资源描述
外文原文Surface settlement predictions for Istanbul Metro tunnelsexcavated by EPB-TBMS. G. Ercelebi H. Copur I. OcakAbstract In this study, short-term surface settlements are predicted for twin tunnels, which are to be excavated in the chainage of 0 ? 850 to 0 ? 900 m between the Esenler and Kirazl stations of the Istanbul Metro line, which is 4 km in length. The total length of the excavation line is 21.2 km between Esenler and Basaksehir. Tunnels are excavated by employing two earth pressure balance (EPB) tunnel boring machines (TBMs) that have twin tubes of 6.5 m diameter and with 14 m distance from center to center. The TBM in the right tube follows about 100 m behind the other tube. Segmental lining of 1.4 m length is currently employed as the final support. Settlement predictions are performed with finite element method by using Plaxis finite element program. Excavation, ground support and face support steps in FEM analyses are simulated as applied in the field. Predictions are performed for a typicalgeological zone, which is considered as critical in terms of surface settlement. Geology in the study area is composed of fill, very stiff clay, dense sand, very dense sand and hard clay, respectively, starting from the surface. In addition to finite element modeling, the surface settlements are also predicted by using semi-theoretical (semi-empirical) and analytical methods. The results indicate that the FE model predicts well the short-term surface settlements for a given volume loss value. The results of semi-theoretical and analytical methods are found to be in good agreement with the FE model. The results of predictions are compared and verified by field measurements. It is suggested that grouting of the excavation void should be performed as fast as possible after excavation of a section as a precaution against surface settlements during excavation. Face pressure of the TBMs should be closely monitored and adjusted for different zones.Keywords Surface settlement prediction _ Finite element method _ Analytical method _ Semi-theoretical method _ EPB-TBM tunneling _Istanbul MetroIntroductionIncreasing demand on infrastructures increases attention to shallow soft ground tunneling methods in urbanized areas. Many surface and sub-surface structures make underground construction works very delicate due to the influence of ground deformation, which should be definitely limited/controlled to acceptable levels. Independent of theexcavation method, the short- and long-term surface and sub-surface ground deformations should be predicted and remedial precautions against any damage to existing structures planned prior to construction. Tunneling cost substantially increases due to damages to structures resulting from surface settlements, which are above tolerable limits (Bilgin et al. 2009). Basic parameters affecting the ground deformations are ground conditions, technical/environmental parameters and tunneling or construction methods (OReilly and New 1982; Arioglu 1992; Karakus and Fowell 2003; Tan and Ranjit 2003; Minguez et al. 2005; Ellis 2005; Suwansawat and Einstein 2006). A thorough study of the ground by site investigations should be performed to find out the physical and mechanical properties of the ground and existence ofunderground water, as well as deformation characteristics, especially the stiffness. Technical parameters include tunnel depth and geometry, tunnel diameterlinegrade, single or double track lines and neighboring structures. The construction method, which should lead to a safe and economic project, is selected based on site characteristics and technical project constraints and should be planned so that the ground movements are limited to an acceptablelevel. Excavation method, face support pressure, advance (excavation) rate, stiffness of support system, excavation sequence and ground treatment/improvement have dramatic effects on the ground deformations occurring due to tunneling operations.The primary reason for ground movements above the tunnel, also known as surface settlements, is convergence of the ground into the tunnel after excavation, which changes the in situ stress state of the ground and results in stress relief. Convergence of the ground is also known as ground loss or volume loss. The volume of the settlement on the surface is usually assumed to be equal to the ground (volume) loss inside the tunnel (OReilly and New 1982).Ground loss can be classified as radial loss around the tunnel periphery and axial (face) loss at the excavation face (Attewell et al. 1986; Schmidt 1974). The exact ratio of radial and axial volume losses is not fully demonstrated or generalized in any study. However, it is possible to diminish or minimize the face loss in full-face mechanized excavations by applying a face pressure as a slurry of bentonitewater mixture or foam-processed muck. The ground loss is usually more in granular soils than in cohesive soils for similar construction conditions. The width of the settlement trough on both sides of the tunnel axis is wider in the case of cohesive soils, which means lower maximum settlement for the same amount of ground loss.Time dependency of ground behavior and existence of underground water distinguish short- and long-term settlements (Attewell et al. 1986). Short-term settlements occur during or after a few days (mostly a few weeks) of excavation, assuming that undrained soil conditions are dominant. Long-term settlements are mostly due to creep, stress redistribution and consolidation of soil after drainageof the underground water and elimination of pore water pressure inside the soil, and it may take a few months to a few years to reach a stabilized level. In dry soil conditions, the long-term settlements may be considered as very limited.There are mainly three settlement prediction approaches for mechanized tunnel excavations: (1) numerical analysis such as finite element method, (2) analytical method and (3) semi-theoretical (semi-empirical) method. Among them, the numerical approaches are the most reliable ones. However, the results of all methods should be used carefully by an experienced field engineer in designing the stage of an excavation project.In this study, all three prediction methods are employed for a critical zone to predict the short-term maximum surface settlements above the twin tunnels of the chainage between 0 ? 850 and 0 ? 900 m between Esenler and Kirazl stations of Istanbul Metro line, which is 4 km in length. Plaxis finite element modeling program is used fornumerical modeling; the method suggested by Loganathan and Poulos (1998) is used for the analytical solution. A few different semi-theoretical models are also used for predictions. The results are compared and validated by field measurements.Description of the project, site and construction methodThe first construction phase of Istanbul Metro line was started in 1992 and opened to public in 2000. This line is being extended gradually, as well as new lines are being constructed in other locations. One of these metro lines is the twin line between Esenler and Basaksehir, which is 21.2 km. The excavation of this section has been started in May 2006. Currently, around 1,400 m of excavationhas already been completed. The region is highly populated including several story buildings, industrial zones and heavy traffic. Alignment and stations of the metro line between Esenler and Basaksehir is presented in Fig. 1. Totally four earth pressure balance (EPB) tunnel boring machines (TBM) are used for excavation of the tunnels. The metro lines in the study area are excavated by a Herrenknecht EPB-TBM in the right tube and a Lovat EPB-TBM in the left tube. Right tube excavationfollows around 100 m behind the left tube. Some of the technical features of the machines are summarized in Table 1.Excavated material is removed by auger (screw conveyor) through the machine to a belt conveyor and than loaded to rail cars for transporting to the portal. Since the excavated ground bears water and includes stability problems, the excavation chamber is pressurized by 300 kPa and conditioned by applying water, foam, bentonite and polymers through the injection ports. Chamber pressure is continuously monitored by pressure sensors inside thechamber and auger. Installation of a segment ring with 1.4-m length (inner diameter of 5.7 m and outer diameter of 6.3 m) and 30-cm thickness is realized by a wing-type vacuum erector. The ring is configured as five segments plus a key segment. After installation of the ring, the excavation restarts and the void between the segment outer perimeter and excavated tunnel perimeter is grouted by300 kPa of pressure through the grout cannels in the trailing shield. This method of construction has been proven to minimize the surface settlements.The study area includes the twin tunnels of the chainage between 0 + 850 and 0 + 900 m, between Esenler and Kirazl stations. Gungoren Formation of the Miosen age is found in the study area. Laboratory and in situ tests are applied to define the geotechnical features of theformations that the tunnels pass through. The name, thickness and some of the geotechnical properties of the layers are summarized in Table 2 (Ayson 2005). Fill layer of 2.5-m thick consists of sand, clay, gravel and some pieces of masonry. The very stiff clay layer of 4 m is grayish green in color, consisting of gravel and sand. The dense sand layer of 5 m is brown at the upper levels and greenish yellow at the lower levels, consisting of clay, silt and mica. Dense sand of 3 m is greenish yellow and consists of mica. The base layer of the tunnel is hard clay, which is dark green, consisting of shell. The underground water table starts at 4.5 m below the surface. The tunnel axis is 14.5 m below the surface, close to the contact between very dense sand and hard clay. This depth isquite uniform in the chainage between 0 + 850 and 0 + 900 m.Surface settlement prediction with finite element modelingPlaxis finite element code for soil and rock analysis is used to predict the surface settlement. First, the right tube is constructed, and then the left tube 100 m behind the right tube is excavated. This is based on the assumption that ground deformations caused by the excavation of the right tube are stabilized before the excavation of the left tube. The finite element mesh is shown in Fig. 2 using 15 stress point triangular elements. The FEM model consists of 1,838 elements and 15,121 nodes. In FE modeling, the MohrCoulomb failure criterion is applied.Staged construction is used in the FE model. Excavation of the soil and the construction of the tunnel lining are carried out in different phases. In the first phase, the soil in front of TBM is excavated, and a support pressure of 300 kPa is applied at the tunnel face to prevent failure at the face. In the first phase, TBM is modeled as shell elements. In the second phase, the tunnel lining is constructedusing prefabricated concrete ring segments, which are bolted together within the tunnel boring machine. During the erection of the lining, TBM remains stationary. Once a lining ring has been bolted, excavation is resumed until sufficient soil excavation is carried out for the next lining. The tunnel lining is modeled using volume elements. In the second phase, the lining is activated and TBM shell elements are deactivated.When applying finite element models, volume loss values are usually assumed prior to excavation. In this study, the FEM model is run with the assumption of 0.5, 0.75, 1 and 1.5% volume loss caused by the convergence of the ground into the tunnel after excavation. Figures 3 and 4 show total and vertical deformations after both tubes are constructed. The vertical ground settlement profile after theright tube construction is given in Fig. 5, which is in theshape of a Gaussian curve, and that after construction of both tubes is given in Fig. 6. Figure 7 shows the total deformation vectors.The maximum ground deformations under different volume loss assumptions are summarized in Table 3.Surface settlement prediction with semi-theoretical and analytical methodsSemi-theoretical predictions for short-term maximum settlement are performed using the Gaussian curve approach, which is a classical and conventional method. The settlement parameters used in semi-theoretical estimations and notations are presented in Fig. 8.The theoretical settlement (Gaussian) curve is presented as in Eq. 1 (OReilly and New 1982): (1)where, S is the theoretical settlement (Gauss error function, normal probability curve), Smax is the maximum short-term (initial, undrained) settlement at the tunnel centerline (m), x is the transverse horizontal distance from the tunnel center line (m), and i is the point of inflexion (m). To determine the shape of a settlement curve, it is necessary to predict i and Smax values.There are several suggested methods for prediction of the point of inflexion (i). Estimation of i value in this studyis based on averages of some empirical approaches given in Eqs. 26:where, Z0 is the tunnel axis depth (m), 14.5 m in this study, and R is the radius of tunnel, 3.25 m in this study. Equation 3 was suggested by Glossop (OReilly and New 1982) for mostly cohesive grounds; Eq. 4 was suggested by OReilly and New (1982) for excavation of cohesive grounds by shielded machines; Eq. 5 was suggested by Schmidt (1969) for excavation of clays by shielded machines; Eq. 6 was suggested by Arioglu (1992) for excavation of all types of soils by shielded machines. As a result, the average i value is estimated to be 6.6 m in this study.There are several suggested empirical methods for the prediction of the maximum surface settlement (Smax).Schmidt suggested a model for the estimation of Smax value for a single tunnel in 1969 as given in Eq. 7 (through Arioglu 1992):where, K is the volume loss (%). Arioglu (1992), based on field data, found a good relationship between K and N (stability ratio) for face-pressurized TBM cases as in Eq. 8:where cn is the natural unit weight of the soil (kN/m3), the weighted averages for all the layers, which is 19 kN/m3 in this study; rS is the total surcharge pressure (kPa), assumed to be 20 kPa in this study; rT is TBM face pressure (kPa), which is 300 kPa in this study; and CU is the undrained cohesion of the soil (kPa), the weighted averages for all the layers, which is 50 kPa in this study assuming that CU is equal to SU (undrained shear strength of the soil). Allaverages are estimated up to very dense sand, excluding hard clay, since the tunnel axis passes around the contact between very dense sand and hard clay. The model yields 17.1 mm of initial maximum surface settlement.Herzog suggested a model for the estimation of Smax value in 1985 as given in Eq. 9 for a single tunnel and Eq. 10 for twin tunnels (through Arioglu 1992):where, E is the elasticity modulus of formation (kPa), the weighted averages for all the layers, which is 30,000 kPa in this study, and a is the distance between the tunnel axes, which is 14 m in this study. The model yields 49.9 and 58.7 mm of initial maximum surface settlements for the right and the left tube tunnel, which is 100 mm behind the right tube, respectively.There are several analytical models for the prediction of short-term maximum surface settlements for shielded tunneling operations (Lee et al. 1992; Loganathan and Poulos 1998; Chi et al. 2001; Chou and Bobet 2002; Park 2004). The method suggested by Loganathan and Poulos (1998) is used in this study. In this method, a theoretical gapparameter (g) is defined based on physical gap in the void, face losses and workmanship value, and then the gap parameter is incorporated to a closed form solution to predict elastoplastic ground deformations. The undrained gap parameter (g) is estimated by Eq. 12:where Gp is the physical gap representing the geometric clearance between the outer skin of the shield and the liner, is the thickness of the tail shield, d is the clearance required for erection of the liner, U*3D is the equivalent 3D elastoplastic deformation at the tunnel face, and w is a value that takes into account the quality of workmanship.Maximum short-term surface settlement is predicted by theoretical Eq. 13 (Loganathan and Poulos 1998):where, t is undrained Poissons ratio, assumed to be of maximum 0.5; g is the gap parameter (m), which is estimated to be 0.0128 m in this study; and x is transverse distance from the tunnel centerline (m) and it is assumed to be 0 m for the maximum surface settlement. The model yields 23.0 mm of undrained maximum surface settlement.Other parameters of settlement such as maximum slope, maximum curvature and so on are not mentioned in this study.Verification of predictions by field measurements and discussionThe results of measurements performed on the surface monitoring points, by Istanbul Metropolitan Municipality, are presented in Table 4 for the left and right tubes. As seen, the average maximum surface settlements are around 9.6 mm for the right tube and 14.4 mm for the left tube, which excavates 100 m behind the right tube. Themaximum surface settlements measured around 15.2 mm for the right tube and 26.3 mm for the left tube. Higher settlements are expected in the left tube since the previous TBM excavation activities on the right tube overlaps the previous deformation. The effect of the left tube excavation on deformations of the right tube is presented in Fig. 9. As seen, after Lovat TBM in the right tube excavates nearby the surface monitoring point 25, maximum surface settlement reaches at around 9 mm; however, while Herrenknecht TBM in the left tube passes the same point, maximum surface settlement reaches at around 29 mm (Fig. 10). If the construction method applied to the site is considered, long-term (consolidation) settlements are expected to be low, since the tail void is grouted immediately after excavation. The results of predictions mentioned above and observed maximum surface settlements are summarized in Table 5. The methods suggested by Loganathan and Poulos (1998) and Schmidt (1969) connected with Arioglus suggestion (1992) can predict the maximum short-term surface settlements only for a single tunnel. Plaxis finite element and Herzog (1985) models can predict deformations for twin tubes.Herzogs model (1985) yields higher maximum surface settlements than the observed ones. The reason for that is that the database of the model includes both shielded tunnels and NATM (New Austrian Tunneling Method) tunnels, of which surface settlements are usually higher compared to shielded tunnels. Schmidt (1969), along withArioglus suggestion (1992), yields predictions close to observed.Plaxis finite element modeling gives the most realistic results, provided there is correct assumption of volume loss parameter, which is usually difficult to predict. The model provides simulation of excavation, lining, grouting and face pressure in a realistic manner to predict surface and sub-surface settlements. The volume loss parameter is usually assumed to be 1% for excavation with facepressure-balanced tunnel boring machines. The realized volume loss in the site is around 1% for this study.Currently, there is difficulty yet in modeling the deformation behavior of twin tunnels. One of the most impressive studies on this issue was performed by Chapman et al. (2004). However, Chapmans semi-theoretical method still requires enlargement of the database to improve the suggested model in his paper.ConclusionsIn this study, three surface settlement prediction methods for mechanized twin tunnel excavations between Esenler and Kiraz
展开阅读全文