Nodal Stiffness Prediction Method Based on Vibration Test

Abstract

The invention relates to a method for node rigidity prediction based on vibration measurement. The method is characterized by comprising the steps of first establishing a node model (a finite element model or a test model), utilizing the model to enable a first principal component score extreme value obtained through generalized likelihood ratio principal component analysis to serve as a control index, and establishing a relation curve of the proportion of the control index exceeding the upper limit and the lower limit and node rigidity under different damage working conditions in a statistical quality control chart; in actual engineering, measuring vibration response of a node, and calculating the proportion of the control index exceeding the upper limit and the lower limit in the statistical quality control chart according to the response; and finally substituting the proportion of the control index exceeding the upper limit and the lower limit into the relation curve to predict the node rigidity. According to the method, the node rigidity can be rapidly and effectively predicted without needing other data, only acceleration response of a structure needs collecting, and the method can be used for structural ageing and after-calamity rapid prediction.

Description

technical field [0001] The invention relates to a technology for predicting node stiffness, in particular to a technology for predicting node stiffness during structure use, that is, a node stiffness prediction method based on vibration testing. Background technique [0002] In order to avoid vicious collapse of bridge piers, improve their dynamic characteristics, and reduce long-term comprehensive maintenance costs, bridge damage identification and assessment based on dynamic tests has become a research hotspot. However, the current structural damage diagnosis methods based on vibration testing have their own limitations. Since only a few first-order frequencies and mode shapes of the structure can be obtained in actual test conditions, it is difficult to directly use frequencies or mode shapes for damage identification. less precise. For this reason, many scholars seek a change in damage-sensitive indicators to diagnose damage. That is, the current state is mainly compar...

AI Technical Summary

Problems solved by technology

At the beginning of this century, some scholars proposed that the structural damage identification technology using statistical knowledge has been developed, but it still has the following problems: (1) The current index mainly considered by using statistical knowledge is the acceleration itself, but the acceleration itself is interfered by the environment Unreliable indicators for large problems; (2) Structural damage techniques using statistical knowledge cannot be combined with structural stiffness

Images