PSI - Issue 84

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

308

MIRET-Tunnel AI is not a commercially sold software; rather, it is part of a risk analysis service involving numerous technical roles and responsibilities. 2.3 Geomechanically characterization The evaluation of the geological and geomechanical parameters is carried out by integrating data from mobile mapping (Lidar point cloud, thermal imaging, GPR, ecc.) with data obtained from in situ analyses performed on observation windows. The geomechanically characterization of the rock mass is integrated into the definition of the attention classes on the base of three main indicators: the rock mass quality assessment, which represents one of the key parameters for defining the global and geotechnical hazard class, the possible collapsing/sliding mechanisms, and the hydrogeological conditions, that contributes to the definition of the global vulnerability class. From an operational perspective, the first two indicators are quantitative and based on objective parameters, whereas the third one follow a qualitative approach. The geological characterization, in accordance with the requirements of the guidelines, shall be carried out for homogeneous sections of the tunnel, typically divided into segments of approximately 20 meters in length. In cases where direct data are not available for each segment either from point cloud analysis or from on-site observations (for example, in tunnel sections with partial lining) it is recommended to use the averaged properties of the two adjacent segments, assuming the same geological conditions. 2.3.1 Rock mass quality assessment The characteristics of the rock mass are expressed in terms of the Geological Strength Index (GSI), to conform to the standard adopted in most of the studies of this type. However, this index is derived from a visual and descriptive assessment of rock mass. Nevertheless, in this study we decided to derive the GSI from the Rock Mass Rating (RMR) of Bieniawski (1989), as we consider the latter to be based on more measurable and quantifiable field parameters. Consequently, it provides a more rigorous and quantitative characterization, relying on the assignment of well-defined parameters. Therefore, after the recognition of the different sets of joints according to their orientation, the RMR can be computed using a calculation matrix that integrates the parameters obtained both from the point cloud analysis and from field survey, in accordance with the International Society for Rock Mechanics (ISRM) guidelines. The parameters considered are: joint spacing, joint characteristics (aperture, filling, roughness, persistence), Uniaxial compressive strength (derived from sclerometric measurements), RQD % (derived from JV – volumetric Joint count). The joints characterization must be assessed in the field through traditional surveying methods, since the available point clouds rarely reach the millimeter, scale resolution that allows the evaluation of these parameters at this level of detail. For the RMR characterization, drained conditions should be assumed, since the influence of water is already considered as one of the parameters used in the determination of the vulnerability attention class. Joint spacing can be estimated either through field-based observation windows or directly from the digitalized point cloud (this approach allows for a quantitative characterization of the discontinuity network only when high-resolution spatial data are available). In absence of borehole data, RQD can be obtained using the Joint Volume Index – JV, from Palmstrom (1974), as measure of the volumetric density of discontinuities; this is inversely proportional to the rock block size (the higher the Jv, the smaller the blocks).