Difficulties in Interpreting Ballast Degradation Level Estimates from Synthetic Ground-Penetrating Radar Data

As fine-grained particles accumulate within railway ballast, it becomes more susceptible to differential deformations, which leads to the loss of proper track alignment and an increased risk for car derailment. Methods for estimating the ballast degradation level from low-frequency (<1 GHz) ground-penetrating radar (GPR) measurements exist, but their applicability in a wide range of track foundation conditions has yet to be evaluated. This analysis, based on simulated GPR data, evaluates the sensitivity of these methods to changing ballast moisture contents, ballast thicknesses and subballast material types. The results highlight that small changes to the track foundation, indepedent of the concentration of degraded ballast, significantly alter the attenuation and reflectivity characteristics of the simualted GPR measurements. As such, ballast degraded to a certain level will manifest with different attenuation and reflectivity characteristics, limiting the ability to accurately and reliably detect these changes using GPR. Radar propagation velocities within the ballast are less influenced by changes in ballast depth and subballast material type. However, ambiguous propagation velocities are observed for certain ballast moisture contents and degradation levels; furthermore, velocities can only be calculated when the ballast thickness is known. These results suggest that while under certain circumstances, with additional information on the track foundation conditons, ballast degradation levels can be estimated from GPR data; in general, the complicated nature of GPR signals makes the quantificaion of ballast degradation levels difficult.