Nectaria Diamanti

Employing ADI-FDTD subgrids for GPR numerical modelling and their application to study ring separation in brick masonry arch bridges

In realistic numerical modelling of ground-penetrating radar (GPR) and when parts of the computational ndomain need to be modelled in detail, the implementation of subgrids into the conventional finite-difference time-domain (FDTD) mesh could greatly economize on computational resources. A novel alternating-direction implicit FDTD subgridding scheme is used to numerically simulate the GPR responses from delaminations located in brick masonry arches. The heterogeneity of these structures renders electromagnetic signals, which originate from the interaction between the GPR system and the bridge, often complex and hence hard to interpret. Therefore, GPR numerical models nwere created in order to study the attributes of reflected signals from various targets within the structure of the bridge. Results from a range of modelling scenarios are presented. The effect of varying the thickness of faults, their location in brickwork, as well as the effect of water ingress in hairline delaminations on GPR signals, are examined. Moreover, GPR vertical resolution and the presence of lossy brickwork are studied through various numerical models.

A study of GPR vertical crack responses in pavement using field data and numerical modelling

The application of ground penetrating radar (GPR) as a non-destructive technique for characterization of pavement structure on road networks has gained considerable attention during recent years. High resolution ground coupled GPR has the potential to provide important additional information on pavement deterioration, defects and cracks, the last being the focus of this study. Crack geometry and the electrical properties of the pavement surrounding the crack can be quite variable, resulting in often complex and hard to interpret data. Therefore, FDTD numerical modelling has been employed to help understand a range of GPR vertical crack responses observed in a variety of pavements.

Quantifying GPR transient waveforms in the intermediate zone

In this paper we examine the transient electromagnetic field variation around dipole antennas placed on the surface of a half-space. To achieve this we employ three-dimensional (3D) finite-difference time-domain (FDTD) numerical modelling. We have previously shown how antenna height, shielding and ground properties impact the directionality and energy flow. Here, we report how the transient fields around a dipole change their amplitude, shape and frequency content. Further, we demonstrate how the aforementioned attributes change as we move away from the source.

Quantifying GPR responses

Antenna height, shielding and subsurface properties all impact the observed GPR responses. While the basic concepts are generally understood, our long term goal is to develop parameterized models that will provide for quantitative interpretation of data acquired with real systems. Our first step was to develop modelling capacity and response presentation tools to help with development. We are using three-dimensional (3D) finite-difference time-domain (FDTD) modelling. Model results are then presented in a variety of forms. In this paper we compute emitted energy characteristics and display in radiation pattern format for infinitesimal dipoles, resistively loaded dipoles and shielded dipoles. Patterns are computed for free-space and over loss-less half-spaces with various properties as a function of height above the surface. The numerical simulations show the advantage of ground-coupling and the impact of shielding on GPR responses. Further, we demonstrate that total radiated energy is a very effective means of characterizing radiated signal directivity for GPR transient emissions.

Influence of interface roughness and heterogeneities on the waveguide inversion of dispersive GPR data

To investigate the influence of interface roughness and heterogeneous media on the inversion of dispersive GPR data, we performed 3D FDTD modeling with GPRMAX. For both the broadside and endfire source-receiver configurations the responses were calculated for models with different interface roughness and for models that include media with heterogeneities in their dielectric properties. Due to the use of multiple source-receiver offsets to calculate the phase-velocity spectrum, the signal to noise ratio was relatively good and for low interface roughness the medium properties could be reasonably well reconstructed. Only for the largest interface roughness the data contained significant diffractions and the medium properties could not be reliably reconstructed. For variations of the relative permittivity values equal to 10% and 20%, the model parameters could still be relatively well reconstructed.

Investigation of choke-ring structures for ground-penetrating radar

For common GPR antennas such as dipoles, shielding, such as with a metallic cavity, is an attractive way to isolate the antenna from the above-ground environment and reduce unwanted coupling and emissions into the air. However, currents induced on the shield by the antenna can travel around the shield and re-radiate, creating unwanted signal paths and offsetting these benefits. This work establishes the viability of choke-ring structures in reducing these surface currents through shaping of the radiation pattern and demonstrates that good bandwidth and time-domain performance may be achieved through the use of resistive, rather than perfectly metallic, surfaces.