Ground penetrating radar (GPR) has been widely applied during the recent restoration of the Façade of St. Peter's Basilica to evaluate the condition of conservation. The photogrammetric restitution data have been analysed and integrated with GPR results. A procedure for the interpretation and processing of the experimental data has been developed, supplying a restitution of a 3D geometric model of the structural element studied, in a CAD platform. The GPR surveys generated significant information on the internal structure of the monument concerning, travertine stone, masonry structure, plugs, cramps irons, cavities, detachments, which otherwise would have been obtained exclusively through a destructive analysis. Detailed information on the thickness and the geometry of the travertine blocks and the repaired plugs has been obtained. A total of 4000 m2 of the monument surface has been mapped, including the large pilaster strips, the large columns and the clocks located on the top of the Façade. The extensive application of the non-destructive technique allows a deep knowledge on the condition of the monument.
Non-destructive investigation techniques can be use in a conservative restoration of a great architectural structure, in order to limit the invasive interventions. Besides, they are also effective to acquire information on the state of the monument and to define to the best the necessary interventions.
Among the non-destructive techniques, the georadar shows the most interesting potentiality. GPR has been successfully applied in civil engineering, environmental, archaeological and geological studies. The GPR technique can be used in a lot of applications:
It is particularly effective in the investigation of not conductive material and detection of metallic components in the same. In comparison to the sonic tomography, the georadar shows more versatile and effective, when it's difficult access to structure in examination. In spite of the his sure potentiality, the use of the georadar in the cultural heritage has been till now limited to applications on reduced surfaces by using transducers to low frequency.
- the measure of the thickness of road layers and of iced lake;
- the assessment of reinforced concrete ceiling of the tunnels;
- the investigation on geological formations, on archaeological sites and the mapping of mined field.
GPR uses electromagnetic waves sent on the surface of the object in examination by using an antenna (hold the transmitter and receiver), moved on the surface in controlled mode. The antenna receives the reflex signal (echo). The interference phenomenon of the radar wave in a material is related to the propagation speed of the same, which is dependent upon the physic characteristic of the material. Generally, the propagation speed of the wave is influenced from the dielectric constant and from the magnetic susceptibility of the material. The attainable maximum depth from the radar impulse is function of the frequency of the same and of the resistivity of the material. Greater is the frequency; smaller is the reached depth. During the propagation in the material, the radar wave suffers an attenuation, whose entity is directly proportional to the electric conductivity of the material same. When structural discontinuities are present (for example, fractures, hollow or metallic components), the wave is partially reflected and that produces secondary impulses. Such impulses are recorded and, if correctly analysed, they allow going back to the related structural characteristic. Note the dielectric constant of the material, it can be determined the depth of the reflecting interface. The choice of the operating frequency has to take into account of mechanical and electric characteristics of the material, of the spatial resolution of the application and of maximum depth of the investigation. High frequency antennas (> 500 MHz) supplied high spatial resolution but restricted penetration depth. They are useful to investigate modest thickness. On the contrary, low frequency antennas allow a superior penetration, with a penalisation, yet, in term of spatial resolution. The result of a GPR survey is a radargraphic map, in practice a section of the object. One of the dimensions represents the line along which the antenna has been moved and the other defines a temporal interval (flight time of the wave). If the propagation speed of the radar wave in the material is known, it may be transformed in a spatial dimension (depth). The times of investigation are relatively limited and that allows the structural inspection of elevate surfaces.
The examined surfaces are highlighted on the photogrammetric model of the façade as shown in Figure 1. The mapped surface is of about 4000 m2 on the ~ 7000 m2 of the whole façade. The survey has concerned all the great architectural elements of the façade (pilasters, columns, and clocks) and the zones with great cracks; in practice all that is critical in terms of its maintenance. These investigations have furnished a series of data relating to the geometry of the base and the body of the façade.
Fig 1: Photogrammetric model of the façade of the St.Peter's Basilica with the indication of the external and inside inspected zones by using georadar technique.
The survey has been carried out using the georadar instrumentation SIR 10B of the Geophysical Survey Systems (GSSI). The instrumentation is constituted from a display unit and experimental condition programmer, a control unit and acquisition, a data recording unit and a 12 V electric supply. The transducer is connected to the unity of control by an opportune cable of about 30 meters. A wide frequency range of the radar transducers 200 MHz - 1.5 GHz has been used. Mainly an antenna from 900 MHz has been employed. Afterwards an antenna from 1.5 GHz has been employed to get greater spatial resolution in the more superficial part (up to 70 cm). Instead, an antenna from 400 MHz has been used to carry out a greater depth of analysis (up to 2.5 m). A suitable mechanical system to move the high frequency antennas has been planned. The experimental conditions of acquisition have been selected on the base of a screening performed on demonstrative materials (travertine stone, stucco and iron components), in order to simulate block dimensions, crack orientation and detachments. Radar measurements have been also carried out on demonstrative models to calculate velocity and dielectric constant of the constructive materials (dielectric constant for travertine stone is er = 8). Typically, the investigated depth has been among 0.5 and 3 meters. For the investigation performed to the ground of the Portico has been employed an antenna from 200 MHz, whose inspection depth reaches the 10-12 meters. The scanning speed used has been of about one meter/minute. Figure 2 shows a typical operation on the scaffolding of the façade.The survey on the four great columns (mean diameter of 2.5 m) has been of the all innovative, treating of a cylindrical geometry of remarkable dimension. The scanning have been performed using the antenna from 900 MHz (depth investigation 1.3 m), moving it on the surface of the column along three vertical lines at 0°, 90° and 180° with respect to the surface of the façade. The scanning along the vertical has been effected on a total of 18 meters (the height of the columns is of about 22 meters). Figure 3 shows the adopted scheme of measure for the four great pilasters, characterised from a scanning in horizontal direction with a 20 cm step of altitude. In this way, about 130 sections for each pilaster have been obtained. The mapping on the mosaics of the clocks "Italian" and "Oltramontano" has been carried out using the antenna from 1.5 GHz, performing an horizontal scanning on the 4 quadrants and reaching an inspection depth of about 70 cm. An experimental data processing and interpretation procedure that allows restitution in the three-dimensional geometric model of the investigated parts in CAD environment has been developed. Program software (GSSI) for analysis of the radar raw data has been used. The first step concerns the identification of the temporal origin of the radar structures (horizon of the surface), because from that depends the measure of the depth. The result of the elaboration and interpretation of the data consists in bidimensional stratigraphic sections that describe the structural characteristics of the object. Normally, processing techniques have been applied in order to correct physical and instrumental effects. In the interpretation of the radar data, the information available from the photogrammetric restitution and that supplied by visual inspection have been used. The photogrammetric model in CAD environment (AutoCAD platform) has been used to plan the insertion both of radargraphic map (bitmap images) and of examined structural section (vectorial images). The analysis procedure has been optimised by using particular software and the semiautomatic management of the data in some point of the process of elaboration has been performed. A synthetic model of the examined solid with relative animation has been also obtained by using 3D Studio software.
Fig 2:Execution mode of the georadar survey on a column with GSSI antenna from 900 MHz supplied with mechanical system for cylindrical scanning.
Fig3: Experimental procedure of the georadar survey on a pilaster. Detail of a section, on which the scanning on the three accessible in general of the pilaster has been performed .
Results and Discussion
As well-known since 1700, on the left side of the façade are present four great cracks, whose formation dates back just after its construction, that is related to the low mechanical characteristics of the left geological site of the monument.
The information obtained by the radar survey has permitted to define the typology and extension of the lesion, the superficial and internal width of the orientation in comparison to the surface of the façade. The mapping of such cracks is shown in Figure 4. These four great cracks orthogonally cross the surface of the façade and the three principal levels of the architectural body (portico, loggias and attic).
Fig 4a,b: Graphic representation of the development of the four great cracks on the Façade (red lines, a). Radar sections related to II and III cracks on the large window placed on the attic level (b).
The study has concerned the four great pilasters that are situated by the sides of the façade (two on the right and two on the left). The radar sections on the three investigated sides of the pilaster have defined the geometry of the investigated domain from the surface until depth of 1.3 m (Figure 5). These sections are explained in finishes thematic (travertine, structures building, cavities, detaches, fractures, dowels, metallic cramps). Detailed information on the thickness (40 - 100 cm) and on the geometry of the travertine block has been obtained. The masonry behind the block, manufactured with tile, tuff, mortar and pozzolan (as underlined across a limited core boring), is resulted rather irregular and characterised by numerous cracks and detaches from the stony covering. There is a good agreement between the georadar and the photogrammetry results: both pointed out the distribution and typology of the fractures and the position of the applied stony plug during the precedent restoration (1985-86). From the three-dimensional CAD model of the whole pilaster and applying to every structural element a fictitious material, a synthetic model of the solid with relative animation has been carried out (total number of inclusive frames between 500 and 700).
Fig 5: Experimental procedure of the georadar survey on a pilaster. Scheme of the georadar data processing for a pilaster portion, from the bidimensional section since the three-dimensional graphic representation of the materials.
Four great columns located on the left side of the façade have been examined (Figure 1). The survey has supplied with detail the structural characteristics of these architectural elements. The columns are constituted from travertine block of thickness between 50 and 105 cm, which cover a central masonry build with similar characteristic to the pilasters. Detachments between stony blocks and masonry, fracturing in the central body and located metallic elements (cramps) have been observed (Figure 6). The presence of cramps is due to fixing the blocks. The lesions in the masonry with the detachments of the stony blocks are frequently detected in correspondence of the inferior part of the columns. That is reasonably related to the past displacements of the structure of the façade as found by the data analysis supplied from the photogrammetric survey. The displacements have been measured and they are present both toward in south place of the façade and toward the plaza. In Figure 7 are shown the vertical georadar maps and the correspondent internal outlines of the travertine blocks.
Fig 7: Radargraphic vertical maps and profile of the internal structures of a column (A), 3D-CAD of the travertine block profiles of the same column with 3D view (B).
Fig 6: Experimental methodology for georadar survey on a column. Scheme of the data processing and interpretation for a column portion, since the vertical bidimensional section of the materials.
The georadar surveys of the clocks have been carried out and structural informations with spatial resolutions of the order of few centimetres have been obtained. These results integrate the thermography surveys. Three layers have been identified:
The peperino slices have been applied on masonry substratum (the depth of investigation is of about 70 cm). The Figure 8, relative to a quadrant of the Oltramontano clock, shows a radargraphic map and the corresponding CAD section. The constituent materials and the structural criticality (detachments, cavities), corresponding to the tesserae/mortar and mortar/peperino stone interfaces are also shown. The survey has underlined the different maintenance condition of the two clocks: Oltramontano clock shows higher criticality than Italiano clock, indeed about the 50% of the tesserae of the mosaic are detached by the mortar substratum. The results have permitted to direct opportunely the intervention of restoration. The photographic survey during the restoration stages has fully confirmed the conclusions of the georadar and thermographic surveys.
- plot mosaic, with thickness of 1.5-2.0 cm,
- mortar, with thickness of 1.5-2.0 cm,
- peperino stone, with thickness of 40-42 cm, fixed by metallic pivots.
Fig 8: Radargraphic map and relative CAD section interpreted of a quadrant of the Oltramontano clock, which show the launch constituent materials and the structural criticality checks (left). Map of the detachment at the tesserae/mortar interface (right). |
The georadar methodology, from the survey since the three-dimensional graphic representation in CAD environment, exhibits an advanced non-destructive analysis tool and it is able to provide information on monumental structures with high spatial resolution. The process of three-dimensional modelling of the structural data carried out from the georadar survey and integrated in CAD environment has been optimised until the realisation of a synthetic model of the materials. The extensive application of the GPR in the restoration of the Façade of St. Peter's Basilica has permitted to increase the documentation on the maintenance condition of the monument. Besides, the comparison with the photogrammetric restitution data has been successfully performed. At last, the georadar survey has provided useful elements to the knowledge of the actual static condition of the Façade and the planning of corrective interventions.
Eni Spa has funded this work. We are very grateful to Mrs. A. Foschini and Mr. G. Troiani from Fabbrica di San Pietro for their technical and in field support.
- D.J. Daniels, Surface radar penetrating, Journal of Electronics & Communication Engineering 8, 1996, 165-182.
- J.H. Bungey S.G. Millard , Radar inspection of structures, Proc. Instn. Civ. Engrs Structs & Bldgs 99 (5), 1993, 173-186