Identifying Pipe Degradation in a Highly Dynamic Environment Using Singular Value Decomposition

Abstract

Pipe corrosion is a common degradation that can cause leakages, breaks, or even explosions. Detecting and monitoring corrosion is necessary to ensure to the structural integrity of pipes. Ultrasonic guided wave techniques are one promising solution due to their sensitivity to small defects and capability to propagate long distance. However, it is difficult to interpret pipe guided wave signals directly or to extract effective corrosion-sensitive features from them due to their multi-path, multimodal, dispersive characteristics. Moreover, pipe structures often experience large variations in environmental and operational conditions (EOC), which can produce large changes in ultrasonic signals and can easily produces false alarms. In previous work, we developed a robust damage detection method based on singular value decomposition (SVD). In this paper, we further explore, through experiments, the capability of our SVD-based method to detect degradation. We demonstrate that our method is effective at identifying progressive corrosion at the same time that large temperature variations are present.