PSI - Issue 64

A. Di Benedetto et al. / Procedia Structural Integrity 64 (2024) 2254–2262 Author name / Structural Integrity Procedia 00 (2019) 000 – 000

2256

3

The data were acquired using the Mobile Laser Scanner (MLS) Leica Pegasus TRK500 Neo, in collaboration with C.U.G.RI. (Inter-University Research Center for the Prediction and Prevention of Major Hazards), Leica Geosystems for the survey, and Consorzio Stabile SIS S.c.p.a. for logistics. The travel speed along the road section was approximately 20 km/h. The output consists of a georeferenced point cloud in UTM/RDN2008 coordinate system, with a maximum density of around 5,000 points/m 2 along the MLS trajectory and a minimum density of 1,000 points/m 2 at a distance of approximately 5 meters from the MLS trajectory. The average distance between two scan lines is 2.3 cm.

Fig. 1. Test Case, (a) Top view of the tunnel (baseMap Google Earth); (b) Map of Italy, with a red dot pinpointing the test area; (c) A picture of the tunnel entrance; (d) Perspective view of the point cloud. 3. Methods The proposed methodology aims to analyze two types of distress: water leakages and concrete spalling/swelling on the tunnel intrados model derived from the MLS point cloud. The process involves the following steps: 1. Segmenting the point cloud into boxes according to the road's longitudinal development, with the size of each box determined by two main factors: Firstly, it's largely influenced by practical and managerial considerations. Typically, distress analysis is conducted on fixed-length portions (sample units), such as 5 meters. Secondly, it's influenced by the transformation of the reference system from 3D to 2D. As the longitudinal development of the box increases, border deformations also increase. Therefore, it's advisable to limit box sizes to minimize deformations and ensure accurate roughness estimation. Over-reducing box sizes may decrease deformations but lead to improper roughness estimates due to border effects (De Blasiis et al., 2020). 2. Transforming the point cloud of each box from the 3D reference system (E, N, h) to the plane reference system (x, y) through unrolling. 3. Automated editing of box point clouds. Individual points belonging to the tunnel intrados are extracted using an algorithm developed by the authors and described in De Blasiis et al. (2020). The key algorithm variables encompass the following: the "maxDistance", denoting the distance between the plane and a generic point considered an inlier; the "ReferenceVector", which imposes a constraint on the orientation of the reference plane; and the "maxAngularDistance", serving as a threshold for the angular distance between the normal vector of the fitted plane and another vector. 4. Calculating roughness to estimate concrete cover release. Roughness value Δz is computed as the distance between the point (center of the sphere) and the fitting plane. Positive Δz values indicate points below the plane