Mines pose a significant risk for civilians, also after a conflict has ended. Mine removal is difficult, risky and time-consuming. Innovative modeling may speed up the removal process.

Traditionally, mines have been detected by use of metal detectors. However, in a battlefield, there may be many metalpieces that do not correspond to an actual mine. This means that several hundred metal objects must be removed for every actual mine.

This project's goal was to detect mines by first making a mathematical model of all the interaction of radar waves with soil and mines, and use this model for later detection through ground penetrating radar. This model would enable detection of explosives, and not just the metal triggering the explosive.

The soil was geometrically modelled by adding particles (pebbles and stones) of random size, form and position to an average background medium (silt, sand). A model of the mine is embedded by the soil. The physical properties of the soil and mine are represented by their complex dielectric constant. A Finite Element Method called DDA is used to calculate the strength of the radar waves scattered in all direction from this object.

The project developed models describing the behaviour of electromagnetic waves (radar waves) in strongly scattering media (soil). The results from this activity was unique modelling skills with respect to light and EM-wave scattering processes. This also enables SINTEF to improve the designs and test methods of optical systems.

For further information please contact Odd Løvhaugen.  

a)

b)

Simulation of ground penetrating radar signals. Rocks and anti-personnel mine in homogeneous soil background.  Moisture level of the soil may decide whether it is possible to detect the mine or not. In figure a) the frequency signature of the mine is drowned in back-scattering noise from rocks due to low soil moisture level. In b) the signature is clearly visible even in the presence of rocks.