Defect Detection Of Spinning Cylinder And Blind Feature Extraction Technology

Metal spinning technology is widely used in aviation,aerospace and other fields.The spinning cylinder made by this technology is an important component of large and heavy carrier rocket.Spinning cylinder workpiece in the process of spinning forming,storage,transportation is easy to produce defects,resulting in the overall mechanical properties and strength of the workpiece decline.Therefore,in order to ensure the integrity and reliability of the workpiece,Non-destructive testing should be carried out during the production process or before the use of the spinning cylinder.In this paper,the ultrasonic detection technology is used to carry out the research of spinning cylinder defect detection method and feature signal blind extraction technology.(1)Aiming at the problem that crack defects at arbitrary angles of spinning cylinder are not easy to be detected,this paper firstly uses CIVA simulation software to carry out numerical simulation of cylinder detection,reveals the mapping law between the probe orientation angle and the crack angle,and designs the detection probe distribution and scanning path scheme.Secondly,the detection effect of three different probe distribution schemes is analyzed.The feasibility of the optimal scheme is verified by the simulation scanning experiment,and the scanning detection of the cracks at different angles of the spinning cylinder is completed.(2)In order to solve the problems of aliasing and scanning imaging quality caused by spinning texture,a blind feature extraction algorithm of spinning cylinder defect based on Variational Mode Decomposition-Fast Independent Component Analysis(VMD-Fast ICA)and frequency domain correlation coefficient is proposed.Firstly,multi-channel observation signals are obtained by signal reconstruction algorithm based on VMD,which solves the problem of under-determined blind source separation of ultrasonic aliasing signal separation.Secondly,Fast ICA algorithm is used for blind source separation of observation signals,and the linear aggregation of each independent component in frequency domain is judged by calculating the correlation coefficient between the independent component and the reference signal frequency domain.The optimal decomposition of aliasing signal is realized.The wavelet threshold algorithm is used to further process the noisy signal,extract the feature information of the defect,and restore the actual location of the defect on the imaging map by using the position correction technology.The results show that the distributed scanning technique of probes and blind extraction of characteristic signals proposed in this paper can effectively realize the identification of spinning cylinder defects.The research in this paper provides some ideas for the detection of spinning cylinder defects and ultrasonic signal processing.