Characterisation of the probability of detection for ultrasonic nondestructive testing of dirty white spots in GH4169

The Dirty White Spot (DWS) defect in GH4169 (In718) alloy is a critical flaw known to cause catastrophic aviation accidents, yet its probability of detection (POD) via ultrasonic testing (UT) has remained unknown. To address this issue, this paper combined the model-assisted detection probability (MAPOD) method with experiments to evaluate the POD of DWSs. The MAPOD method was constructed based on a spherical wave model to predict the signal strength of DWS at different depths and sizes, reducing the number of experimental samples required. Experiments prepared samples containing artificial DWSs with varying depths, sizes, and grain sizes to obtain signals and noise, addressing the scarcity of natural defects. Finally, the ability of UT to detect DWSs was analysed and quantitatively characterised. Results showed that the average signal reflectance of DWS was less than 0.04 and was related to DWS grain size. The low signal reflectance of DWS makes its POD susceptible to depth effects, with an average maximum POD difference of 0.495 when the depth differs by 50 mm. This study provides a reference for assessing UT capabilities in damage tolerance analysis related to DWS in aeroengine disks.