|
Title |
Application of Multi-type Nondestructive Testing Techniques for Evaluating the Internal Deterioration of Bridge Decks
|
|
Authors |
김현수(Hyeonsu Kim) ; 강준구(Jungu Kang) ; 정재현(Jaehyun Jeong) ; 한기표(Gi-Pyo Han) ; 송호민(Homin Song) |
|
DOI |
https://doi.org/10.4334/JKCI.2026.38.1.061 |
|
Keywords |
충격-반향; 위상배열 초음파 탐상; 지표투과레이더; 다중유형 비파괴검사; 교량 바닥판 impact-echo (IE); phased array ultrasonic testing (PAUT); ground penetrating radar (GPR); multi-modal NDT; bridge deck |
|
Abstract |
Bridge deck slabs are highly susceptible to rapid deterioration due to repeated traffic loading and environmental factors such as moisture and chloride penetration. However, early detection of internal damage is difficult through visual inspection alone, and reliance on a single nondestructive testing (NDT) method is insufficient for quantitative assessment. In this study, impact echo (IE), ground penetrating radar (GPR), and phased array ultrasonic testing (PAUT) were applied to deck slab specimens with simulated delamination and disintegration damage. For each NDT method, a machine learning?based deterioration evaluation model was developed. The IE data were converted into 2D spectrograms, while GPR and PAUT data were represented as B-scan images to construct 2D convolutional neural network (CNN) models based on ResNet-50. As a result, the classification accuracies of the IE, GPR, and PAUT models were 80.17 %, 75.32 %, and 80.80 %, respectively. This study identified the detectable damage types and limitations of each NDT method based on their wave propagation characteristics. The findings can serve as a foundation for developing quantitative condition assessment and multimodal fusion?based deterioration prediction models for bridge deck slabs.
|