Fourier spectral–based model curvature analysis and its application to damage detection in beams

Zhi-bo Yang , M. Radzieński , Paweł Kudela , Wiesław Ostachowicz


In this paper, a simple Fourier spectral–based method is proposed to calculate the modal curvature (MC) of beams instead of the traditional central difference method. Based on the present method, damages in beam–like structures are localized. The present method provides an alternative selection to estimate MC in damage detection. There are two advantages of the present method. Firstly, the spectral calculation of spatial derivatives is conducted globally, which provides the suppression for noise. In addition, signal processing in the wavenumber domain provides an alternative choice for spatial filtering for mode shapes. Secondly, the proposed method provides a precise estimation of the MC which is related to original definition. With the absence of numerical derivative, the estimated results can be more stable and robust. Statistical analysis is conducted to show the effectiveness and noise immunity of the proposed method. In order to obtain the better identification, the MC calculated by the proposed method is employed as the input of continuous wavelet transform, and then the hybrid method is generated. The validations of the present method and comparison with the traditional central difference method are numerically and experimentally demonstrated.
Author Zhi-bo Yang
Zhi-bo Yang,,
, M. Radzieński
M. Radzieński,,
, Paweł Kudela
Paweł Kudela,,
, Wiesław Ostachowicz IP
Wiesław Ostachowicz,,
- Institute of Automotive Engineering
Journal seriesMechanical Systems and Signal Processing, ISSN 0888-3270
Issue year2016
NoPart A
Publication size in sheets0.9
Keywords in EnglishFourier transform; Modal curvature; Damage detection; Beam; Wavenumber filtering
Languageen angielski
Score (nominal)45
ScoreMinisterial score = 45.0, 29-11-2017, ArticleFromJournal
Ministerial score (2013-2016) = 45.0, 29-11-2017, ArticleFromJournal
Publication indicators WoS Impact Factor: 2016 = 4.116 (2) - 2016=4.874 (5)
Citation count*0
Share Share

Get link to the record

* presented citation count is obtained through Internet information analysis and it is close to the number calculated by the Publish or Perish system.