High-Sensitivity Ultrasonic Guided Wave Monitoring of Pipe Defects Using Adaptive Principal Component Analysis

Area of Science:

• Structural Health Monitoring
• Non-Destructive Testing
• Signal Processing

Background:

• Ultrasonic guided wave monitoring is crucial for industrial pipe integrity.
• Environmental and structural factors complicate the detection of small pipe defects.
• Existing methods struggle with sensitivity and accuracy for subtle damage.

Purpose of the Study:

• To develop a high-sensitivity algorithm for detecting small defects in industrial pipes.
• To improve the accuracy of ultrasonic guided wave structural health monitoring.
• To provide an intuitive method for visualizing and locating pipe defects.

Main Methods:

• Proposed an adaptive principal component analysis (APCA) algorithm.
• Introduced a sensitivity index for signal analysis.
• Optimized principal component selection within PCA.
• Developed a comprehensive damage index (K) using subspace signal features.

Main Results:

• APCA successfully detected small defects (0.075%–0.18% cross-section loss ratio) across straight, spiral, and bent pipes.
• The APCA method demonstrated superior performance compared to optimal baseline subtraction (OBS) and average Euclidean distance (AED).
• The damage index (K) curve intuitively displayed defect trends and roughly indicated defect locations.

Conclusions:

• The APCA algorithm offers a significant advancement in detecting small defects in industrial pipes using ultrasonic guided waves.
• This method provides a more sensitive and accurate approach to structural health monitoring.
• The developed damage index offers intuitive defect visualization and localization capabilities.