|
It is well known that guided ultrasonic waves are suitable to detect damages in composite plates. It has also been shown that these Lamb waves can be utilized to infer material properties through nondestructive measurements. More recently, it was shown that this may be used to determine regional inhomogeneity that is inherent to composite materials due to manufacturing imperfections. In this project, it is investigated whether automated processing of Lamb wave-based data is generally suitable to detect such imperfections. Woven prepreg and short-fiber composite panels are manufactured. A large set of nondestructive measurements are conducted to determine dispersion and attenuation characteristics for multiple regions across each panel. Automated signal processing is performed to extract characteristic features of the signal, which are in turn used to identify any differences within the panels. Moreover, it is studied which type of sensing technology, such as contact transducers, air-coupled transducers or a laser Doppler vibrometer are most suitable for this task. That is, ultrasound measurements with different actuator and sensor combinations are accompanied by additional transducer characterization measurements. Optimal frequency ranges for each transducer are determined in addition to studying potential effects of transducer orientation. Based on all findings, it can be concluded that detecting regional inhomogeneity remains challenging due to various compounding limitations, such as optimal transducer frequency ranges, human error and generally low signal-to-noise ratios in Lamb wave-based measurements, especially at longer propagation distances. In turn, the development of guided wave-based nondestructive evaluation methods require a holistic approach with careful considerations of the employed transducers.
|
