Filtern
Dokumenttyp
Sprache
- Englisch (3)
Referierte Publikation
- ja (3)
Schlagworte
- Damage detection (3) (entfernen)
Organisationseinheit der BAM
The fourth dataset dedicated to the Open Guided Waves platform presented in this work aims at a carbon fiber composite plate with an additional omega stringer at constant temperature conditions. The dataset provides full ultrasonic guided wavefields. Two types of signals were used for guided wave excitation, namely chirp signal and tone-burst signal. The chirp signal had a frequency range of 20-500kHz. The tone-burst signals had a form of sine modulated by Hann window with 5 cycles and carrier frequencies 16.5kHz, 50kHz, 100kHz, 200kHz, 300kHz. The piezoceramic actuator used for this purpose was attached to the center of the stringer side surface of the core plate. Three scenarios are provided with this setup: (1) wavefield measurements without damage, (2) wavefield measurements with a local stringer debond and (3) wavefield measurements with a large stringer debond. The defects were caused by impacts performed from the backside of the plate. As result, the stringer feet debonds locally which was verified with conventional ultrasound measurements.
Understanding on how a machine learning model interprets data is a crucial step to verify its reliability and avoid overfitting. While the focus of the scientific community is nowadays orientated towards deep learning approaches, which are considered as black box approaches, this work presents a toolbox that is based on complementary methods of feature extraction and selection, where the classification decisions of the model are transparent and can be physically interpreted. On the example of guided wave benchmark data from the open guided waves platform, where delamination defects were simulated at multiple positions on a carbon fiber reinforced plastic plate under varying temperature conditions, the authors could identify suitable frequencies for further investigations and experiments. Furthermore, the authors presented a realistic validation scenario which ensures that the machine learning model learns global damage characteristics rather than position specific characteristics.
Understanding on how a machine learning model interprets data is a crucial step to verify its reliability and avoid overfitting. While the focus of the scientific community is nowadays orientated towards deep learning approaches, which are considered as black box approaches, this work presents a toolbox that is based on complementary methods of feature extraction and selection, where the classification decisions of the model are transparent and can be physically interpreted. On the example of guided wave benchmark data from the open guided waves platform, where delamination defects were simulated at multiple positions on a carbon fiber reinforced plastic plate under varying temperature conditions, the authors could identify suitable frequencies for further investigations and experiments. Furthermore, the authors presented a realistic validation scenario which ensures that the machine learning model learns global damage characteristics rather than position specific characteristics.