Monitoring Structural Health of Concrete Structures Using Dielectric Constant Prediction at Microwave Frequencies

This paper introduces methodologies that utilize microwave sensing to predict the electrical properties of concrete structures, incorporating parameters such as dielectric constant and loss tangent to assess their quality for structural health monitoring. The significance of structural health monitoring lies in its role in enhancing the safety, reliability, and longevity of infrastructure. By continuously monitoring the condition of structures such as bridges, buildings, and dams, it enables early detection of defects, deterioration, or damage. In this article, a two-port rectangular waveguide configuration is employed alongside an Artificial Neural Network (ANN) to forecast the electrical properties of M20-grade concrete. The ANN’s training dataset is generated using a commercially available high-frequency structure simulator (HFSS), and the trained network is subsequently validated on laboratory-manufactured M20-grade concrete samples. Tests were conducted on concrete specimens with varying water-to-cement ratios of 0.4, 0.5, and 0.6 under three distinct curing conditions: ambient curing, water curing, and steam curing. The findings highlighted differences in the electrical properties of these concrete specimens, offering valuable insights for evaluating the quality of concrete structures in real-time monitoring applications.