Table of Content

Open Access iconOpen Access


Electro-Mechanical Impedance Based Position Identification of Bolt Loosening Using LibSVM

Yuxiang Zhanga, Xin Zhanga, Jiazhao Chenb, Jianhai Yanga

a 602 office, Power Engineering Department, Xi’an Research Inst. of Hi-Tech, Xi’an, China;
b 601 office, Power Engineering Department, Xi’an Research Inst. of Hi-Tech, Xi’an, China

* Corresponding Authors: Yuxiang Zhang, email; Xin Zhang email

Intelligent Automation & Soft Computing 2018, 24(1), 81-88.


Bolt loosening is a common structural failure, which received extensive attention from many industrial departments. Because the uneven stress on different directions of a bolt is the common reason that the bolt becomes loose, it is quite important to carry out the research about sensitive detection of bolt loosening. Using Agilent 4924A instrument, this paper precedes the loosening test of bolts based on Electro-Mechanical Impedance (EMI), on an aluminum plate instead of flange plate for simplification. And the electromechanical admittance is given according to the fundamental equation of PiezoMaterial Lead Zirconate Titanate (PZT). Specifically, the paper first studies the detecting sensitivity of EMI on bolt loosening; then, it shows that RMSD can be seen as a good damage index to identify damage; at last, our experiment result shows that by using LibSVM to process big data, the position of a loose bolt can be correctly identified from 12 possible bolt positions. The method mentioned in this paper shows the great potential to be used for the damage monitoring of bolted structure.


Cite This Article

Y. Zhang, X. Zhang, J. Chen and J. Yang, "Electro-mechanical impedance based position identification of bolt loosening using libsvm," Intelligent Automation & Soft Computing, vol. 24, no.1, pp. 81–88, 2018.

cc This work is licensed under a Creative Commons Attribution 4.0 International License , which permits unrestricted use, distribution, and reproduction in any medium, provided the original work is properly cited.
  • 1167


  • 812


  • 0


Share Link