Numerical analysis and automatic identification of GPR responses to highway surface & subgrade diseases

This paper deals with modeling the GPR response from usual road diseases by the software tool called GprMax and presents the automatic identification of these disease responses. There are totally six kinds of road diseases being modeled, such as surface crack, base crack, base replacement, base breaking, void, ea al. Simulation results show that a 900MHz antenna has good detecting capability to such diseases, all kinds of diseases could cause corresponding anomaly on the GPR pictures. Generally, the vertical interface with different dielectric properties could cause arc-shaped features, while the level interface could lead to corresponding reflection axis. The little dielectric constant leads to less GPR wave travel time, and vice versa. All the antenna moving space, antenna frequency and water in disease area have influence on the GPR image: little antenna moving space and high antenna frequency could lead to nice resolution; moisture in disease area could increase the reflection signal. After determining the disease feature, the authors exploited a program for disease automatic identification basing on the graphics theory, which proved to be reliable in a field GPR scan.