PSI - Issue 84

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

311

3. Case studies The proposed method has been applied (Fig. 8) to the assessment of both Global and Local Attention Class (CdA) in two partially lined natural tunnels, located in northern Italy. The rock mass intersected by the tunnel consists of well-bedded micritic limestones. These tunnels are characterized by an alternation of sections with different types (concrete lining, shotcrete, lightweight sheet-metal cladding), while some are left unlined. The initial dataset was derived from a mobile mapping survey carried out using the ARCHITA system, supported by on-site inspections. The orientation of discontinuities (bedding planes and fractures) and the spacing of the discontinuity sets were derived directly from the point cloud, while information concerning joint characteristics (aperture, filling, roughness and persistence) and Schmidt hammer measurements were collected in the field through seven observation windows approximately uniformly distributed along the tunnel length.

Fig. 8. (a) Attention Classes defined for each twenty-meter tunnel section are summarized, (b) Attention Classes of sectors 3, 4 and 5 in the MIRET-Tunnel AI.

The quality of the rock mass was assessed following the approach; for each analyzed segment, we obtained GSI values varying between 45 and 51, corresponding to rock mass class A of the guidelines (rock mass with medium and medium-low geomechanical characteristics) (Foria et al. 2024). The assessment of defects 1.9 and 1.10 was carried out only for the vault and not for the sidewalks. This decision was based on the analysis of the bedding attitude, which, combined with the vertical orientation of the walls and their distance from the roadway, did not indicate any hazardous conditions. The hydraulic conditions were evaluated for each tunnel section based on thermographic data acquired with the ARCHITA system and subsequently processed and analyzed using the MIRET-Tunnel-AI. 4. Conclusion This study presents an innovative approach to the assessment of unlined tunnels, introducing a semi-quantitative methodology. The use of multidimensional mobile mapping systems (ARCHITA) allows advanced data acquisition (photogrammetry, thermal imaging, laser scanning). The acquired data, analyzed with MIRET-Tunnel AI software, is combined with in situ surveys to analyze the parameters required to define the attention class of tunnels. We propose a set of quantitative indicators that support a more rigorous characterization of rock mass defects and tunnel conditions. This framework could contribute to a more consistent and objective definition of the Hazard and Vulnerability parameters of unlined tunnels, informing the classification into Global and Local Attention Classes. References

Bergeson, W., Ernst, S., 2015. Tunnel operations, maintenance, inspection, and evaluation (TOMIE) manual. Federal Highway Administration, Washington, DC, Report No. FHWA-HIF-15-005.