Author Description

Subsurface drainage features are routinely incorporated in the design of pavement systems as they are believed to increase pavement service life provided that they are installed correctly and maintained. Maintenance, however, is challenging in that location and subsequent inspection of these systems can be time consuming and laborious. With this in mind, some departments of transportation have turned to ground penetrating radar as one means to rapidly locate subsurface drainage features in pavements and thus alleviate some of the cost and complexity of maintaining these systems, but with mixed results.In this context, this study pursued a two-pronged approach to improve GPR-based location of sub-pavement drainage systems, involving: (1) software-based signal processing and (2) modifications of hardware test configurations. From a signal processing perspective, two complementary signal processing approaches were developed in this work. Method 1 involved algorithms that are designed to reduce GPR signal background clutter and noise by taking advantage of the somewhat uniform nature of the strata underlying constructed pavements and to systematically remove anomalous signals. Method 2 focused on enhancing 2-D image quality to facilitate recognition of hyperbolic signal returns indicative of drain detection. From a hardware perspective, field experiments were also carried out in this work to validate the signal processing algorithms and assess the potential for alternative antenna configurations to enhance detection success. Five different antenna configurations were tested in total. When employed in field settings, the signal processing algorithms demonstrated an ability to routinely detect X-drains (shallow depth, PVC) with 2-3 false alarms per successful detection. Similarly, all known K-drains (moderate depth, metal/clay) in the studied field test regions were successfully identified, although each successful K-drain detection was accompanied by a significant number