Listen2theSOURCE

Listen2theSOURCE develops the necessary tools for reliably identifying different types of damage in fiber reinforced plastics by acoustic emission analysis.

Structural Health Monitoring (SHM) systems for fiberreinforced plastics (FRP) offer the potential to reduce downtime as well as inspection costs. System costs and reliability are currently the biggest market entry barriers for this technology. This study is dedicated to both factors with the development of a low-cost demonstrator, which is able to reliably localize artificial sound sources (pencil lead breaks) in an FRP structure on the basis of their acoustic emissions. The localization of acoustic events in real components is a challenge due to anisotropic material properties and variations in wall thickness. Hence, in this study, neural networks are implemented, which are available through a targeted training to capture the complexity of the task in the network architecture. Piezoelectric sensors, which are characterized by small installation space and low weight, are used for the detection of the waves. Measurement data acquisition and processing is realized via a digital 4-channel oscilloscope and a single-board computer.

This approach demonstrates in an impressive way the implementation of a resource-saving SHM system that can be transferred to other structures and applications.