Tunnelling and Underground Space Technology

Tunnelling and Underground Space Technology 23 (2008) 210214

incorporating Trenchless Technology Research

www.elsevier.com/locate/tust

Technical note

Numerical analysis of the interaction between twin-tunnels: Inuence of the relative position and construction procedure
F. Hage Chehade a, I. Shahrour
b

b,*

a Libanaise, BP 813 Saida, Lebanon Institut Universitaire de Technologie, Universite canique de Lille, Universite des Sciences et Technologies de Lille, Polytech-Lille, 59 655 Villeneuve dAscq, France Laboratoire de Me

Received 21 August 2006; received in revised form 28 January 2007; accepted 19 March 2007 Available online 4 May 2007

Abstract The development of transportation in large cities requires the construction of twin-tunnels or the construction of new tunnels close to the existing ones. Since, both the relative position of tunnels and the construction procedure aect the soil movement and internal forces in the lining, it is of major interest to study the inuence of these factors on the tunnel design. This paper presents analysis of this issue with a particular interest for the optimization of both the relative position of the twin-tunnels and the construction procedure. For this concern, a parametric study is conducted for the investigation of the inuence of these two factors on the soil settlement and internal forces resulting from the tunnel construction. The paper presents successively the numerical model and then analyses conducted for three congurations of the twin-tunnels: aligned-horizontally, vertically and inclined. It shows that the construction procedure aects the soil settlement and internal forces. The construction of upper tunnel at rst leads to both higher settlement and bending moment. The highest soil settlement is obtained for vertical aligned tunnels, while horizontal aligned tunnels cause the lowest settlement. 2007 Elsevier Ltd. All rights reserved.
Keywords: Bending moment; Construction procedure; Finite element; Plasticity; Settlement; Tunnel design; Twin-tunnels; Thrust

1. Introduction The development of large cities requires the use the underground area for the construction of transportation infrastructures and facilities. In some cities, the geotechnical and underground conditions impose the construction of new tunnels close to existing ones. In other cases, the solution of twin-tunnels presents major advantages, such the reduction of the both the tunnel diameter and the soil movement resulting from the tunnel construction. Both numerical modeling and in situ observations were used to analyze the interaction between twin-tunnels (Soliman et al., 1993; Kawata and Ohtsuka, 1993; Perri, 1994; Saitoh et al., 1994; Yamaguchi et al., 1998; Shahrour and Mroueh, 1997). Results show that in some congurations,
Corresponding author. E-mail addresses: fchehade@ul.edu.lb (F. Hage Chehade), Isam. Shahrour@univ-lille1.fr (I. Shahrour). 0886-7798/$ - see front matter 2007 Elsevier Ltd. All rights reserved. doi:10.1016/j.tust.2007.03.004
*

the interaction could largely aect the soil settlement and that the design of twin-tunnels requires numerical analyses associated to monitoring during the tunnel construction. This paper concerns the design phase. It presents analysis of the interaction between twin-tunnels with a particular interest for the optimization of both the relative position of the twin-tunnels and the construction procedure. For this concern, a parametric study is conducted for the investigation of the inuence of these two factors on the soil settlement and internal forces resulting from the tunnel construction. The paper presents successively the numerical model and then analyses conducted for three congurations of twin-tunnels: aligned-horizontally, vertically and inclined (Fig. 1). 2. Numerical modeling Analyses are conducted using the nite element method. The soil behavior is described using an elastic perfectly

F. Hage Chehade, I. Shahrour / Tunnelling and Underground Space Technology 23 (2008) 210214

211

Sx Sy

Sx

Tunnels with horizontal alignment

Tunnels with inclined alignment

Sy

Tunnels with vertical alignment

Fig. 1. Congurations considered in the analysis of the interaction between twin-tunnels.

plastic constitutive relation based on the non associated MohrCoulomb criterion. The Youngs modulus of the soil E is supposed to increase with depth according to the following expression: Ez E0 P m =P 0
0:5

where Pm denotes the mean stress at the depth z; E0 is constitutive parameter, which corresponds to the Youngs Modulus at the mean pressure Pm = P0. This expression takes into account the variation of the Youngs modulus with the mean pressure, which increases with depth due to the soil self-weight. The behavior of the lining is assumed to be linear-elastic. The nite element modeling of the construction of twintunnels is carried out as follows: (i) Construction of the rst tunnel using the convergence-connement method with a stress release factor b = 0.5. This factor corresponds to the ratio of the stress release before the lining installation. (ii) Construction of the second tunnel using also the convergence-connement method, as for the rst tunnel with a stress release factor b = 0.5. This factor is applied to the stresses exercised around the tunnel after the excavation of the rst tunnel.

Finite element analyses were conducted using the nite element program PLAXIS. Fig. 2 shows the mesh used for the analysis of horizontally aligned tunnel with a ratio spacing Sx/D = 2(D, Sx denote the tunnel diameter and the distance between tunnel axes, respectively). It contains 2036 triangular 6-nodes elements. The soil layer is underlined by a sti one at a depth H = 8D. The lateral extension of the soil mass is equal to 20D. This extension ensures the absence of lateral boundary eect on the numerical modeling of the tunnel construction. Concerning the boundary conditions, the displacements are constrained in both directions at the bottom, while zero horizontal displacement is imposed at lateral boundaries (Fig. 2). Table 1 summarizes the properties of the soil and the lining used in this study. The soil corresponds to medium sand. The coecient of the lateral stress (K0) is equal to 0.5. The thickness of the lining is equal to 0.5 m.
Table 1 Properties of both the soil ratio and lining materials Material E0 (MPa) 30 35000 Poissons ratio 0.3 0.25 Cohesion (kPa) 3 Friction angle () 33 Dilancy angle ( ) 7 Unit weight kN/m3 18 25

Soil Lining

Fig. 2. Mesh used in the analysis of tunnels with horizontal alignment.

212

F. Hage Chehade, I. Shahrour / Tunnelling and Underground Space Technology 23 (2008) 210214

3. Result of analyses Analyses are conducted for the following congurations (Fig. 1, Table 2): tunnels with horizontal alignment tunnels with vertical alignment tunnels with inclined alignment

2.5 2.5 DD

Sx

Geometric configuration
Settlement at the soil surface (mm)

3.1. Tunnels with horizontal alignment Analyses were conducted for ves values of the tunnel spacing ration Sx/D (2, 2.5, 3, 4 and 5) (Fig. 3a). Fig. 3b shows the settlement pattern at the ground surface at the end of construction of the second tunnel. It shows that both the settlement pattern and amplitude depend on the distance between tunnels. The maximum soil settlement is observed for the conguration with close tunnel (Sx/ D = 2). In this case, the maximum soil settlement is induced between the two tunnels, it attains about 50 mm. The increase in the distance between tunnels induces a decrease in the settlement in the central part of the twintunnels and leads to a stabilization in the settlement above each tunnel. Beyond the distance (Sx = 3D), the construction of the rst tunnel does not aect the second one. The non-symmetry of the settlement results from the asymmetry of the plasticity induced in the soil mass as illustrated in Fig. 4. Fig. 3c and d show the distribution of the bending moment and the thrust in the right tunnel, respectively. It can be observed that the both the tunnels spacing and construction do not aect the internal forces in the tunnel. 3.2. Tunnels with vertical alignment Fig. 5a shows the tunnel conguration considered in this section. The upper tunnel center is located at 2.5D below the soil surface; the distance between the tunnel axes is equal to 2D (Fig. 5a). Two analyses were carried out. In the rst one, the upper tunnel is constructed at rst (reference case), while in the second analysis, the lower tunnel is constructed rst (inverted case). Results are presented at the achievement of the construction of tunnels. Fig. 5b, c and d illustrate the inuence of the construction procedure on the soil settlement, bending moment and thrust, respectively. It shows that the construction of the upper tunnel at rst leads to higher settlement and internal forces compared to that obtained by the construction of the lower

0 -10 -20 -30 -40 -50 -30 2D 2.5 D 3D 4D 5D -20 -10 0 10 20 30

X/D

Soil settlement induced by the construction of the twin-tunnel

Bending moment in the lining of right tunnel 2 (KNm)

300 200 100 0 -100 -200 -300 0 50 100 150 200 250 300 350 400
2D 2.5 D 3D 4D 5D

Radial Angle

Bending moment in the right tunnel

d
Thrust (kN)

-200 -300 -400 -500 -600 -700 0 50 100 150 200 250 300 350 400
2D 2.5 D 3D 4D 5D

Radial Angle

Thrust in the right tunnel

Fig. 3. Tunnels with horizontal alignment: Inuence of the construction procedure and congurations on the soil settlement and internal forces.

Table 2 Congurations of twin-tunnels analyzed in this paper Conguration Horizontal alignment Vertical alignment Inclined alignment Sx/D 2, 2.5, 3, 4 5 0 2, 2.5 Sy/D 0 2 2 Inclination angle (a) () 0 90 45, 39

tunnel at rst. The maximum settlement (Fig. 5b) in the rst case is about 12% higher than that in the second case, while the bending moment in the rst case is higher by about 23% than that induced in the second case (Fig. 5c);

F. Hage Chehade, I. Shahrour / Tunnelling and Underground Space Technology 23 (2008) 210214

213

Plasticity after the excavation of the first tunnel

Plasticity after the excavation of the second tunnel

Fig. 4. Plasticity in horizontal alignment tunnels (Sx = 2D).

a
2.5 D 2D

2.5 D

Sx Sy

Geometric configuration
Settlement at the soil surface (mm)

Settlement at the soil surface (mm)

Geometric configuration
0

-20

- 20

-40
reference case inverted case

- 40

alpha = 45 (reference case) alpha = 45 (inverted case) alpha = 39 (reference case) alpha = 39 (inverted case)

-60 -25

-20

-15

-10

-5

10

15

20

25

- 60

Soil settlement induced by the construction of the twin-tunnels Reference case: upper tunnel constructed at first Inverted case: lower tunnel constructed at first
Bending moment in the lining of tunnel 2 (kNm)

X/D

-10

-5

10

X/D

Bending moment in the lining of upper tunnel (kNm)

Settlement induced at the achievement of the construction Reference case: upper tunnel constructed at first Inverted case: lower tunnel constructed at first
400
alpha = 45 (reference case) alpha = 45 (inverted case) alpha = 39 (reference case) alpha = 39 (inverted case)

300 reference case 200 100 0 -100 -200 -300 0 50 100 150 200 250 300 350 400 inverted case

200

-200

-400

50

100

150

200

250

300

350

400

Radial angle

Bending moment in the upper tunnel at the end of construction

Radial angle

Bending moment in the lower tunnel

- 400

-20 -200 -30 -300

d
Thrust r (kN)

Thrust (kN)

- 600

-40 -400 -50 -500 -60 -600 -70 -700 -80 -800
alpha = 45 (reference case) alpha = 45 (inverted case) alpha = 39 (reference case) alpha = 39 (inverted case)

- 800

1 000

reference case inverted case

0 50 100 150 200 250 300 350 400

1 200

50

100

150

200

250

300

350

400

Radial angle

Radial angle

Thrust in the upper tunnel at the end of construction

Fig. 6. Tunnels with inclined alignment: Inuence of the construction procedure and tunnels conguration on the soil settlement and internal forces.

Thrust in the lower tunnel

Fig. 5. Tunnels with vertical alignment: Inuence of the construction procedure on the soil settlement and internal forces.

214

F. Hage Chehade, I. Shahrour / Tunnelling and Underground Space Technology 23 (2008) 210214

After the excavation of the upper tunnel

After the excavation of the lower tunnel

Fig. 7a. Plasticity in inclined alignment tunnels (a = 45) (Reference case).

After the excavation of the lower tunnel

After the excavation of the upper tunnel

Fig. 7b. Plasticity in inclined alignment tunnels (a = 45) (Inverted case).

the thrust in the reference case is higher by about 10% than that obtained in the inverted case (Fig. 5d). 3.3. Tunnels parallel inclined Two congurations were analyzed (Fig. 6a). The vertical distance between the tunnel axes is equal to Sy = 2D. In the rst conguration, the horizontal distance between the tunnel axes is equal to Sx = 2D(a = 45); in the second conguration Sx = 2.5D(a = 39). Fig. 6bd shows the inuence of both the tunnel conguration and construction procedure on the soil settlement and internal forces in the upper tunnel. It can be observed that the construction of the lower tunnel at rst (inverted case) leads to higher soil settlement than that induced when the upper tunnel is rst constructed (Fig. 6b). This result is similar to that obtained with vertical aligned tunnels. Both the bending moment (Fig. 6c) and thrust (Fig. 6d) in the upper tunnel are moderately aected by the order of construction of the tunnels. The inuence of the construction order on the plasticity distribution is illustrated in Figs. 7a and 7b. It shows that the construction order aects the distribution of plasticity induced around the tunnels. A lower plasticity level is observed around the tunnel constructed rst. 4. Conclusion This paper included a numerical analysis of the construction of twin-tunnels with a particular focus on the inuence of both the construction procedure and geometric conguration on the soil settlement and internal forces due to tunnels construction. The comparative study shows that the construction procedure (order of construction of tun-

nels) aects the soil settlement and bending moment. The construction of the upper tunnel at rst leads to higher settlement and bending moment, compared to that obtained by the construction of the lower tunnel at rst. The highest soil settlement is obtained for vertical aligned tunnels, while horizontal aligned tunnels cause the lowest settlement, but with a larger lateral extension of the settlement. The design of twin-tunnels must take into consideration other constraints related to the environment and to the underground congestion. The numerical modeling helps in considering such constraints to nd an optimal solution for the construction of twin-tunnels. References
Kawata, T., Ohtsuka, M., 1993. Observational construction of largescaled twin road tunnels with minimum interval. In: Reith, J.L. (Ed.), Infrastructures Souterraines de Transports. Balkema, Rotterdam. Perri, G., 1994. Analysis of the eects of the new twin-tunnels excavation very close to a big diameter tunnel of Caracas Subway. In: Salam, Abdel (Ed.), Tunnelling and Ground Conditions. Balkema, Rotterdam, pp. 523530. Saitoh, A., Gomi, K., Shiraishi, T., 1994. Inuence forecast and eld measurement of a tunnel excavation crossing right above existing tunnels. In: Salam, Abdel (Ed.), Tunnelling and Ground Conditions. Balkema, Rotterdam, pp. 8390. Shahrour, I., Mroueh, H., 1997. Three-dimensional non linear analysis of a closely twin tunnels. In: Sixth International Symposium on Numerical Models in Geomechanics (NUMOG VI), vol. 2. Montreal, Quebec, Canada, pp. 481487. Soliman, E., Duddeck, H., Ahrens, H., 1993. Two and three-dimensional analysis of closely spaced double-tube tunnels. Tunelling Underground Space Technol. 8 (1), 1318. Yamaguchi, I., Yamazaki, I., Kiritani, K., 1998. Study of ground-tunnel interactions of four shield tunnels driven in close proximity, in relation to design and constructions of parallel shield tunnels. Tunelling Underground Space Technol. 13 (3), 289304.