Inverse problem analyses with noised signals in eddy current testing

Haoyu Huang, Hiroyuki Fukutomi, Toshiyuki Takagi

Research output: Contribution to journalArticlepeer-review

2 Citations (Scopus)

Abstract

This paper demonstrates crack reconstruction from the relative signals, which overlap with noises in eddy current testing. A method for the reconstruction from the signals, a kind of inverse problem, has been developed with fast signal predictions and the steepest descent algorithm. The signal predictions use a pre-computed unflawed database approach based on edge-based finite elements and the reciprocity theorem. This approach makes it possible that the problem domain of an electromagnetic field is reduced to a small region where a crack/cracks possibly exist. In the inspection of steam generator tubes in nuclear power plants, this test detects deposits on the surfaces of the tubes and structures to fix the tubes such as support sheets or plates, and rigs, as well as defects. The signals from the structures are thought to be noises when finding out the defects. Crack profile identification is given from the crack signals with residual noises, which remain after multifrequency processing of the signals disturbed by the presence of copper/magnetic deposits and support plates.

Original languageEnglish
Pages (from-to)63-70
Number of pages8
JournalNihon Kikai Gakkai Ronbunshu, A Hen/Transactions of the Japan Society of Mechanical Engineers, Part A
Volume66
Issue number641
DOIs
Publication statusPublished - 2000

Keywords

  • Copper/magnetite deposits
  • Crack reconstruction
  • Eddy current testing
  • Finite element method
  • Inverse problems
  • Magnetic vector potentials
  • Noised signals
  • Pre-computed unflawed database approaches
  • Steam generator tubes
  • Support plates

ASJC Scopus subject areas

  • Materials Science(all)
  • Mechanics of Materials
  • Mechanical Engineering

Fingerprint Dive into the research topics of 'Inverse problem analyses with noised signals in eddy current testing'. Together they form a unique fingerprint.

Cite this