NDT for the diagnosis of modern, historical and archaeological structures

This thesis has been developed with the aim to explore thoroughly potential and limit of the GPR and ERT methods for monitoring heterogeneous structures where different construction materials are combined together. Firstly we analysed the GPR response, in various construction materials related to different modern, historical or archaeological structures. In particular, three real examples were investigated during the thesis, namely: the Pyramid of Caius Cestius, the Passage of Commodus and the Colle Oppio Ninpheum, all in Rome. According to the different types of material and frequency antennas, different GPR responses and therefore dissimilar degree of resolution and of attenuation was obtained. In light of this, the interposition between the surface of the investigated medium and the GPR antenna of a dielectric material (e.g. Plexiglas) was performed in order to improve the resolution. Furthermore, an application of the GPR and ERT methods for monitoring a load test executed on masonry samples was presented. This panels were built up in the laboratory controlled conditions using tuff and bricks (widespread materials employed in Italy for decades for masonry buildings) and also were reproduced in the phase of theoretical modeling. The laboratory samples are reinforced with a conductive fibre fabric, where a high-conductive material (steel wires) is combined together with a dielectric material (basalt fibre). In order to improve the sample-antenna coupling in the presence of conductive reinforcements, a Plexiglas (polimetilmetacrilato - PMMA) plate was added underneath a 2 GHz antenna. GPR data were acquired along profiles spaced 0.1 m apart and ERT measurements were executed on a 0.1 m regular spaced grid with a dipole-dipole array operating in a three-dimensional configuration. GPR datasets were also analysed in non-conventional mode, by means of the picking of the reflection time of the EM wave from the rear face of the wall samples. Results show that GPR and electrical resistivity tomography were both able to detect fractures and weakness zones caused by the load application, even though with a higher resolution for the georadar with respect to the geoelectrical method. Moreover, mapping the GPR data in terms of the dielectric constant and mean absolute amplitude is particularly diagnostic to detect the effective fracturing pattern, after the application of the diagonal load. Therefore, GPR and ERT methods can reduce the degree of uncertainty in the detection of fractures, voids or cavities, with respect to the standard processing, by the combined analysis of radargrams, time-slices and resistivity ERT models. Furthermore, for the GPR laboratory data acquired directly on the reinforced face of samples, it is demonstrated how interposing a layer of dielectric material between the antenna and the structure can substantially improve the antenna coupling and consequently the capability to detect fractures and to reach the rear face of the sample, despite losing resolution in the case of shallow high-conductive layers. Finally, three-dimensional synthetic simulations on the same samples validate the experimental evidences. Therefore, we demonstrate that this approach can be a reliable tool to monitor static load tests and it can be further extended to the whole load cycle (before, during and after the experiment).