Ultrasonic Signal Processing for Non-Destructive Testing and Evaluation

• Author / Creator

  Hoseini, Seied Mohammad Reza

• In this thesis, ultrasonic B-scan signal processing is investigated. A B-scan displays a cross-sectional view of the test piece and defects within it. Compared to A-scan signals, B-scan images offer more reliable fault detection. Nonetheless, traditional ultrasonic signal processing methods are mostly based on A-scan signals and little research has been reported on ultrasonic B-scan signal processing. The aim of the current thesis is to investigate different aspects of ultrasonic B-scan signal processing including denoising, parameter estimation and fault identification.

  The stationary wavelet transform (SWT) is used to denoise B-scan signals. SWT exhibits a good denoising performance. A thresholding scheme based on the amplitude of the analytic signal was reported to further improve the denoising performance of the stationary wavelet transform. In this work, the application of the analytic wavelet thresholding is extended to two dimensional signals for removing noise from B-scans. In addition, an extra step is proposed for removing noises caused by the waves reflected at the wedge-specimen interface.

  The ultrasonic signal is further processed to extract fault related features. A model-based method is proposed for estimating parameters of ultrasonic echoes such as the time of arrival of echoes. Despite many advantages such as excellent estimation accuracy, the current model-based methods applied to ultrasonic signals suffer from a major disadvantage. These methods often involve solving an optimization problem with many parameters. Some researchers calculated the envelope of the ultrasonic echo to reduce the number of parameters by removing phase and frequency from the parameter set. A quasi maximum likelihood estimator is proposed for estimating parameters of the echo envelope. Using experimental and simulated signals, it is shown that the proposed method improves the parameter estimation compared to the state of the art available in the literature.

  Current crack sizing techniques often neglect the effect of the crack orientation when estimating the crack length. This introduces errors in estimating the length of inclined cracks. A modified relative arrival time technique is proposed for estimating the crack length and inclination angle using the relative time of arrival of the echo diffracted from the crack tip with respect to the echo reflected from the crack corner.

• Subjects / Keywords

  • Signal Processing
  • Non-Destructive Testing and Evaluation
  • Analytic Wavelet Thresholding
  • Time of Flight Estimation
  • Ultrasonic Testing

• Graduation date

  Spring 2013

• Type of Item

  Thesis

• Degree

  Doctor of Philosophy

• DOI

  https://doi.org/10.7939/R3T36M

• License

  This thesis is made available by the University of Alberta Libraries with permission of the copyright owner solely for non-commercial purposes. This thesis, or any portion thereof, may not otherwise be copied or reproduced without the written consent of the copyright owner, except to the extent permitted by Canadian copyright law.