Research On Structural Fatigue Crack Monitoring And Life Prediction Based On Lamb Waves

With the advancement of science and technology and the improvement of industrial level,the engineering structures are becoming more and more complicated,and fatigue cracks have become one of the main reasons for the failure of engineering structures.Therefore,it is of great significance to carry out research on structural fatigue crack monitoring and life prediction methods.This paper takes Q235 steel commonly found in engineering structures and mechanical equipment as the research object,then combines Lamb waves detection technology,fatigue crack growth analysis methods,and data-driven algorithms to carry out research on fatigue crack growth and life prediction method.The main contents of the paper are as follows:1.This paper reviews the research status of Lamb waves-based structural health monitoring,fatigue crack growth and life prediction at home and abroad.It focuses on the basic theory of Lamb waves detection,explains the Lamb waves modeling,the piezoelectric effectbased Lamb waves excitation,and the Lamb waves signal analysis method,and then builds a test system based on the Lamb waves detection technology.2.This paper studies the effectiveness and recognition ability of Lamb waves detection technology applied to structural crack detection.Taking H-beam as an example,the finite element simulation analysis model of Lamb waves detection is established using Abaqus software,and the propagation process of Lamb waves in the undamaged H-beam model and the damaged H-beam model with cracks are simulated,and effect of different length cracks on lamb waves is studied.Based on the defect echo method,the crack damage location in the H-shaped steel structure is studied,and the effective location of the 6 mm crack in the web is achieved through the verification of the H-shaped steel crack detection experiment.The relative positioning errors of the simulation results and the experimental results are 4.45% and-7.98%respectively,and the relative errors of the group velocity are-4.43% and 7.07% respectively.3.Based on the theory of nonlinear filtering,the fatigue crack growth state model and the crack observation model based on Lamb waves are constructed respectively.By discretizing the Paris formula,a noisy crack growth state model is established;the part of S0 mode wave packets in the Lamb waves crack monitoring signal are extracted to calculate the damage index DI,and a crack observation model is established.Then based on the fatigue crack growth experiment of the Q235 steel unilateral notch T1-T4 specimens based on the Lamb waves online monitoring,the Paris formula parameters m and log C of the structural crack growth are obtained,and the Lamb waves crack monitoring of the T1-T4 specimens are analyzed and used to calculate the Pearson distance,and the observation equation is obtained by parameter fitting.4.Aiming at the problem of the low prediction accuracy of the traditional Paris fatigue crack growth model and the inability to consider the influence of various uncertain factors in the crack growth process,a fatigue crack growth prediction method based on nonlinear predictive filtering(NPF)algorithm combined with the state model based on the Paris formula and the observation model based on Lamb waves is proposed.The results are verified by the unilateral fatigue crack growth experiment of Q235 steel.The experimental results show that the NPF algorithm can effectively correct the prediction error of Paris formula in the fatigue crack growth prediction,its prediction accuracy is better than the extend Kalman filter(EKF)and particle filter(PF)algorithm,and the algorithm efficiency is significantly better than the PF algorithm.5.Aiming at the problem of poor prediction accuracy caused by the lack of particle diversity of material parameters when using PF algorithm to predict fatigue crack growth and life,an improved particle filter(BAS-PF)algorithm based on BAS optimization is proposed for fatigue crack growth prediction and life prediction in this paper.It is verified by experiments that this method can combine the real-time observation information of Lamb waves to effectively improve the particle diversity of material parameters,and its crack growth prediction accuracy is higher than that of PF and NPF algorithms,and the remaining useful life(RUL)prediction ability is better than PF algorithm,which is more suitable for high-precision prediction of fatigue crack growth and life.