A dynamic detection and identification method for bridge structure defects of bridge cranes

Abstract

The invention discloses a method for dynamic detection and identification of bridge structure defects of bridge cranes. Set the material performance parameters of each component in the initial finite element model, set the boundary conditions of each component in the initial finite element model, conduct theoretical modal analysis of the bridge crane to be tested, and determine the dynamic test time of the bridge crane to be tested The cross-sectional position where the sensor is located, records the vibration parameters of the measuring point, conducts the operation mode analysis of the bridge crane to be tested, determines the actual operation mode parameters of the bridge crane to be tested, and corrects the original finite element model; The actual operating vibration parameters of the crane and the theoretical vibration parameters based on the modified finite element model.

Description

technical field [0001] The invention relates to a method for dynamic detection and identification of defects, in particular to a method for dynamic detection and identification of structural defects of bridge crane bridges. Background technique [0002] The existing non-destructive testing methods of cranes in non-working state mainly include radiographic testing (RT), ultrasonic testing (UT), magnetic particle testing (MT), penetrant testing (PT) and other methods, and the testing methods in working state mainly include acoustic emission testing. (AE). Radiographic testing (RT) is a method that uses a certain type of radiation to inspect internal defects in welds. When the ray passes through the weld to be inspected, the intensity of the ray falling on the film is different due to the different absorption ability of the weld defect to the ray, and the sensitivity of the film is also different, so that it can be accurate, reliable and non-destructive. Shows the shape, loca...

AI Technical Summary

Problems solved by technology

However, although the actual measurement of the structural frequency is accurate, it is not sensitive to local changes in the structure; the mode shape, especially the high-order mode shape, is sensitive to local stiffness changes, but it is difficult to measure accurately

The model correction method mainly uses the dynamic test data to continuously correct the stiffness distribution of the structural model through condition optimization constraints, so as to obtain the information on the change of the structural stiffness, and realize the damage identification and location of the structure. Incomplete modal set and insufficient test degrees of freedom can easily lead to non-uniqueness of the solution

According to the response of the structure in different states, the neural network selects parameters sensitive to structural damage as the input vector of the network through feature extraction, and the damage state of the structure as the output, and establishes a training sample set for damage classification. The main limitation of the artificial neural network method is that The accuracy of the acquisition of the training data set is largely limited by the completeness of the training data set

Images