| Title |
Construction of High-quality CFRP Acoustic Emission Failure Dataset Through Representation Learning |
| Authors |
김다현(Da Hyun Kim) ; 황병일(Byeong Il Hwang) ; 김호연(Ho Yeon Kim) ; 서영주(Young Joo Suh) ; 김동주(Dong Ju Kim) |
| DOI |
https://doi.org/10.5573/ieie.2025.62.3.74 |
| Keywords |
CFRP; Failures; Deep learning; Representation learning; Acoustic emission testing |
| Abstract |
Carbon Fiber Reinforced Polymer (CFRP) is widely used in various fields such as aerospace, civil engineering, and military due to its lightweight, high tensile strength, and low thermal expansion coefficient. However, the combination of different molecular structures makes CFRP prone to fatigue-induced failures, making it crucial to detect these failures promptly. Identifying these failures is very important, but it is challenging, making it difficult to build accurate datasets, leading to a lack of related research. In this paper, CFRP specimens were fabricated and Acoustic Emission Testing (AET) and tensile tests were conducted to collect event waveforms generated during the material failure process. Through deep learning-based representation learning, the characteristics of the failures were compared with those of the materials to build a high-quality dataset. This dataset was then used to classify failure types and verify its quality using a deep learning model. This study contributes to the development of datasets and models for real-time failure detection in CFRP materials. |