Conference Agenda

We present some results obtained in the ongoing research project entitled “IN MOOSHEAC - INnovative damage MOnitoring Of Self-HEAling Composites by acoustic emission in civil and aerospace applications”. The research program aims at utilizing Nondestructive Evaluation (NDE) techniques for monitoring the healing process of healable composites and to investigate their structural properties post healing. In this campaign, samples of composite thin rectangular strips were designed and manufactured on purpose, and then tested in the lab under tension. The samples were supplied by a company that has developed a new type of resin - HealTech™ - giving composite materials the ability to heal micro cracks and delamination in about one minute after a localized heat supply. Acoustic Emission (AE) technique was used to detect damage, while Digital Image Correlation (DIC) was adopted to measure surface deformation. Data from both techniques were employed to assess material repair.

This study was carried out for the French National Radioactive Waste Management Agency (Andra) as part of the Cigéo project, a deep geological disposal facility for radioactive waste. Cigéo consists of a network of concrete galleries where the radioactive wastes will be disposed. After the construction and operation phase, the galleries will be sealed, and no further human intervention will be possible. Therefore, to monitor the health of the galleries remotely, structural health monitoring (SHM) techniques are required. The study focused on using three different monitoring techniques, namely acoustic emission (AE), strain measurements using optical fibers, and electrical measurements with embedded sensors to test the self-sensing ability of the concrete. The objective was to cross-reference data from these various techniques to propose methods for detecting cracks in concrete elements. The obtained results were compared with those from digital image correlation (DIC). These tests were carried out on structural elements, i.e. beams subjected to 4-point bending test. With the dual aim of improving the relevance of these monitoring techniques and better controlling crack propagation, fibers were incorporated into the concrete mixes. Two types of fibers were used: amorphous metallic fibers (AMF) and carbon fibers (CF). The results showed that CF primarily enhanced crack detection through electrical measurements without significant influence on mechanical performance. However, AMF demonstrated their ability to restrain crack propagation and improved the relevance of electrical and AE measurements to monitor crack initiation and propagation.

Much of the existing literature has focused on improving concrete's fracture response by focusing on binder composition. However, the critical role of aggregates and their packing efficiency on fracture response has not been studied in detail. Here, we investigate the impact of packing enhancement on fracture behavior by designing 4 unique concrete mixes with customized aggregate skeletons, utilizing a continuous packing model. Fracture assessment using the two-parameter fracture model (TPFM) and high-speed cross-sectional Digital Image Correlation (DIC) was employed to acutely capture strain distribution and crack propagation from 95% of the pre-peak to the peak load. We find that improving coarse aggregate packing efficiency from 88% to 93%, while maintaining constant binder content, increased fracture energy (Gf) by 35.29% and 39.48% for 3 and 7 days respectively. Likewise, for 3 and 7 days the critical stress intensity factor (K1C) increased by 46.08% and 51.66% respectively with the increase in packing efficiency. In summary, the results obtained quantify the impact of aggregate packing on concrete’s fracture response.

Understanding the mechanism of crack growth in cement-based materials under mechanical loading involves complex interactions between microstructural components, including aggregates, voids, and cement paste. This paper presents a unique approach that combines X-ray computed tomography (XCT) with deep learning to segment these components precisely. By leveraging XCT's high-resolution 3D imaging capabilities and the robustness of deep learning algorithms, our method provides a detailed characterization of the microstructure of cement-based materials. This detailed structural information is crucial for understanding crack initiation and propagation processes, ultimately contributing to developing more durable and sustainable concrete. Our results highlight the significant potential of deep learning in enhancing our understanding of damage and failure mechanisms in cement-based materials, providing valuable insights that can lead to improved material performance and longevity.