Amirkabir NATM Tunnel- A case study of design challenges in a mega project of tunnel in soft ground

In spite of enhancement in modeling techniques as well as site investigation methods, uncertainty exists seriously in the process of construction of underground structure in soft soils. These uncertainties arise from limited data of geological data, measurement errors, interpolation of spatially geological properties, and extrapolation of results of experiments and natural analogue studies over times and conditions relevant to the project. Robust and optimize design of tunnel support pattern consists many important parameters including advance rate and excavation method. However, quantitative definition of these parameters is difficult because of restricting in site investigation data and uncertainties related to them. Furthermore, erroneous evaluation in these parameters can affect in incorrect projection of tunnel stability or economic loss. The work reported in this paper specifically dealt with Amirkabir highway tunnel with approximately 1.5Km long in each tube (i.e.,north and south tubes).It is one of the important projects excavating below one of the highest traffic region of Tehran, including the difficulty of excavating through a heterogeneous sedimentary basin mainly composed of recent alluvial. Construction has been performed by different methods i.e., Underground excavation, Cut and Cover and Top/Down methods in different sections of the project. Deformations and settlements have been monitored during and after construction in order to avoid unpredictable deformations and as a result any possible failure. Similar to other part of the project, a probabilistic hypothetical elasticity modulus (PHEM) approach has been employed to evaluate the uncertainty in lining design of a horseshoe shape NATM(New Austrian Tunneling Method) tunnel in T4 section of the project.The hybrid model, PHEM, consisting two essential parts is used to evaluate the performance of the system. A MATLAB interface program for generating the ABAQUS-based parametric model in one hand; and a Monte Carlo algorithm (Latin Hyper Cube Sampling method) to simulate the uncertainty existed in the system on the other hand are applied to produce the probabilistic density function of surface settlement of tunnel excavation and lining phases. In comparison with the monitoring data, the numerical results show that the PHEM approach introduced has had an appropriate prediction in surface settlements, improved the classical deterministic approaches.