Classification of steam generator tube defects for real-time applications using eddy current test data and self-organizing maps (Q1770577)

For the steam generators incorporated in a pressurized water reactor, a new method for isolating tube defects is proposed. Using eddy current test (based on the electromagnetic interaction between a bobbin coil and an electrically conductive specimen) and the self-organizing maps (a neural network) it is possible to identify and classify the tube defects. The classification algorithm used by the authors permits to identify defect types for nuclear plant steam generators, and with the codebook maps can be estimated the main signature for each defect.