Thermal NDE of delaminations in plastic materials by neural network processing

Neural networks are applied to the problem of detecting and classifying voids inside an opaque material at different depths. A classic one-side dynamic thermographic testing procedure is applied to artificial defects buried in 14 mm specimens of PVC. Experimental data are automatically processed. extracting the maximum and the corresponding time of thermal contrast profile versus time. A two steps procedure was developed and tested using an intentionally uneven heating of the sample. The obtained results are presented, demonstrating the robustness and accuracy of the developed technique.