher requirements for the statue of bridges.As the main type of the railway bridge,the steel truss bridge has the advantages of light weight and convenient construction.Bolt looseness is a common disease of steel truss bridges.A large number of bolts looseness will seriously endanger the safety of steel truss bridges.Under this background,it is necessary to study the method for identifying bolt looseness disease applied to steel truss bridges.Through the comparison of the existing bolt issues detection method,a new method that could be used to locate and quantitatively identify the bolt looseness disease of steel truss bridge is proposed,based on model updating and probabilistic neural network.The bolt looseness disease of the main truss member of the steel truss bridge is studied by numerical simulation.The main research contents and results are as follows:(1)A simplified simulation method for bolt looseness disease of steel truss bridges is proposed.Through the calculation and analysis of the joints connected by bolts,the relationship between the bolt looseness disease and the rotational stiffness of the joint is obtained,and it is proposed to simplify the simulation of bolt looseness disease by reducing the bending stiffness of the element at the rod end.(2)A location identification method for bolt looseness disease of the steel truss bridge based on probabilistic neural network is proposed.According to the geometric position of the bar,the steel truss bridge is divided into four substructures.Through parameter sensitivity analysis,the curvature modal change rate is determined as the input parameter of the neural network.The probabilistic neural network is used to locate the member with bolt looseness disease into the substructure.Numerical simulations prove that the accuracy of location identification of this method is above 90%.(3)A quantitative identification method for bolt looseness disease of the steel truss bridge based on model updating is proposed.On the basis of locating the disease to the substructure,the objective function is constructed in curvature mode,and the model updating method is used to identify the bolt looseness disease degree of each member in the substructure.Numerical simulations show that the quantitative identification can effectively quantify single disease and multiple disease cases.(4)Using the curvature modal parameters,a test method for identifying bolt looseness disease based on the free vibration method is proposed.By adding white noise to the free vibration test signal,the influence of noise on the identification results of bolt looseness disease is analyzed.The analysis results show that this method can effectively identify the bolt looseness disease when the signal to noise ratio of the test signal is higher than 40 dB.