Matched-field processing for leak localization in a viscoelastic pipe: An experimental study

Xun Wang, Jingrong Lin, Alireza Keramat, Mohamed S. Ghidaoui, Silvia Meniconi, Bruno Brunone

Research output: Journal article publicationJournal articleAcademic researchpeer-review

34 Citations (Scopus)

Abstract

This paper applies the matched-field processing (MFP) method to leakage localization in a viscoelastic pipe. The viscoelasticity of pipe wall is included in the governing equations of transient wave via the generalized Kelvin-Voigt model and its effect is finally translated into a frequency-dependent wave speed. Then, a leak is localized by MFP via a 1D search of leak location along the pipe, independent of the leak size. Transient experiments with viscoelastic pipe in the Water Engineering Laboratory at University of Perugia and in the Water Resources Research Laboratory at Hong Kong University of Science and Technology are studied. Experimental results demonstrate that the inclusion of pipe wall viscoelasticity and using more frequencies (instead of using only resonant frequencies) improve significantly the leak localization accuracy. It is shown that the MFP leak localization is accurate even for a small leak (the flow ratio of leak and main pipe is approximately 10%) in a noisy environment: among 50 transient experiments, the maximum error of MFP leak localization is only 1.14 m and in the other 49 experiments the error is always lower than 1 m.

Original languageEnglish
Pages (from-to)459-478
Number of pages20
JournalMechanical Systems and Signal Processing
Volume124
DOIs
Publication statusPublished - 1 Jun 2019
Externally publishedYes

Keywords

  • Complex environment
  • Leakage localization
  • Matched-field processing
  • Pipe viscoelasticity
  • Transient wave

ASJC Scopus subject areas

  • Control and Systems Engineering
  • Signal Processing
  • Civil and Structural Engineering
  • Aerospace Engineering
  • Mechanical Engineering
  • Computer Science Applications

Cite this