Research On Laws Of Connecting Structures’ Loosing Between Tower-crane Sections And Detection Technology

Tower-crane collapse accidents are very common during building construction.The damage to the connecting structures is often the direct cause of these accidents;tower-cranes produce severe vibrations during using,and the structures are subjected to frequent cyclic loads and impacts.The power of the connecting structures is gradually degraded under the influence of long-term vibrations,which is specifically manifested as loosening and falling of bolts.However,in the current maintenance works,manual inspection is costly and efficiency is unstable,and the performance of bolted connecting structures can’t be fully guaranteed,resulting in low level of safety management efficiency and high level of structural safety risk,which often bring hidden dangers and lead to collapse accidents.Based on the above factors,we could get that,studying the dynamic characteristics of tower-cranes and the law in degradating process of connecting structures,exploring the auxiliary or alternative manual monitoring methods to maintain the safety state of the tower-cranes during using,make a lot of sense.This thesis first analyzes the impact of vibration on the bolt-connecting structures in different parts.In this part,a FEA model of the T6013 flat-top tower-crane is established,and its accuracy is verified by analysis of both static stiffness and static strength.Then based on this model,simulation analysis was carried out on the three major operation processes,and the laws of dynamic responses were found,which explains the loosening of the bolt-connecting structures during tower-cranes’ using.Subsequently,the tracking survey data of the tower-crane in service of the project was used to verify the current situation of bolt-connecting structures’ frequent loosening;statistical analysis of the loosening location is performed with these data,its conclusion is consistent with what simulation analysis get.Then,based on the laws of dynamic responses,a method for dividing connecting nodes of the tower-crane and a “two-step method” for detecting the loosening of the boltconnecting structures are proposed.In this “two-step method”,fist,the connecting nodes are divided into several sub-regions,and the sub-regions where bolt loosening occurs are identified by the natural frequency change ratio of the structure.In order to take the true influence of the bolt loosening on the stiffness of the node into consideration,partial entity finite element analysis is used to obtain precise mechanical properties of bolt-connecting structures’ stiffness;based on this analysis,a model composed by several sections is established,96 sets of representative working condition samples are obtained,identifying the damaged sub-regions is realized through BP neural network which is trained and verified by those samples.Finally,in the second step,based on the partial strain modal amplitude change ratio,the bolt-connecting structures that are loosened in the target damage sub-regions are accurately located.In order to verify the feasibility of this method,an experiment is designed and carried out,the specific location of the damage is accurately located by comparing and analyzing the rate of change.