Calculating probability of detection of short surface cracks using inductive thermography

Inductive thermography experiments on artificial defects were used to calculate the POD of this inspection technique. In nickel-based austenitic superalloy samples defects in the range of 0.3–1.9 mm length and with 0.25–1.5 mm depth were inspected. 11 samples, each of them with 9 different defects, were used for the experiments. Additionally, finite element simulations were carried out to support the POD calculation and in order to investigate how different parameters affect the results. The measured and the simulated temperature sequences were evaluated by Fourier transform to a phase image, and the signal-to-noise ratio of the phase contrast around the crack was used as signal â for the â versus a POD calculation. As the phase contrast depends not only on the defect length but also on the defect depth, POD depending on both variables was calculated, which can be visualised as a POD surface. Finally, experimental results on samples with real cracks (fatigue cracks and cracks obtained through the Varestraint test machine) were compared to the results obtained by the FEM simulation and in the experiments for the artificial defects.