Debonding Detection Technology Of Honeycomb Sandwich Structure Based On Air-coupled Ultrasound

Honeycomb sandwich structure is widely used in aviation,aerospace and other fields due to its excellent characteristics such as high specific strength and low weight.However,the honeycomb sandwich structure composite material may debond between the honeycomb core and the skin due to improper operation and material defects during the manufacturing and use process,which poses a great threat to the safety of the overall structure.In order to prevent the structure from being damaged,effective It is necessary to detect defects and determine the location and size of defects.Aiming at the above problems,this paper uses air-coupled ultrasonic detection technology to detect the debonding of honeycomb sandwich structure.The research contents and achievements are as follows:Firstly,a two-dimensional model of honeycomb sandwich structure was established using COMSOL Multiphysics finite element simulation software,and the detection process of aircoupled ultrasonic waves was simulated to study the effect of different debonding defect sizes on the signal.The simulation results show that when there is debonding inside the material,the ultrasonic energy will be attenuated,so the received signal amplitude decreases,and as the debonding size increases,the signal amplitude decreases,so the received signal amplitude change can be used to detect the material debonding defects.Secondly,to address the problems of low signal-to-noise ratio of received signals and difficult identification of echo signals in actual detection,the combination of Variational Mode Decomposition(VMD)and Independent Components Analysis(ICA)algorithm,VMD-ICA algorithm is proposed to The results show that the echo signal in the signal can be clearly distinguished after noise reduction using VMD-ICA algorithm,indicating that the VMD-ICA algorithm can realize the signal noise reduction of air-coupled ultrasonic detection of honeycomb sandwich structure.Finally,for the long time of ultrasonic C-scanning,it is difficult to quickly judge the defect size.Using the signal classification technology of Convolutional Block Attention ModuleConvolutional Neural Networks(CBAM-CNN)based on the attention mechanism,the classification research of defect-free signals and 1mm?6mm defect signals is realized.The accuracy rate in the test set reaches 91.73%,and the classification of a single signal only takes200 ms,which has a faster speed and can provide a new solution for the judgment of defect size in ultrasonic inspection.