• CN: 11-2187/TH
  • ISSN: 0577-6686

Journal of Mechanical Engineering ›› 2021, Vol. 57 ›› Issue (14): 177-187.doi: 10.3901/JME.2021.14.177

Previous Articles     Next Articles

PCA-WHMM-based Frequency-domain Synthetic Aperture Focusing Imaging Using Ultrasonic Leaky Rayleigh Waves

HU Hongwei1,2, ZHOU Gang1, SHEN Xiaowei1, XU Xiaoqiang1,2   

  1. 1. College of Automotive and Mechanical Engineering, Changsha University of Science and Technology, Changsha 410114;
    2. Hunan Provincial Key Laboratory of Intelligent Manufacturing Technology forHigh-Performance Mechanical Equipment, Changsha 410114
  • Received:2020-08-25 Revised:2021-03-05 Online:2021-07-20 Published:2021-09-15

Abstract: Ultrasonic leaky Rayleigh waves can be used to detect surface and sub-surface defects. The advantage of non-contact detection makes it easy to realize automatic inspection. However, due to the waveform conversion and propagation attenuation, the echo amplitude of leaky Rayleigh waves is quite small, which is not conducive for defect detection and imaging. It simulates and analyzes the propagation characteristics and defect echo characteristics of the leaky Rayleigh waves, applies principal component analysis (PCA) to separate the interference waves from the echo signal, then uses the Wavelet-based hidden Markov models (WHMM) algorithm to separate the systematic noise from the whole signal, combines the two methods to extract the defect information in the leaky Rayleigh waves signal, and finally uses the frequency-domain synthetic aperture focusing technology(F-SAFT) to conduct a high resolution ratio reconstruction for the data of the leaky Rayleigh waves scanning. The results show that compared with the conventional B-scan imaging, the PCA-WHMM-based ultrasonic leaky Rayleigh waves F-SAFT imaging method improves the SNR of the echo signal by 10.05 dB and reduces the average imaging error by 26.3%,which provides an effective method for the detection of metal surface and sub-surface defects.

Key words: ultrasonic nondestructive testing, leaky Rayleigh waves, principal component analysis, wavelet-based hidden Markov models, frequency-domain synthetic aperture focusing

CLC Number: