Classification of artificial near-side cracks in aluminium plates using a GMR-based eddy current probe

Nestor Orlando Romero Arismendi, Eduardo Ramirez Pacheco, Oswaldo Panzo Lopez, J. H. Espina-Hernandez, Jose Alberto Perez Benitez

Research output: Chapter in Book/Report/Conference proceedingConference contributionpeer-review

2 Scopus citations

Abstract

In the present work a method for the nondestructive classification of crack s width, depth and orientation by using an asymmetrical eddy current GMR-Coil Configuration is proposed. The experimental measurements where performed along five directions in a series of artificial flat cracks with different widths and depths in aluminum. DV and DX values for each of the defects are obtained. The results show that there is a single pair DV, DX for each studied cracks width, depth and orientation. The results also reveal that it is possible to make a classification and estimation of the cracks width, depth and orientation using supervised learning models.

Original languageEnglish
Title of host publication2018 28th International Conference on Electronics, Communications and Computers, CONIELECOMP 2018
PublisherInstitute of Electrical and Electronics Engineers Inc.
Pages31-36
Number of pages6
ISBN (Electronic)9781538623633
DOIs
StatePublished - 27 Mar 2018
Event28th International Conference on Electronics, Communications and Computers, CONIELECOMP 2018 - Cholula, Mexico
Duration: 21 Feb 201823 Feb 2018

Publication series

Name2018 28th International Conference on Electronics, Communications and Computers, CONIELECOMP 2018
Volume2018-January

Conference

Conference28th International Conference on Electronics, Communications and Computers, CONIELECOMP 2018
Country/TerritoryMexico
CityCholula
Period21/02/1823/02/18

Keywords

  • GMR sensor
  • eddy currents
  • machine learning
  • nondestructive testing

Fingerprint

Dive into the research topics of 'Classification of artificial near-side cracks in aluminium plates using a GMR-based eddy current probe'. Together they form a unique fingerprint.

Cite this