Ultrasonic Nondestructive Evaluation Of Debonding Damage Of Honeycomb Plate

Honeycomb plate is a kind of composite structure which is bonded by the upper and lower skin and the middle honeycomb core.It has the advantages of high specific strength,high specific stiffness and high energy absorption ability,and has good impact resistance and shock absorption performance.Honeycomb plates are widely used in the infrastructure,aerospace automotive industry and other fields.Due to the manufacturing process,service life and repeated loading,debonding damage will occur between the honeycomb core and the skin,which endangers the safety of the structure.In order to avoid the structural damage and failure,it is necessary to identify the damage location and damage degree effectively in the early stage of damage.Ultrasonic guided wave is widely used in structural non-destructive evaluation because of its advantages of long propagation distance,fast detection speed and sensitivity to structural internal damage.Therefore,the research on nondestructive evaluation of debonding damage of honeycomb plate by ultrasonic guided wave is of great academic significance and application value.In this paper,based on the characteristic frequency method,the actual cell model of honeycomb plate is established by using finite element software COMSOL Multiphysics,and the group velocity dispersion curve of guided wave is plotted.The three-dimensional model of the honeycomb plate is then established by using the finite element software,COMSOL Multiphysics,and the propagation of the guided wave in the presence or absence of the debonding damage is simulated.A three-dimensional model of honeycomb plate is established,and the propagation law of guided wave in the presence and absence of debonding damage is simulated.The influence of debonding damage on guided wave propagation is studied.Furthermore,the influence of the debonding size on guided wave propagation is studied by establishing a honeycomb plate model with different debonding damage areas.Finally,the experimental study of debonding damage in honeycomb plate is carried out by using the Polytec laser vibrometer test system.In the experiment research,by introducing the damage index(Damage Index,DI),using the artificial neural network algorithm,the simulated data are trained and applied to the prediction of the experimental results,so as to realize the location and quantitative identification of the debonding damage.The results show that:1.Based on the characteristic frequency method,the actual cell model of honeycomb plate can be established by finite element software,and the group velocity dispersion curve of guided wave in honeycomb plate can be obtained by calculating.2.When the guided wave propagates in the well-bonded honeycomb plate,some energy leaks into the honeycomb core,resulting in the leaky wave.When the guided wave propagates in the honeycomb plate with debonding damage,the amplitude of the guided wave in the debonding position increases,and the amplitude of the AO mode increases most obviously.According to the characteristics of the shortest wavelength of AO mode and the most sensitive to small size debonding damage,ultrasonic nondestructive testing and evaluation of debonding damage can be carried out by using the change of AO modal amplitude.3.With the increase of debonding area,the amplitude of AO mode increases significantly,and the damage index also increases.Therefore,the damage index can be introduced as a qualitative index to judge the degree of debonding damage.The damage index DI is more sensitive and stable than the amplitude of AO mode in characterizing the degree of debonding damage.4.The propagation signal of guided wave in honeycomb plate is collected by Polytec laser vibration test system.The results show that the amplitude of AO mode and the damage index DI increase with the increase of debonding area,and the damage index DI is more obvious.The characterization of debonding damage is more stable and accurate.The artificial neural network is trained by the data of finite element simulation.The trained neural network algorithm is applied to the prediction of experimental results,which can effectively identify the location of debonding damage and evaluate the degree of debonding damage.The accuracy of the neural network algorithm is equal to that of the damage probability algorithm.Increasing the number of training samples can improve the accuracy of the neural network algorithm.The neural network algorithm is helpful to the development of intelligent nondestructive testing in the future.