Abstract

In this presentation, I will outline the concept and practical implementation of using dispersive guided waves to measure remaining wall thickness. Dispersive modes offer greater sensitivity to variations in wall thickness than conventional, nearly non-dispersive guided-wave modes. However, due to the inherent complexity of their signals, they have historically seen limited application in NDT practice.

We analyse cut-off frequencies derived from dispersion curves to assess their correlation with the minimum remaining wall thickness and employ this approach for the quantitative assessment of corrosion at pipeline supports. Incorporating machine learning techniques allows for the automatic analysis of complex multi-parameter dispersion data, improving measurement accuracy. 

A practical NDT instrument that integrates newly developed algorithms with electromagnetic ultrasonic transducers and motion systems provides a reliable solution for demanding corrosion inspection tasks.