Monitoring And Numerical Simulation Of Looseness And Damage Of Bolt Group Joints Based On Piezoelectric Ceramics

In recent years,with the development of transportation industry vigorously and the expansion of the green environmental protection idea,steel structure bridges occupy an increasing proportion in bridge construction,the high-strength bolts connection as the structure of the connection mode is commonly used in steel bridge,during operation,it is impacted by adverse factors such as dynamic impact,fatigue load and harsh environment,As a result,the high-strength bolts are corroded,loose and fall off.In order to avoid the threat of such problems to bridge safety,it is very necessary to judge the loosening damage of high-strength bolts in time,therefore,this paper proposes the monitoring research on the loose damage of steel structure bolt group joints based on piezoelectric ceramics,and the specific research work is as follows:(1)Bolt loosening damage monitoring of high-strength bolted connection member of steel structure was studied based on piezoelectric wave method.Six piezoelectric ceramics are arranged on the test component to form a "three rows and two longitudinal" driving sensor monitoring network.The experimental results show that with the increase of the number of loose high-strength bolts,the absolute mean value of time domain signal and wavelet packet energy of the three groups of transverse driving sensing devices continue to decrease,which is reflected in the attenuation of 76.4%,87.2%and 84.2%respectively compared with the absolute mean value of time domain signal in healthy state,and the attenuation of wavelet packet energy is 99.9%,100%and 100%respectively,The absolute mean value of time-domain signal and wavelet packet energy of the two groups of longitudinal driving sensing devices are increasing,which is shown in that the absolute mean value of time-domain signal increases by 39.1%and 28.7%respectively,and the wavelet packet energy increases by 93.6%and 60.4%respectively;In the case of loosening the same number of high-strength bolts,the closer the damage position is to the driving sensing device,the greater the impact on the signal collected by the driving sensing device.(2)BP neural network and support vector machine(SVM)are used to identify the high-strength bolts loosening damage based on piezoelectric wave test.The effects of different algorithms,different damage indexes and sample sizes on the identification of loose damage of high-strength bolts in steel structures were studied.The results show that the recognition accuracy of BP neural network trained by wavelet packet node energy change rate damage index "small sample" and "large sample" for high-strength bolt loosening damage is 75%and 100%respectively,and the recognition accuracy of BP neural network and SVM trained by wavelet packet node energy entropy damage index "small sample" and "large sample" for high-strength bolt loosening damage is 92.7%,100%,100%and 100%respectively,It shows that wavelet packet node energy entropy as damage index can better reflect the loosening damage characteristics of high-strength bolts than wavelet packet node energy change rate;Under the condition of sufficient training samples,BP neural network and SVM can better identify the loosening damage of high-strength bolts,but SVM has more advantages in the case of "small samples".(3)The finite element analysis software ABAQUS is used to establish the numerical model of high-strength bolted connection member of steel structure,and the finite element analysis is carried out under three kinds of high-strength bolts loosening damage conditions(bolt loosening damage in turn,single bolt loosening damage and double bolt loosening damage),The results show that the more serious the interface damage of high-strength bolt connection is,the greater the attenuation of time-domain signal amplitude of transverse driving sensing device is,and the more obvious the increase of time-domain signal amplitude of longitudinal driving sensing device is,which is in good agreement with the experimental results,which verifies the accuracy and effectiveness of the numerical simulation results.For the double bolt loosening damage monitoring,when the loosening damage positions of two high-strength bolts are adjacent and on the linear propagation path of the transverse drive sensing device,the attenuation effect of the corresponding stress wave signal is the greatest,that is,when loosening 1#,3#bolts and loosening 3#,4#bolts,the absolute amplitude attenuation of the time domain signal received by the transverse drive sensing device"pzt3-pzt4" is the largest,and the attenuation degree is 52.6%and 44%respectively.