Loading...

Elliptical Crack Detection in Steel Beam Using Lamb Wave and SVM Learning Model

Soheilzad, Ensieh | 2024

77 Viewed
  1. Type of Document: M.Sc. Thesis
  2. Language: Farsi
  3. Document No: 58425 (45)
  4. University: Sharif University of Technology
  5. Department: Aerospace Engineering
  6. Advisor(s): Hosseini Kordkheili, Ali
  7. Abstract:
  8. The use of Lamb waves as an effective method for structural health monitoring (SHM) and non-destructive testing (NDT) has attracted significant attention. The application of machine learning algorithms reduces complexity and minimizes errors in the use of Lamb waves. Elliptical cracks are among the common types of defects, and their detection plays a crucial role in ensuring structural integrity. In previous studies, there has been no specific focus on identifying elliptical cracks in terms of their size, orientation, and location. The proposed strategy for identifying elliptical cracks in this research consists of three main stages :Data generation, Lamb wave signal processing and feature extraction, and The use of learning models to predict the size, orientation, and location of the elliptical crack. Various types of elliptical cracks with different sizes, orientations, and locations were generated using the finite element method (FEM). Lamb wave signals were processed in the time and time–frequency domains using the RMS and discrete wavelet transform (DWT) energy coefficients. Subsequently, a Support Vector Machine (SVM) model was used to identify the size and orientation of the elliptical cracks, while a regression model was employed to predict their location. The proposed method offers advantages such as high prediction accuracy—particularly in determining the size and orientation of elliptical cracks—as well as low computational cost
  9. Keywords:
  10. Lamb Wave ; Elliptical Crack ; Structural Health Monitoring ; Support Vector Machine (SVM) ; Regression Analysis

 Digital Object List

 Bookmark

No TOC