Pipeline Flange Bolt Loosening Detection Technology Based on Stress Waves and Deep Learning

Area of Science:

• Mechanical Engineering
• Industrial Safety
• Non-destructive Testing

Background:

• Flanged connections are vital in industrial production, necessitating bolt loosening detection for safety.
• Existing research primarily addresses flat-face flanges without gaskets, leaving a gap in understanding gasketed pipe flanges.
• Bolt loosening in gasketed pipe flanges poses significant risks to industrial operations.

Purpose of the Study:

• To develop a method for detecting bolt loosening in gasketed pipe flanges.
• To analyze the impact of bolt loosening on stress wave propagation within gaskets.
• To establish a reliable system for assessing the severity of bolt loosening in pipeline flanges.

Main Methods:

• Utilized stress wave principles and finite element simulation to analyze wave propagation influenced by bolt loosening.
• Employed the hammer impact method for experimental detection of bolt loosening.
• Determined optimal experimental parameters including knock force and hammer head material.
• Applied and optimized a Support Vector Machine-Recursive Feature Elimination (SVM-RFE) model with feature enhancement and cost-sensitive learning.

Main Results:

• The hammer impact method effectively detected bolt loosening in gasketed pipe flanges.
• The SVM-RFE model demonstrated high accuracy and efficiency in assessing bolt loosening degree.
• Optimized SVM-RFE model provided enhanced performance for severity identification.
• Experimental validation confirmed the reliability of the proposed method.

Conclusions:

• The developed method offers a reliable solution for rapid identification of bolt loosening severity in pipeline flanges.
• The study contributes to enhanced industrial safety through improved bolt loosening detection techniques.
• The optimized SVM-RFE model represents a significant advancement in non-destructive testing for bolted connections.