Reliability of Dye Penetrant Inspection Method to Detect Weld Discontinuities

Abstract

The dye penetrant inspection method used to reveal surface weld discontinuities is an important factor for quality verification in the manufacture of structural components; however, it is probable that certain size-dependent discontinuities may or may not be detected. Then, how reliable can it be? In this sense, the objective of the research has been to estimate the reliability of dye penetrant inspection to detect discontinuities in relation to their size. Six experimental tests were performed by three inspectors, with visible and fluorescent dye penetrants, on twenty welded joints of similar surface characteristics, containing 63 typical weld discontinuities arranged according to shape and size, whereas POD reliability quantitative estimates were developed by the hit-or-miss statistical method. For the test system, the fluorescent penetrants, due to their greater sensitivity compared to the visible ones, registered greater reliability in revealing smaller discontinuities. The POD estimators were a₅₀ (1.469 mm < 1.978 mm), a₉₀ (6.348 mm < 7.474 mm), a_90/95 (14.58 mm < 15.77 mm). Fluorescent dyes allowed a higher rate and probability of detection; both factors showed a tendency to increase as the discontinuities size increased.