Study On Ultrasonic Near-surface Defect Detection Method Of Metal Alloys

As one of the most active research methods in industrial non-destructive testing,ultrasonic testing technology is widely used in the defect detection of metal alloy materials,which can effectively judge the performance and life of materials,and ensure the safety and reliability of equipment.The existence of internal defects can easily cause damage and aging of metal materials,especially those located near the surface of the material.Due to the existence of surface waves,there is a blind spot for detection of near-surface defects,which makes it impossible to identify and locate defects.Therefore,based on the theory of ultrasonic testing,this paper studies the ultrasonic testing method of near surface defects of metal alloy materials.Aiming at the problem that the near-surface defect echo of metal alloy is submerged in the interface echo,which makes it difficult to extract the near-surface defect features of the material.Useing the different characteristics of the bottom echo,a method based on Hilbert spectrum for identifying near-surface defects in bearing inner ring is proposed.The near-surface defect signals of 440 C stainless steel bearing inner ring were compared in time-frequency domain based on Hilbert-Huang transform and wavelet transform.Through the convolutional neural network classification,cross-experiment was carried out to verify the effectiveness of the method proposed in this paper for the near-surface defect identification of the bearing inner ring.The experimental results show that the Hilbert spectrum has more advantages than the wavelet time spectrum,and it can classify the small defects on the upper and lower surfaces at the same time.The method effectively improves the detection accuracy of the near-surface defects.An average accuracy of recognition of 98.83% is achieved.The study provides a new method for the recognition of near-surface defects.Based on the amplitude of the defect echo is smaller than interface echo,the arrival time and amplitude of the defect echo can?t be directly identified,so that the depth of the near-surface defect can?t be correctly judged.In this paper,a method based on mathematical morphological filtering and EEMD for localization detection of near-surface defects in metal alloy is proposed.After mathematical morphological filtering and EEMD decomposition of the echo signal,the components with better correlation are selected to reconstruct the signal.Then calculate the difference between the maximum local echo amplitude of the defect signal and the reference signal,if(35)> 0,take the position corresponding to the maximum value as the defect position.The experimental verification is carried out using the near-surface defect signals of the inner ring of 440 C stainless steel bearings and 304 stainless steel.The experimental results show that the signal processing method combining mathematical morphological filtering and EEMD can accurately detect the position of near-surface defects,and the detection error is less than 0.05 mm,which provides a reference for the localization and detection of near-surface defects.