Detection and discrimination of buried metal and low-metal content
landmines can be accomplished using dielectrometry technology. Dielectrometry
measurements can detect and discriminate both non-metal and metal objects
buried in the ground, whereby the presence of low dielectric constant plastic
landmines in a high dielectric constant sand will decrease the measured capacitance
while a metal landmine would increase the capacitance for most configurations.
Further improvements in measurement sensitivity and object discrimination
is gained from low frequency measurements of terminal conductance as well
as capacitance and from their variation with frequency. Signatures from dielectrometry
frequency spectroscopy measurements can be used to enhance detection sensitivity
and to minimize false alarms. Computer based models can relate sensor terminal
admittance measurements to identification of the size, location, and material
of buried objects. Measurements show good agreement with computer simulations.
Dielectric properties of permittivity and conductivity in dry and moist sand
have been measured with a guarded parallel plate capacitor in the frequency
range of 0.1 Hz to 10 KHz. Sand was found to have a loss peak between 1 and
10 Hz, depending on moisture content, which is a convenient range for dielectrometry
measurements. Since sugar is considered to be an adequate simulant for explosives,
its dielectric properties were also measured as a function of frequency.
Capacitive detection measurements were performed in a laboratory sandbed
using a capacitive landmine detector to be used as a model validation tool
supplied by our industry partner Jentek Sensors, Inc. Long cylinders of
aluminum, solid Teflon, and hollow Teflon were used as surrogate landmines
and were buried just under the sand surface. The hollow Teflon cylinder
could be filled with other materials to simulate the presence of explosive
material or other landmine components. Frequency sweeps and surface scans
were performed with this apparatus.
Three dimensional computer simulations were compared to analogous two dimensional
computer simulations in order to validate the correctness of approximating
3D geometries with 2D geometries to give electrical terminal admittances
on a per unit length basis. Since 3D simulations are very time consuming,
it is advantageous to use the much faster 2D computer simulations and then
to multiply by the length of the object. For our case studies, the 3D simulations
and the 2D simulations and the 2D simulations multiplied by object lengths
agreed to within 0.28% or better. Thus, most computer simulations used the
much faster 2D analysis.
The laboratory measurements were compared to a few three dimensional computer
simulations, but more frequently with two dimension computer simulations
that well approximate experimental configurations on a per unit length basis.
The sensor response was studied with respect to the sensor lateral and vertical
position, frequency of electrical excitation, and surrogate landmine material.
The presence of surrogate landmines is clearly identifiable when the sensor
is 1 cm above the sand-bed level with cylinders just under the sandbed surface.
Preliminary measurements have been done with manual positioning of the landmine
detector but we are in the process of adding stepper motor control to automate
the data acquisition. Dielectrometry spectroscopy data show interesting
signatures of the different tested surrogate landmine materials.