Research On Damage Identification Of Grid Structure Based On Random Vibration

With the variation of manufacturing and increase of production capacity,more and more equipment are used between grid structures of in service in industrial construction,especially in some grid structures of cement industry,like prehomogenization silo,clinker silo,and aerospace A380 hanger,there are a large number of hanging equipment and moving load.In addition to increase the dead load on the structure,hanging equipment operation will have random excitation on the structure continuously.Grid structure will produce random vibration under the random excitation,structure damage will occur in the long time,such as partial bolts looseness,rods bent and rods broken.If the key rod occurs damage,it is likely to lead to the collapse of the whole structure,resulting in construction accidents,threatening the safety of people’s lives and property.Therefore,it is imperative to be security measures to detect damage status of in service grid structure under random excitation regularly.It has important practical significance to prevent the happening of accident when finding damage status and putting forward damage warning timely.In this paper,according to the properties and characteristics of the cross-correlation function in random process,the principle of cross-correlation function is used to analyze time history response data of before and after the damage of grid structure,and MATLAB is used to calculate those data to gain cross-correlation function amplitude vector(CorV).Modeled on the definition of modal confidence criterion,cross correlation function amplitude vector assurance criterion(CVAC)is used to identify structural damage and damage degree.In order to make structural damage identification more convenient and fast,CorV is normalized,the K that changes of corresponding element in CorV of before and later structural damage is used as damage index to judge structural damage location.The same grid structure model is used to numerical simulation and experiment analysis in this paper.In the numerical simulation,in order to improve the accuracy of the damage detection,all members in grid structure model are distinguished upper chord,lower chord and web member to detect structural damage in different areas.First,three conditions that upper chord,lower chord and web member damage completely are set to verify the correctness of damage detection method based on cross-correlation function amplitude vector.Second,damage location is identified when damage degree of three kinds of bars is set as stiffness reduction 30%,50%,70% and 90%,different stiffness reduction values as X-axis and CVAC at different stiffness reduction of three bars as Y-axis are made liner fitting,quadratic curve fitting and cubic curve fitting,CVAC of stiffness reduction 60% verifies the quadratic curve fitting with the highest of alignment.Third,two conditions that two bars damage with common node and no common node are set,damage detection method based on cross-correlation function amplitude vector is used to identify damage location.In the experiment analysis,two pathway are laid along the longitudinal string node of grid structure,angle iron car moves on the pathway as random excitation effect on grid structure.Because of the limitation of experiment,just two conditions that one web member and two web members with common node damage completely are set.Although some interference factors influence the identification accuracy,location of damage bar is identified accurately.Through research on numerical simulation and experiment analysis,grid structure is divided into upper chord,lower chord and web member to detect classification,not only the accuracy of damage identification improved,but also cut down detection time.Research shows that CVAC is more sensitive to structural damage,and the quadratic curve fitting of different damage degree is the best representative,and change degree K of corresponding element in CorV can judge out structural damage location.