PSI - Issue 84

Federico Foria et al. / Procedia Structural Integrity 84 (2026) 304–312

310

For Defect 1.10, (detachment of tabular/stratified rock) the evaluation is based on the analysis of the point cloud data, supported by on-site inspections. This combined approach enables the identification of tabular blocks in the tunnel crown and provides a quantification of their extent. According to our assessment, the extent should be classified as high when it affects more than 51% of the tunnel crown, medium if it affects 21–50% of the crown, and low if it is limited to small portions of the crown or the side walls.

Fig. 6. Workflow of the stability analysis, for the definition of Defect Severity class, G, based on point-cloud data. Discontinuity surfaces are extracted from the 3D point cloud and analyzed through Schmidt projection to identify dominant joint sets. Kinematic stability is then evaluated using the Markland test to assess potential failure mechanisms. Finally, the results are synthesized through statistical analysis to quantify the likelihood of different instability scenarios and support slope stability assessment. In general, from a geological perspective, in cases of tabular block detachment occurring in the crown, the spatial extent of the affected area provides the magnitude of the phenomenon. Furthermore, in the case of naturally excavated tunnels, the detachment of a rock mass affecting more than 50% of the crown generally fully compromises the tunnel’s operational functionality. 2.3.3 Water detection The influence of water is highlighted by the thermographic data analysis using the MIRET-Tunnel AI software (Fig. 7) and confirmed by on-site inspections. Three levels of severity have been identified: DRY conditions (no presence of water detected either in thermographic surveys or on-site), MOIST conditions (evidence of moisture on sidewalls and/or crown, signs of water passage, occasional drips), and WET conditions (widespread dripping and water inflows), that are evaluated for each segment of the tunnel.

Fig. 7. MIRET-Tunnel AI – Water detection. (a) the 2D photogrammetric survey with the mapping of defects associated with the presence of water (in green). (b) the 3D thermographic survey (linked to the photogrammetric survey) with an indication of the thermal gradient.