Settlement Predictions of The Noordtunnel: A Numerical Simulation

More Info
expand_more

Abstract

The Noordtunnel is an immersed tunnel open to roadway service since 1990 in The Netherlands. Over the past thirty years of its operation time, a significant differential settlement behaviour has been observed, and this ongoing settlement potentially imposes safety concerns to the tunnel, such as joint leakage. However, the underlying factor triggering this differential settlement behavior remains unknown. To ensure the tunnel's serviceability, this thesis aims to investigate the underlying causes of the occurring excessive settlement, predict future settlements, and assess its impact on tunnel structural safety within its designed lifetime.

The analysis starts by reconstructing the settlement time history of the Noordtunnel. The process involves determining the most reliable reference point and performing back analysis to estimate the settlement magnitude during the unmeasured period. Further, the settlement history is reconstructed by combining all the settlement data of all the periods. It is found that the settlement history at all immersion joints show a logarithmic trendline, with a maximum estimated settlement of about 94.36 mm occurring at immersion joint 2. Subsequently, the soil profile and geotechnical parameters were determined for the simulation. The provided Cone Penetration Test (CPT)s data and borehole ensure the soil profile depicted in the given situation map. Additionally, in the absence of laboratory data, the Hardening Soil Model (HS Model) and Soft Soil Creep Model (SSC) parameters are estimated based on CPT - NEN Table 2b correlations.

Afterward, the Two Dimensions (2D) Finite Element Method (FEM) simulations were carried out in PLAXIS (a commercial simulation software), while considering the variations of load acting on top of the subsoil. Two types of soil constitutive law were chosen to simulate the settlement in the Noordtunnel: HS Model and SSC Model. The optimum model, which can simulate the settlement behaviour in the field, was selected by aligning the simulation outcomes to the reconstructed historical settlement. The simulation results show that only settlement at immersion joint 2 has the same tendency as the SSC Model, while the other immersion joints tend to have a similar tendency to the HS Model. The simulation outcomes also indicate that excessive settlement at immersion joint 2 occurs due to the soft soil underneath the tunnel and the natural sedimentation on top of it. The soft soil is responsible for at least 20 mm of settlement, while the sedimentation contributes to a minimum of 8 mm of settlement during 30 years of tunnel operation. Subsequently, the sensitivity analysis is conducted to examine how much the simulation outcomes may deviate when accounting for soil variability in the field. Due to the narrow distribution of the reference values, adjusting the most sensitive parameter will only deviate the results by a maximum of 5.48% at immersion joint 2 and 4% at the other immersion joints. These results indicate that the model is robust enough and expected to generate reasonable future settlement predictions. An additional settlement of 15.13 mm of settlement is predicted to occur at immersion joint 2, while 4 to 5 mm of additional settlement is anticipated at the other immersion joints over the tunnel's remaining lifespan.

The differential settlement at the tunnel longitudinal direction has triggered element tilting and further induced compression and decompression to the GINA gasket at immersion joints. It has been observed that while uneven settlement contributes to joint decompression, the external forces acting on the GINA gasket remain considerably lower in magnitude compared to the overall friction force. Therefore, the impact of uneven settlement on the water tightness is generally minimal. Additionally, considering the limitations of the current monitoring procedure, an optimized monitoring plan based on the Distributed Optical Fiber Sensor (DOFS) system is proposed. Finally, future recommendations to improve the current thesis are also put forward.