PSI - Issue 64

Leonardo Paris et al. / Procedia Structural Integrity 64 (2024) 2222–2229 Leonardo Paris, Maria Laura Rossi / Structural Integrity Procedia 00 (2019) 000 – 000

2224

3

The experiments involved some particularly emblematic case studies of some of the most famous ancient bridges of the Roman era such as the Ponte Emilio e Fabricio in Rome, the Bridge of Augustus in Narni, the Roman Bridge in Rieti, the Alconetar and Alcantara bridges in Lusitania, Spain (Amadei 1948; Fernandez Casado, 2008; Galiazzo, 2004; Gazzola, 1963; Malizia, 1997; Tagliaferri & Varriale, 2007). 2. Knowledge through modeling The tools and methods of investigation for the knowledge of architectural artefacts have received a notable boost in the last twenty years thanks to the rapid evolution of technologies such as the 3D laser scanner, the new topographical tools (Range-based Modeling) and digital photogrammetry (Modeling based on images). The coordinated use of these technologies, in an integrated digital methodology, allows us to provide accurate three dimensional models of architecture (Bianchini, 2001). The application of survey methodologies in a specific area such as that of ancient bridges has highlighted some peculiarities: the first is that the bridge cannot be surveyed unless in continuity with the environmental context of which it is an integral part; the second is that the bridge has a strong archaeological component and therefore the same survey criteria used for archaeological areas must be adopted (Borghini & Carlani, 2011; D'Amelio & Lo Brutto, 2009; Demeterscu, 2011); finally, the bridge, being in fact also an infrastructural work (we would say today of civil engineering), must also be investigated from the point of view of its structural behaviour. Although the investigation activity must always be understood as a single scientific process of knowledge, from the methodological point of view of the evaluation of the main peculiarities and critical issues, three phases can be distinguished: acquisition, processing and interpretation, each of which presents specific problems. Every significant activity evidently represents a unicum, just as today an object is a unicum. However, the experiments carried out in carrying out our research on some of the most important ancient stone bridges from the Roman era (all unpublished surveys) have made it possible to define a procedural protocol and to codify a methodology that can be used and repeatable in similar examples. The preparation of such a protocol is not to be understood as a simplification and standardization of the detection process, but rather as an aid within a critical path, in order to obtain an optimization of the entire detection and analysis process. The use of digital acquisition techniques, which has become increasingly widespread in recent years - also thanks to greater manageability, cost reduction and the development of software capable of better managing the size and interchangeability of files - leads to a tendency towards objectification of the data acquisition phase. In this phase therefore there is not, or at least there should not be, any level of interpretation or selection of the information relating to the conformation of the detected object. On the other hand, as experienced in all case studies, an acquisition carried out without adequate planning that takes into account environmental conditions and inevitable constraints inevitably leads to a reduction in the quality of the processing of the interpretative models. The model resulting from the acquisition phase, following the application of scientific methodologies (therefore necessarily objective, not subject to interpretation), must be understood as a product in which the degree of correspondence with the detected object depends on the relationship between quality and quantity that must be made explicit by the choices and conditions determined during the acquisition phase (Paris, 2010). In the experiments carried out on the case studies, a recurring difficulty emerged in positioning the laser scanner, due to the strong environmental constraints, in identifying the optimal shooting points capable of guaranteeing maximum homogeneity of the real resolution of the acquired points. Even the distance between the instrument and the surfaces detected, as well as the distance between the different stations, was very often conditioned by orographic impediments and by the presence, for example, of vegetation, with the need to carefully evaluate the use of automatic recognition, necessary for the optimization of the post-processing phase. When the data, appropriately processed, return information, it is necessary to highlight how in the archaeological field the development of interpretative models takes on a specific value. The three-dimensional model is increasingly associated with "traditional" graphic models. In fact, as mentioned, the two-dimensional graphic model very often derives from the three-dimensional model which in turn derives directly from the point cloud. The conversion of the numerical model for points into mesh does not determine any implementation from the metric point of view. On the contrary, it is very likely that the application of mesh surface generation algorithms leads to a loss of data to increase the manageability of the software to the detriment of the quality of correspondence between the model and reality.