Semi-automatic determination of layer depth, permittivity and moisture content for unbound granular pavements using multi-offset 3-D GPR

Abstract

ABSTRACT A semi-automated multi-offset ground penetrating radar (GPR) analysis method called Ray-path Modelling Coherence (RM–C) was developed and field tested. It was used to quasi-continuously predict the depth, permittivity and volumetric moisture content of unbound granular (UBG) layers along a pavement site during two investigations using ground-coupled measurements from a 3-D GPR array. Predictions from the later site visit were compared to physical measurements of layer depth and moisture content at a number of locations along the site, which in general agreed well. RM-C permittivity predictions were compared to results from time domain reflectometery (TDR) sensors and diffraction hyperbola fitting of buried reflectors. The new approach reported higher permittivity values compared to these other methods, although showed similar trends in permittivity change with respect to pavement depth and ⁠variation over time. Reproducibility of the RM–C analysis was confirmed by comparing layer depth and moisture content predictions for repeat runs along the site, although occasional discrepancies were observed.