Assessing the efficacy of non-destructive testing methods to detect pitting corrosion

Calibration of the non-destructive testing (NDT) methods for pipeline inspections is commonly done via machined defects with simple shapes and known sizes. However, such defects do not represent, e.g. an actual pit morphology observed in the field, which could lead to the unreliable interpretation of the inspection results. This work proposed a systematic approach to create and characterise real pits in stainless steel pipe specimens to assess the sensitivity of selected NDT methods. Pipe samples with three different diameters and thicknesses were used. Pits of different geometries and dimensions were generated by electrochemical methods, and immersion in an acidic ferric chloride solution filled in pre-drilled holes on the external surface of the specimens. NDT methods including radiography, Eddy current and dye penetrant testing were used for detecting and sizing the pits. The results were then compared with true pit depths obtained from cross-section analysis. Radiography and Eddy current methods failed to detect small pits with depths less than 70 µm. The porous metal cover on the pit mouth affected the detectability of radiography, while it had little impact on Eddy current testing. The findings of this investigation could be used to design training and calibration blocks for NDT methods.