Estimating moisture changes in concrete using GPR velocity analysis: potential and limitations

Abstract

Chloride-induced corrosion of steel rebars is a key problem for the durability and safety of reinforced concrete buildings such as bridges. Penetrating rainwater transports the chlorides into exposed parts of these buildings. Hence, the characterization of moisture conditions and their tempospatial variability is a fundamental part of an on-site practical investigation. Ground-penetrating radar (GPR) is a common tool for efficient non-destructive imaging of small-scale structural defects in concrete. When evaluating constant-offset GPR data, the analysis of diffraction hyperbolas yields quantitative information on GPR velocity, and thereby on the water content of the medium. However, when performing a thorough velocity analysis to estimate moisture content, precise information on time zero is a key problem. In this study, a GPR monitoring experiment has been performed under laboratory-like conditions across a reinforced concrete specimen. We show the results of a typical processing flow providing a highly-resolved structural image. Furthermore, we develop and apply a migration-based velocity and time-zero analysis and calculate changes in moisture content. We compare our results to independent measurements of concrete moisture to evaluate the potential and limitations of GPR for estimating tempo-spatial changes in concrete moisture content.