Classification Analysis Of High Frequency Stress Wave For Autonomous Detection Of Defect In Steel Tubes

• Main Authors: Abd Halim, Zakiah, Jamaludin, Nordin, Junaidi, Syarif, Syed Yahya, Syed Yusaini
• Format: Article
• Language: English
• Published: AENSI 2014
• Subjects: T Technology (General); TJ Mechanical engineering and machinery
• Online Access: http://eprints.utem.edu.my/id/eprint/21042/2/2014%20AJBAS.pdf; http://eprints.utem.edu.my/id/eprint/21042/; http://www.ajbasweb.com/old/ajbas/2014/Special%203/251-257-special14.pdf
• Institution: Universiti Teknikal Malaysia Melaka

Interpretation of propagated high frequency stress wave signals in steel tubes is noteworthy for defect identification.This paper demonstrated a successful new approach for autonomous defect detection in steel tubes using classification analysis of high frequency stress waves.Classification analysis using Principal Component Analysis (PCA) algorithm involved feature extraction to reduce the dimensionality of the complex stress waves propagation path.Two defective tubes containing a slot defect of different orientation and a reference tube are inspected using Vibration Impact Acoustic Emission (VIAE) technique.The tubes are externally excited using impact hammer.The variation of stress wave transmission path are captured by high frequency Acoustic Emission sensor.The propagated stress waves in the steel tubes are classified using PCA algorithm.Classification results are graphically illustrated using a dendrogram that demonstrated the arrangement of the natural clusters of the stress wave signals.The inspection of steel tubes showed good recognition of defect in circumferential and longitudinal orientation.This approach successfully classified stress wave signals from VIAE testing and provide fast and accurate defect identification of defective steel tubes from non-defective tubes.