Numerical investigation of new austrian tunneling method and tunnel boring method for tunnel construction

Abstract

To alleviate traffic congestion and curtail environmental deterioration GoB and JICA in cooperation with the World Bank formulated the ‘Strategic Transport Plan for Dhaka’ which includes establishment of Mass Rapid Transit system (MRT). Among all lines of MRTs, Line 1 (Standard gauge, 28.8km route length) will be the part of the integrated transportation network including the underground (12 stations: Shield tunnel by Tunnel Boring Machine with outer diameter 7m) and elevated (7 stations: PC Box girder and RC pier) rail line systems. Scope of this research is to identify the differences between New Austrian Tunneling Method (NATM) and Tunnel Boring Machine (TBM) methods considering proposed underground tunnel alignment of MRT Line 1 based on 3D numerical analysis. To validate the effectivity of PLAXIS 3D in tunnel modeling, two metro lines, Mashhad Metro Line 2 and Delhi Metro Phase 3 have been chosen and comparison between FLAC 3D and Optum G2 with PLAXIS 3D have been focused as well as with empirical formulas. From the comparative analysis of Mashhad Metro Line 2, it can be seen that O’ Reily & New empirical relation shows closest values (3.91%, 1.66% and 4.81% deviation with FLAC 3D, MC model in PLAXIS 3D and MCC model in PLAXIS 3D respectively). From the comparative analysis of Delhi Metro Phase 3, it can be seen that the vertical surface settlement found from Peck’s formula, Optum G2 abd PLAXIS 3D are 34.10, 29.70 and 31.20mm respectively. In this research, the tunnel depth of MRT Line -1 has been considered not below than 30m considering the deep foundations of surrounding structures. The tunnel depth and diameter of NATM method are kept constant (depth 35m and diameter 7m) as NATM method is not as flexible as TBM method. For TBM method, three types of depths (30m, 32m, and 35m) and three types of diameters (5m, 6m, and 7m) are considered to take account the effect of parameters in settlement values. Both of the methods are modelled in three different types of systems (MC, MCC and HS) to evaluate the appropriate numerical analysis method by comparing the results with established empirical solutions provided by different researchers. As the variation in meshing is found considerable (4% for TBM and 10% for NATM), for saving the computation time and finding the close results, medium mesh is considered. It can be concluded after analyzing the results that the total settlement decreases with an increase in depth of the tunnel (almost 11% decrement for every 5m increment of depth) and increases with an increase in diameter (almost 20% increment for every 1m increment of diameter). MCC model shows relatively precise value to the empirical solutions and best fit shape to Gaussian curve. The average deviated values between the numerical result and empirical result shows that, the O’Reily & New equation is better to be used to predict the transverse surface settlement and the deviated percentages from this equation are 5.38%, 3.84% and 6.39% for MC, MCC, and HS models respectively. Increasing the TBM depth results to increase around 4% in distance of inflexion point from center whereas increasing in radius results to decrease around 5% in inflexion point distance from center of the tunnel. Jacobsz formula for predicting longitudinal shows approximately close value to numerical value (around 2 to 16% for NATM and 3 to 10% for TBM). From comparison of longitudinal and lateral settlement, it can be shown that NATM method shows more settlement (10 to 30% more) than TBM method as it includes blast technique which induce more ground surface variation than TBM machine advancement, especially in soft soil. Therefore, Preferring TBM to NATM for constructing the Dhaka MRT Line-1 is the accurate answer.