講演情報
[Data_Collection-32]Effect of Differences in the Relative Permittivity of Pavement Layers on the Electromagnetic Characteristics of Cement Concrete Pavements with Subsurface Voids
*Kohko Inaba1, Mizuki Tawara2, Shinichi Watanabe2, Takuya Maeshima4, Hideki Naito3 (1. Nagaoka University of Technology (Japan), 2. NIPPO Corporation (Japan), 3. Tohoku University (Japan), 4. Nihon University (Japan))
キーワード:
Cement concrete pavement、Subsurface voids、Ground-penetrating rader(GPR)、Finite-difference time-domain (FDTD) method、Machine learning
Subsurface voids beneath cement concrete pavements are known to develop as a consequence of the pumping phenomenon. As these voids grow, they may induce deterioration of the concrete slab and compromise vehicle driving safety. Therefore, the early detection of subsurface voids and the implementation of appropriate maintenance measures are essential for ensuring the long-term performance and safety of concrete pavements.Ground-penetrating radar (GPR) is widely employed as a nondestructive testing method for detecting subsurface voids. However, the interpretation of GPR B-scan images relies heavily on human visual judgment, which often results in variability in detection accuracy among evaluators. To mitigate this issue, machine learning techniques have increasingly been introduced to automate void detection and evaluation. In machine learning–based approaches, characteristic features are extracted from large volumes of GPR measurement data using computational algorithms, and these features are subsequently used to assess the condition of subsurface voids.Despite the demonstrated potential of machine learning for subsurface void detection, the influence of pavement layer properties on electromagnetic responses has not been sufficiently investigated. In particular, the effects of differences in the relative permittivity—closely associated with the moisture conditions—of individual pavement layers, including the concrete slab, base course, and subgrade, on the electromagnetic characteristics of cement concrete pavements with subsurface voids remain unclear.In this study, electromagnetic field analysis based on the finite-difference time-domain (FDTD) method is conducted to clarify the relationship between the relative permittivity of pavement layers and the electromagnetic characteristics of cement concrete pavements with subsurface voids. Furthermore, based on the insights obtained from the numerical analysis, this study discusses key features to be extracted and reasonable data acquisition strategies for improving the reliability and robustness of machine learning–based evaluations of subsurface void conditions.
