Enhancing generalizability of a machine learning model for infrared thermographic defect detection by using 3d numerical modeling Vladimir Vavilov, Arsenii Chulkov, Alexey Moskovchenko Frattura Ed Integrita Strutturale, 2024 The paper describes the implementation of 3D numerical simulation in machine learning models used in infrared thermographic nondestructive testing. The enhancement of generalizability of such models emerges as a decisive factor for producing trust-worthy test results. First, it is demonstrated that the models trained on datasets with fixed parameters yield limited defect detection capabilities. The concept of training datasets, which include subtle variations in material thickness, thermal conductivity, as well as various combinations of material density and heat capacity, provides the best learning results and a noticeable ability to identify defects in all test datasets. Second, the model robustness in respect to noise is explored to demonstrate its ability to withstand additive and multiplicative random noise. Third, potentials of some known techniques of thermographic data processing, such as Thermographic Signal Reconstruction, Fast Fourier Transform and Temperature Contrast, are examined. In particular, the use of the Temperature Contrast data ensured sensitivity (True Positive Rate) better than 98% across all test datasets.
EFFECT OF ANTI-CORROSION PROTECTIVE PAINT ON THERMOGRAPHIC INSPECTION OF CURVED STEEL TUBE PARTS Michal ŠVANTNER, Alexey MOSKOVCHENKO, Lukáš MUZIKA Metal International Conference on Metallurgy and Materials Conference Proceedings, 2024 The use of anti-corrosion paint coatings on steel pipes is a common practice for their protection against corrosion. However, such coatings may have an impact on the results of the thermographic inspection, which can be used, for example, for their corrosion damage identification. In this study, we investigated the influence of anti-corrosion paint on the thermographic inspection of curved steel tube parts. Long pulse thermography inspection was conducted on painted and unpainted samples, and the results were compared. It was found that the used painting reduced the absorbed energy, however, the contrast of the found defect indications was better on painted samples. The experiments indicated that any inhomogeneity of an inspected surface due to, for example, the painting process, the presence of old painting layers, or the presence of surface corrosion, can cause irregular surface absorption patterns. It can result in signals from this unevenness that can reduce the contrast of the indications of defects. These findings can have significant implications for the use of thermography as a non-destructive testing technique for curved steel tube parts, for example, steel pipes, especially those in operation with correction paint or corrosion layers.
