Impact of Overcut on Interaction Between Shield and Ground in the Tunneling with a Double-shield TBM

Transcription

1 Rock Mech Rock Eng (2016) 49: DOI /s x TECHNICAL NOTE Impact of Overcut on Interaction Between Shield and Ground in the Tunneling with a Double-shield TBM Rohola Hasanpour 1 Jamal Rostami 2 Yılmaz Özçelik 1 Received: 20 November 2013 / Accepted: 15 August 2015 / Published online: 23 August 2015 Springer-Verlag Wien 2015 Keywords Double-shield TBM Overcut Shield jamming Squeezing ground 3D numerical simulation Thrust force 1 Introduction Double-shield TBMs (DS-TBM) are among the most technically sophisticated excavation machines in use by tunneling industry. The use of shields around the TBM allows the machine to pass through weak grounds and adverse geological conditions. However, there are limitations in applicability for DS-TBM in some ground conditions where large deformations are anticipated. The presence of the shield limits access to the tunnel walls for observation of ground conditions. This means limited possibilities of observing and analyzing ground conditions to avoid certain problems. Similarly, the presence of the shield does not allow the intrusion of the ground into the tunnel envelope, which is the main objective of using a shielded machine in the first place, yet it also creates the possibility of ground pressing against the shield. In such conditions, TBM may get stuck (including shield jamming & Rohola Hasanpour roha93@gmail.com Jamal Rostami rostami@psu.edu Yılmaz Özçelik yilmaz@hacettepe.edu.tr 1 2 Department of Mining Engineering, Hacettepe University, Beytepe, Ankara, Turkey Department of Energy and Mineral Engineering, Pennsylvania State University, University Park, PA, USA and cutterhead blocking) in complicated geological structures, especially under high ground cover or in weak rocks, where large convergences are expected. This could cause major delays and impose a heavy and expensive burden on the tunneling operation. Some of the issues related to application of DS-TBMs in squeezing ground have been discussed in Hasanpour (2014) and Hasanpour et al. (2014a, b) and some possible scenarios and concepts for mitigating the related problems are offered. There are several performance parameters that should be considered with high accuracy at the design stage of a TBM for preventing machine entrapments. Size of the annular space or gap between ground and shields (created by overcut), length and diameter of shields, thrust force and torque, and machine advance rate are the most important performance parameters in tunneling by a shielded TBM. However, selecting the correct overcut, compared to other performance parameters, has a significant impact on preventing shield jamming. Selecting an appropriate or optimum value for overcut at the design stage of DS-TBM tunnel and implementing the predetermined overcut is the easiest way to address machine jamming, with the possibility of adjustment along the tunnel by using movable gage cutters. The adjustments can be directly related to ground properties and optimized to reduce the risk of machine jamming, while minimizing both the amount of material that is excavated and hauled out of the tunnel and the amount of grout that is placed behind the segments. For preventing the shield seizure, increasing the annular gap between the rock and shield is often utilized at the machine design stage. This feature can be included in the design of the cutterhead to accommodate a given overcut as a base design, and as needed, the excavated diameter of the tunnel, and hence the gap above the shield can be increased to react to bad ground where large convergences are

2 2016 R. Hasanpour et al. expected, or in cases when longer and normal machine delays are planned. Furthermore, the stepwise increase of the annular gap by decreasing the diameter of the rear shield relative to the front shield is another solution that is incorporated in machine design and can be observed in all double-shield TBMs manufactured in recent years. In order to choose a suitable value for the overcut, more comprehensive and detailed three dimensional numerical analysis of the ground is needed at the design stage of a double-shield TBM. This is due to the fact that ground convergence and hence the pressure imposed upon the shield and thus the thrust force needed to propel the shield is a function of the complex interaction between the rock mass, the tunneling machine, its subsystems and components, and the final tunnel support in tunneling by a shielded TBM. Therefore, three dimensional models that include all of these components are essential for modeling the interactions correctly and avoid the errors created by assumption of plane strain conditions or axisymmetric modeling. There are a number of studies in the literature that related to the numerical analyses of mechanized tunneling in squeezing conditions (Lombardi and Panciera 1997; Einstein and Bobet 1997; Graziani et al. 2007; Sterpi and Gioda 2007; Wittke et al. 2007; Ramoni and Anagnostou 2006, 2007, 2008, 2010, 2011; Amberg 2009; Schmitt 2009; Zhao et al. 2012). The developed model in this study enhances the computational model of Zhao et al. (2012) and describes a comprehensive 3D modeling used for simulation of the DS-TBMs for excavation of long deep tunnels through rock masses that exhibit squeezing behavior. The model has some properties that distinguish it from other 3D models that have been developed for numerical simulation of shield TBMs in the past. Finite difference analysis with large strain assumption was used in this paper for numerical computation. Moreover, the results of analysis were presented in simple 3D and geometrically correct shapes that are practical for engineering applications. The model estimates tunnel convergence during excavation and predicts the loads on the shields at various time steps, corresponding to short term response of the ground (Hasanpour et al. 2014a, b). In addition, to avoid errors in the analysis due to large displacements in weak grounds, the method of displacement control has been applied to represent contact surfaces between the ground and the shield. For this purpose, a FISH routine was developed in FLAC 3D that controls all displacements with respect to non-uniform overcut at each solving step of numerical analysis. Increasing of gap due to conical shape of the shield is also considered in the geometric setting of the model and in the contact detection code. These properties of model distinguish the simulations used in this study from other 3D models that have been developed for numerical simulation of shield TBMs in the past (Hasanpour 2014). A numerical study on the impact of changes in overcut was subsequently carried out to allow for observation of impact of overcut on the possibility of machine jamming in squeezing ground. The results demonstrate that the contact forces on shields are considerably lower when a larger overcut is provided. But, in squeezing ground, it leads to a bigger plastic zone and instabilities above shields and lining and creates higher dead weight loads that could cause jamming of the backup system or failure of the concrete segmental lining. On the other hand, the smaller overcut will result in high contact forces, leading to a smaller plastic zone; however, shields may be entrapped due to high frictional forces. Design of overcut should be performed by accounting for the above mentioned phenomenon. In this paper, the results of 3D numerical modeling of rock mass and TBM components are discussed with reference to previous research work by Hasanpour et al. (2014a, b) and Hasanpour (2014). The output of the modeling includes longitudinal displacement and contact forces and also sectional ground pressure on shields versus different amount of overcut. Furthermore, the size of plastic zones for four different values of overcut and the required thrust force to overcome frictional forces when shield is in contact with the rock mass are investigated. 2 Numerical Modeling The study presented in this paper is based on 3D numerical modeling and simulation of ground response when using a double-shield TBM and large ground convergences are anticipated. The current part of the study is an extension of the previous work by the authors on this topic (Hasanpour 2014; Hasanpour et al. 2014a, b). The current study builds upon the 3D models previously used by changing the overcut values for evaluation of the impact of overcut on contact forces between the shield and the ground. Rock mass parameters, geometric dimensions, and mechanical properties for DS-TBM components are given in Tables 1 and 2. Various excavation stages are programmed as several steps in the numerical model. These stages were implemented based on the design and excavation diameter of the cutterhead, diameters of the front and rear shields for a given double-shield TBM, and incorporating the time steps. In this study, a total of 41 excavation steps were simulated including the initialization step 1 and 40 excavation steps (each excavation step is modeled by advancing of the face by 1 m).

3 Impact of Overcut on Interaction Between Shield and Ground in the Tunneling with a Double 2017 Table 1 Rock mass parameters and geometric dimensions for DS-TBM components DS-TBM components Rock mass parameters Cutterhead length (m) 0.75 Elastic modulus, E (GPa) 1.40 Front shield length (m) 5 Poisson s ratio, m ( ) 0.25 Rear shield length (m) 6 Cohesion, c (MPa) 0.60 Shield thickness (cm) 3 Friction angle, / ( ) 28 Lining segment width (m) 2 Dilatancy angle, w ( ) 8 Lining segment thickness (cm) 45 Tunnel inner diameter (m) 8.10 Table 2 Mechanical properties of DS-TBM components Material properties Unit Shield Segmental lining Soft backfill Hard backfill Elastic modulus GPa Poisson s ratio Unit weight kn/m The contact between the TBM main shield and the rock mass was modeled by interface elements and considering the gap between the ground and the shield, with a nonuniform overcut in the shielded TBM. This allows for accurate simulation of the overcut which is maximum at the crown and zero at the invert, where the TBM shield slides on the invert by its own weight. Normal and shear stiffness values (k n, k s ) were assigned to interface elements for simulation of interaction phenomena. k n and k s values of shields were selected to be ten times the equivalent stiffness of the softer neighboring zone: K þ ð4=3þg ; DZ min where K and G are the bulk and shear moduli of the ground, respectively, and DZ min is the smallest width of an adjacent zone in the normal direction, equal to 1 cm (FLAC 3D manual, 2006). k n and k s are calculated to be 2.7e13 Pa/m according to the above-noted relationship. Advance rate of machine was taken into account by controlling of the unbalance forces that are created after each excavation step during face advance. For this purpose and in order to correctly represent the continuous excavation by a shielded TBM and considering the advance rate of machine into step-by-step analyses, an 87 % relaxation of the unbalance forces for each step of analysis was used (Zhao et al. 2012). 3 Results of Numerical Modeling The modeling results are illustrated in the following by giving the displacements, forces, and pressure along the lines at the tunnel crown and side-walls. The overcut is defined as the gap between the ground and the front shield. The overcut between the ground and the rear shield is higher due to a stepwise reduction in diameter of the shields. Shield thickness is considered to be 3 cm. The gap between the rock mass and the cutterhead is slightly smaller than that of the front shield when overcut in the crown is 20 cm. For other overcut settings, this magnitude is different and is adjusted with respect to the overcut in the front shield. Figure 1 shows the different overcut settings at the crown. DR is the gap between the front shield and the ground, which is also shown as DR f. Given that the overcut between the ground and shields is non-uniform in the cross section of the tunnel, there are different values of overcut at the tunnel sectional boundaries. This means that the overcut has the maximum value at the crown and gradually decreases to its minimum value equal to zero at the invert. The numerical analyses were performed to capture the non-uniform overcut as introduced in the models. 3.1 Shield Ground Interaction One of the important parameter in DS-TBM design is the stepwise reduction of the shield diameter, thus defining the variation DR of the radial gap along the shield based on the initial excavation diameter defined by the cutterhead. The positive effect of a stepwise construction is reducing the contact forces (which govern the required thrust force) acting upon the shields. Figure 2a, b illustrates the simulation results in terms of the longitudinal contact force profile (LFP), which is proportional to the longitudinal displacement profile (LDP) at tunnel circumference along the tunnel crown and side-walls, respectively. As expected, the contact forces are considerably low (both for the front and rear shields) when a larger overcut is provided. In the case of a very large overcut DR = 20 cm, the gap between the ground and shield closes at and the shield experiences lower contact forces.

4 2018 R. Hasanpour et al. Fig. 1 Illustration of four assumed different overcut sets (1 4) used in numerical analysis The impacts of overcut and stepwise reduction of the shield diameter on decreasing the total ground pressures acting on the shields are summarized in Table 3. The results show that a wide gap is more important for the rear shield, because the convergence of the ground increases with the distance behind the face. For example, the total ground pressure acting upon the front shield decreases from 23.3 MPa where DR = 1 cm to 4.2 MPa at DR = 20 cm, corresponding to a reduction of about 81.9 %. This percentage is about 86.3 % for the rear shield where the ground pressure reduces from 19.6 to 2.7 MPa. 3.2 Comparison of the Plastic Zones for Different Depths of Overcut A comparison of the plastic zone around the tunnel for different sizes of overcut has been performed to evaluate the impact of depth of overcut on the size of plastic zone. As shown in Fig. 3, the size of the plastic zone for different overcut increases linearly with the size of the radial gap, DR (DR = 1, 5, 10, and 20 cm). The larger overcut requires more time to close the gap but leads to a bigger plastic zone in the longitudinal direction. Therefore, if DR = 20 cm of overcut gap remains open for longer time and takes shield length L for initial contact between the rock and shield, the extent of plastic zone will increase due to additional room to expand and related increased volume of rock involved in ground deformation. Although a larger overcut creates a lower load on the shield and subsequently a lower frictional force during machine advance, however, a larger overcut leads to larger deformations around the tunnel and consequently forms an extended zone of overstressed ground. Thus, the ground is loosened and softened due to large deformation. This creates higher loads above the shield at the crown of tunnel and can cause backfilling problems (Ramoni and Anagnostou 2011). The problems related to loosening and softening of ground in tunnels are particularly important for the design of a yielding support, because both strength loss and major loosening call for a higher yield pressure in the support system (Anagnostou and Cantieni 2007). When an additional dead load is placed against the shield and ground support due to development of extended plastic/weak zones, it may appear as overstressed segments that can fail or develop cracks in various directions. This could explain the ground behavior in the case of tunnels where failed segments were observed after passing of the machine. One example is the T26 Tunnel (Istanbul-Ankara high speed rail project) in Turkey which has experienced serious failure in the segments and backup jamming, and sometimes resulted to the entrapment of the shields (Hasanpour and Rostami 2013). It should be noted that proper placement of backfill and grouting is essential for uniform redistribution of pressures around segmental lining. This issue has direct implication on the design of the lining system and optimizing the amount of backfill/grouting around segments. Applying larger overcut to relieve the shield loading and reduction in the required thrust force for propelling the TBM in bad ground can be considered as a reasonable solution. However, the amount of overcut should be reduced in good ground where it could result in the need for higher amounts

5 Impact of Overcut on Interaction Between Shield and Ground in the Tunneling with a Double 2019 Fig. 2 LDP and LFP for a 1-m length of cutterhead, 5 m front shield and 6 m rear shield at different overcut sets, DR, of 1, 5, 10 or 20 cm a at the crown b at side- wall

6 2020 R. Hasanpour et al. Table 3 Ground pressure acting upon machine components Overcut, DR (cm) Ground pressure acting on (MPa) Pressure reduction comparing to 1 cm overcut (%) on Cutter head Front shield Rear shield Cutter head Front shield Rear shield Fig. 3 Plastic zone for a 1-m cutter head, 5 m front shield and 6 m rear shield at variable DR a 1 cm, b 5 cm, c 10 cm, d 20 cm of material for backfill/grouting that means increased cost of grouting and more expensive excavation of tunnel on a per foot basis. 3.3 Thrust Force Calculations The thrust force required to overcome shield skin friction can be calculated by integrating the contact pressure over the shield surface and multiplying the results by the skin friction coefficient. Sectional contact pressure profiles, between the ground and front shield as well as rear shield, are shown in Fig. 4a, b. Figure 4c, d also shows the required thrust force to overcome the frictional forces on the shield versus the overcut for the two operational stages including ongoing excavation and restart after a standstill. The skin friction coefficient was assumed to be l = for ongoing excavation and l = for restart after a standstill, where the lower friction coefficient values aim to illustrate the positive effects of lubrication of the shield extrados, e.g., by bentonite or other lubricants (Ramoni and Anagnostou 2010).

7 Impact of Overcut on Interaction Between Shield and Ground in the Tunneling with a Double 2021 Fig. 4 Sectional contact pressure profile between ground and a front shield, and required thrust force for the ongoing excavation, l = 0.25 and b rear shield restart after standstill, l = c Amount of propel force needed to move the machine forward in ongoing excavation and d in case of restart after long delay or stand still 4 Conclusion A comprehensive 3D modeling study of mechanized excavation by a DS-TBM was performed to allow for the assessment of the ground shield interaction at various points along the shield for various overcut size. Numerical analysis was used in the simulations to evaluate the magnitude of the ground loading on the shield and possible TBM jamming. The effect of a stepwise shape of the shield in reducing the ground pressure acting against it, for different values of the overcut was investigated. The results show that a larger overcut decreases shield loading and therefore would lead to a lower frictional resistance during shield advance. The required thrust force to overcome frictional forces was determined to be lower for increased depth of overcuts. This was true for two operational cases of ongoing excavation and restart after long delay or standstill. The results also show that a higher overcut can lead to a larger plastic zone and increases the possibility of experiencing instabilities above the shields and increased loading against lining by imposing higher values of dead weight loads on the shield. This could cause failure of the segmental rings. On the other hand, smaller overcut results in high contact forces, leading to a smaller plastic zone. However, shields may be entrapped due to high frictional forces resulting in high contact forces. The results indicate that a larger overcut is not by itself a solution for coping with squeezing conditions and avoiding shield entrapment. Increasing the overcut can be

8 2022 R. Hasanpour et al. considered for preventing machine jamming in squeezing ground in special cases but should be carefully optimized. This paper has described the type of a parametric study that can be used for evaluating the impact of overcut on the shield loading and the extension of plastic zone. This procedure can be used to optimize the overcut size in specific ground conditions and geometry of the shield. For optimizing the amount of overcut, ground properties along tunnel should be determined with high degree of accuracy. This is essential for selecting the input parameters for simulation of the shielded machine. A full 3D numerical simulation of the shielded TBM tunneling should be performed for evaluation of the different rates of overcut on shield loading as well as on plastic zone around tunnel. Appropriate value of overcut can be selected based on a sensitivity analysis of the overcut and resulting ground load on the shield, including the extent of the plastic zone around the tunnel. Acknowledgments The authors gratefully acknowledge the financial support of the Scientific and Technological Research Council of Turkey (TUBITAK) under Project No. MAG-114M568. References Amberg F (2009) Numerical simulations of tunnelling in soft rock under water pressure. ECCOMAS thematic conference on computational methods in tunnelling, EURO:TUN 2009, Bochum, Aedificatio Publishers, Freiburg, pp Anagnostou G, Cantieni L (2007) Design and analysis of yielding support in squeezing ground, The second half century of rock mechanics. In: 11th congress of the international society for rock mechanics (ISRM) Lisbon Taylor & Francis Group, London, 2: Einstein HH, Bobet A (1997) Mechanized tunnelling in squeezing rock-from basic thoughts to continuous tunneling. Tunnels for people, ITA World Tunnel Congress 97, Vienna 2 Graziani A, Ribacchi R, Capata A (2007) 3D-modelling of TBM excavation in squeezing rock masses. Brenner Basistunnel und Zulaufstrecken, Internationales Symposium BBT 2007, Innsbruck, Innsbruck University Press, pp Hasanpour R (2014) Advance numerical simulation of tunneling by using a double shield TBM. Comput Geotech 57:37 52 Hasanpour R, Rostami J (2013) Numerical modeling of tunneling by a single shield TBM. In: UYAK 2013, 3rd international symposium and exhibition on underground excavations for transportation, Istanbul Hasanpour R, Rostami J, Barla G (2014a) Impact of advance rate on entrapment risk of a double shield TBM in squeezing grounds. Rock Mech Rock Eng. doi: /s Hasanpour R, Rostami J, Ünver B (2014b) 3D finite difference model for simulation of double shield TBM tunneling in squeezing grounds. Tunn Undergr Space Technol 40: Lombardi G, Panciera A (1997) Problems with TBM & linings in squeezing ground. Tunnels and tunnelling international no. 29, 6 June 1997, Miller Freeman plc., London, pp Ramoni M, Anagnostou G (2006) On the feasibility of TBM drives in squeezing rock conditions. Tunn Undergr Space Technol 21(3 4):262 Ramoni M, Anagnostou G (2007) Numerical analysis of the development of squeezing pressure during TBM standstills. The second half century of rock mechanics, 11th congress of the international society for rock mechanics (ISRM). Lisbon, Taylor & Francis Group, London 2: Ramoni M, Anagnostou G (2008) TBM drives in squeezing rockshield-rock interaction. Building underground for the future, AFTES international congress Monaco, Montecarlo, Edition specifique Limonest, pp Ramoni M, Anagnostou G (2010) Tunnel boring machines under squeezing conditions. Tunn Undergr Space Technol 25: Ramoni M, Anagnostou G (2011) The interaction between shield, ground and tunnel support in TBM tunneling. Rock Mech Rock Eng 44:37 61 Schmitt J.A (2009) Spannungsverformungsverhalten des Gebirges beim Vortrieb mit Tunnel bohr maschinen mit Schild. Heft 89 Dissertation, Institut für Grundbau und Bodenmechanik. Heft 89-Dissertation, Institut für Grundbau und Bodenmechanik, Technische Universität Braunschweig Sterpi D, Gioda G (2007) Ground pressure and convergence for TBM driven tunnels in visco-plastic rocks. ECCOMAS Thematic conference on computational methods in tunnelling, EURO:- TUN 2007, Vienna. University of Technology, pp Wittke W, Wittke-Gattermann P, Wittke-Schmitt B (2007) TBMheading in rock, design of the shield mantle. ECCOMAS Thematic conference on computational methods in tunnelling, EURO:TUN 2007, Vienna, Vienna University of Technology, p98 Zhao K, Janutolo M, Barla G (2012) A completely 3D model for the simulation of mechanized tunnel excavation. Rock Mech Rock Eng 45(4):