Radar in non-destructive testing of thin granular dielectric layers

Ground-penetrating radar (GPR) is an electromagnetic probing and imaging method of subsurface which is applied in the non-destructive evaluation of roads and bridges as well. In Finland, the GPR is a commonly used quality control method for the air void content of newly laid asphalt pavements. The air void content has an effect on the long-term performance of the pavement. The GPR method is based on deriving the relative permittivity, or dielectric constant, from the reflected radar signal. The air void content is then estimated from the relative permittivity based on the drilled calibration asphalt core. One of the problems related to the current GPR method is the insufficient depth resolution of conventional, 1-2 GHz systems. The objective of this study was to analyse if GPR, or microwave technology in general, can be applied in the air void content assessment of thin asphalt layers. Furthermore, this study aims to provide an understanding about factors affecting the relative permittivity measurements of heterogeneous, granular and layered media. A 12-18 GHz microwave radar with centimetre scale depth resolution was developed for studying thin asphalt layers. Unfortunately, the necessary solution of increasing the centre frequency produced other problems due to the asphalt granularity. The largest aggregate grains in asphalt are about the same size as the wavelength at employed frequencies. Different laboratory tests were conducted for better understanding the relative permittivity variations of pavements. Resonator based surveys included both the magnetic permeability and relative permittivity measurements of rock specimens. Most of the laboratory studies were free space measurements with the vector network analyzer in the transmission configuration at the 7-17 GHz. Studied materials included asphalt raw materials, drilled asphalt samples and granulated plastic specimens. The relative permittivity results of asphalt pavements and samples showed a wide range of values. It was concluded that the observed variation in permittivity does not simply result from changes in air void content. For example, the real part of the relative permittivity of the studied metavolcanic rock was 5.90-6.15, whereas the granular sample of the same rock type had values 3.17-8.12. Multiple reflections from individual granules can increase the observed permittivity of the studied mixture when the grain size is larger or of the order of the signal wavelength. The most of the permittivity variation is likely due to the mineralogical changes in rock aggregates, and changes in the volumetric portions of aggregates and bitumen in asphalt mix. Results of this study suggest that the current GPR method is not reliable in the air void content evaluation of thin pavement layers.