Combining acoustic emission with passive thermography to characterize damage progression in cross-ply CFRP laminates during quasi-static tensile loading

Kelkel, Benjamin; Institut fur Verbundwerkstoffe GmbH; Germany

Kelkel, B.; Institut für Verbundwerkstoffe GmbH; Germany
Popow, V.; Institut für Verbundwerkstoffe GmbH; Germany
Gurka, M.; Institut für Verbundwerkstoffe GmbH; Germany

ID: ECNDT-0078-2018
Download: PDF
Session: Composite Material - AE
Room: G1
Date: 2018-06-12
Time: 09:00 - 09:20

Understanding the gradual failure process of carbon fiber reinforced plastics (CFRP) is the key for exploiting their full potential for lightweight applications. Acoustic emission (AE) can support this process through the detection and evaluation of transient acoustic signals released from loaded CFRP specimen in the moment of damage initiation and progression. This way, not only the presence of damage, but also its location, severity and type can be determined by analyzing arrival times as well as energy and frequency contents of the acquired acoustic signals.
In order to differentiate between different types of damage, one has to identify their acoustic signature. This is usually done by correlating extracted AE parameters from acquired signals during a mechanical test with the resulting damage pattern of the specimen that is visualized offline in a time consuming process by imaging methods such as X-ray tomography or microscopy.
In this study, passive thermography is utilized to identify occurring damage inline on the basis of their released heat patterns to support the AE analysis in the identification of acoustic fingerprints and the characterization of damage progression.
Mechanical tests are performed on cross-ply CFRP coupon specimen subjected to quasi-static tensile loading in the 0° and 90° fiber direction while an IR camera and two wide band AE sensors are utilized to monitor the specimen during the test. Heat patterns are extracted from the series of IR data through advanced image processing techniques and correlated with the generated AE signals in order to identify damage modes such as fiber breakage or matrix fracture. An unsupervised pattern recognition approach is then utilized to find similar AE signals and characterize the gradual failure process in the specimen. The outcome is validated with the resulting damage pattern that is visualized offline via X-ray tomography after the mechanical test.