Type-IV carbon-fiber-overwrapped pressure vessel in burst chamber, equipped with acoustic emission sensors.
Source: BAM
The study investigates two type IV pressure vessels made of carbon fiber reinforced polymer (CFRP) that were hydraulically loaded to burst. The aim is to analyze damage progression and identify critical failure mechanisms such as fiber breaks using acoustic emission (AE) – a non-destructive testing method that detects microscopic material damage. The vessels were manufactured differently and therefore fail in different fiber layer orientations (hoop and helical layers). More than 200,000 AE events were recorded during stepwise pressure increases. The evaluation includes frequency features, amplitudes, cluster analyses, as well as the Felicity and Shelby ratios, which provide information about the load history and the sensitivity of a material to repeated loading.
Although previous studies suggest that fiber breaks are characterized by high frequencies (typically 300–600 kHz), the measurements here reveal a more complex picture: high-frequency signals do tend to occur at high loads, but they cannot be clearly and consistently assigned to specific damage types. Reasons include the superposition of different damage mechanisms, frequency-dependent attenuation, differences in sensor coupling, and the complex laminate structure. However, one finding is clear: there is a pronounced localization of AE activity immediately before burst failure in the region of the eventual fracture surface—an indicator that reliably signals imminent failure.
Overall, the study shows that robust identification of individual failure mechanisms—especially fiber breaks—based solely on AE frequency signatures is not reliably achievable in real pressure vessels. The most reliable information is provided by the spatial and temporal clustering of AE events shortly before failure.
Monitoring Failure of Composite Pressure Vessels with Acoustic Emissions
Emanuel D. Kästle, Eric Duffner, Bartosz Popiela, Ali Ghaznavi
Journal of Nondestructive Evaluation, 2025