Keywords: Concrete Engineering
-
Application of unsupervised AI-assisted acoustic wave sound analysis for non-destructive detection of steel corrosion induced deterioration
Reinforced concrete structures require reliable monitoring to ensure safety and efficient maintenance. Non-destructive testing methods such as tapping sound inspection are widely applied. However, the diagnosis results often depend on technical expert skill and experience. This study proposes an easy-to-use, AI-based evaluation method for tapping sounds using unsupervised deep learning. Laboratory tests were carried out on reinforced concrete beams with simulated steel bar corrosion. The method proposes an anomaly index that reflects corrosion progress and surface cracking. The results demonstrate that acoustic inspection with AI can support early damage detection and improve condition assessment of concrete structures.
Nopphanan Phannakham, Katsufumi Hashimoto, Yasuhiko Sato, and Naoshi Ueda
Mesoscale modeling of anisotropic compressive behavior and pull-out performance of 3D printed concrete with steel bars using 3D RBSM
This study uses a 3D Rigid Body Spring Model (RBSM) to analyze the anisotropic behavior of 3D-printed concrete (3DPC) with steel reinforcement. Validated by experiments, the research highlights how the mesoscale structure—specifically porous interlayer interfaces—affects performance. Results indicate that specimens loaded parallel to the printing direction exhibit superior compressive strength and bond performance. Conversely, loading perpendicular to the layers leads to stress concentrations and weaker bonds due to interfacial zones. Overall, this research provides a predictive framework for optimizing the structural integrity of 3DPC through mesoscale modeling.
Jiaxu Yao, Jie Luo, Minghong Qiu, and Kohei Nagai
Impact-echo for different level cracks detection in concrete with artificial intelligence based on un/supervised deep learning
Aging concrete infrastructure such as bridges and tunnels requires effective inspection to ensure safety and durability, particularly for detecting invisible internal cracks subjected to structural integrity. Impact-echo, which is one of non-destructive testing methods, is widely used but costly and time-consuming with relying on skilled and experienced analysis. This study integrates AI with impact-echo data to improve crack detection. Supervised deep learning using FFT-transformed signals enables accurate classification of multiple crack levels, including intact condition of concrete. However, data labeling for each existing structures is impractical, so an unsupervised approach using an auto-encoder is proposed to identify internal crack levels through anomaly-based indices without labeled data.
Jeero Pandum, Katsufumi Hashimoto, Takafumi Sugiyama, Wanchai Yodsudjai