The Volume Loss: Real Estimation and Its Effect on Surface Settlements Due to Excavation of Tabriz Metro Tunnel

• Published in: Geotechnical and geological engineering 2020-06, Vol.38 (3), p.2663-2684
• Main Authors: Rezaei, Amir H., Ahmadi-adli, Mohammad
• Format: Article
• Language: English
• Subjects: Boring machines; Case depth; Case studies; Civil Engineering; Computer simulation; Dredging; Earth and Environmental Science; Earth Sciences; Empirical equations; Excavation; Finite element method; Geotechnical Engineering & Applied Earth Sciences; Ground motion; Groundwater table; Grout; Grouting; Hydrogeology; Mathematical models; Mohr-Coulomb theory; Original Paper; Simulation; Subway tunnels; Subways; Terrestrial Pollution; Tunneling; Tunnels; Waste Management/Waste Technology; Water table
• ISSN: 0960-3182; 1573-1529
• DOI: 10.1007/s10706-019-01177-5

A case study including monitoring data of surface settlements induced by tunneling in Tabriz metro line 2 is presented. This study focuses on about 4.6 km EPB-TBM tunneling works from west depot to S5 Station, where the tunnel was being excavated in sandy and fine grained alluvial, generally below the water table. Although volume loss ( V L ) has an important effect in estimating the ground movements, the prediction of optimum V L relies on the data from previous case studies with similar ground condition. Realizing unavailability of V L data based on Tabriz geological conditions, from the actual occurred settlements, V L was back-calculates by Gaussian empirical equation to get the V L percentage. Also a back-analysis was done to obtain settlement trough width constant K in both longitudinal and transverse sections. Furthermore, face support and tail grouting pressures were monitored, and compared with COB method. Likewise, a 3D finite element multi-step simulation was carried out by ABAQUS to predict surface settlements. There was concluded that, in longitudinal sections, numerical simulation is not capable of reproducing the recorded S max , applying Mohr–Coulomb model. S max can be estimated accurately by Gaussian curves, but the distribution of longitudinal surface settlement doesn’t match with these curves. The obtained V L varies from 0.13 to 1.35%, with an average value of 0.46%. The range of the V L increases significantly with the decrease of the tunnel depth. For the studied case, in shallow tunnels (C/D ≃ 1) high values of V L are occurred despite good control of face and grout injection pressures.