Enhancement of damage identification in composite structures with self-heating based vibrothermography

The self-heating vibrothermography non-destructive testing method is a promising approach for testing polymeric composite structures with a single-side access or inability of application of external heating sources for thermal excitation. The method is based on excitation of low-frequency mechanical vibrations with multiple resonant frequencies, which allows for local heating of a tested structure, and based on a captured thermal response, identification of structural damage. Although the method is quite sensitive for damage, the resulting damage signatures in thermograms are often blurred or invisible due to the measurement noise and non-uniform temperature evolution on the surface of a tested structure. In this paper, the authors examine various post-processing techniques applied for infrared thermograms in order to enhance damage detectability in polymeric composite structures. Three types of methods were examined: methods based on statistical features, a new method based on trend estimation of the temperature profiles, and a method based on similarity maps. The results of the processing reveal visible enhancement of the damage detectability using the proposed methods, which allow detecting and identifying more damage signatures compared to single raw thermograms.