Underwater inspection of plastic pipes using guided waves from mobile piezoelectric transducers

Guided wave inspection of water-filled plastic pipelines is challenging due to high attenuation, limiting the effectiveness of methods designed for metallic pipes. This paper proposes a method for inspecting water-filled plastic pipes from inside the pipes using mobile robots equipped with water-coupled piezoelectric transducers. The region of interest (ROI) is determined by comparing the predicted defect signal with the noise signal under realistic environmental conditions. The defect signal is modelled based on guided wave propagation and defect scattering, while noise is measured experimentally. The detection threshold is established using the probability of detection (POD) and false alarm (PFA) criteria, defining the ROI where defects can be reliably detected. The proposed method was experimentally validated on an underwater UPVC pipe with manually manipulated transducers. Both pulse-echo (PE) and pitch-catch (PC) configurations successfully detected the target defects. While the PC configuration offers larger coverage, it relies on baseline signals, limiting robustness. In contrast, the PE configuration is more sensitive, detecting smaller defects with fewer false alarms. Further analysis reveals that angular misalignment impacts detection coverage, with the PC configuration being more sensitive to misalignment. This highlights the importance of alignment control during robotic inspection to ensure reliable performance.