Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447

Contents lists available at ScienceDirect

Journal of Rock Mechanics and

Geotechnical Engineering
journal homepage: www.jrmge.cn

Full Length Article

Assessing the deformation response of double-track overlapped tunnels

using numerical simulation and field monitoring
Yao Hu a, b, Huayang Lei a, c, d, *, Gang Zheng a, c, d, e, Liang Shi a, Tianqi Zhang a, Zhichao Shen b,
Rui Jia a, c
a
School of Civil Engineering, Tianjin University, Tianjin, 300350, China
b
Department of Civil and Environmental Engineering, National University of Singapore, 117576, Singapore
c
Key Laboratory of Coast Civil Structure Safety (Tianjin University), Ministry of Education, Tianjin, 300350, China
d
Key Laboratory of Earthquake Engineering Simulation and Seismic Resilience of China Earthquake Administration (Tianjin University), Tianjin, 300350, China
e
State Key Laboratory of Hydraulic Engineering Simulation and Safety (Tianjin University), Tianjin, 300350, China

a r t i c l e i n f o a b s t r a c t

Article history: The unprecedented rate of metro construction has led to a highly complex network of metro lines.
Received 23 December 2020 Tunnels are being overlapped to an ever-increasing degree. This paper investigates the deformation
Received in revised form response of double-track overlapped tunnels in Tianjin, China using finite element analysis (FEA) and
8 June 2021
field monitoring, considering the attributes of different tunneling forms. With respect to the upper
Accepted 17 July 2021
Available online 28 August 2021
tunneling, the results of the FEA and field monitoring showed that the maximum vertical displacements
of the ground surface during the tail passage were 2.06 mm, 2.25 mm and 2.39 mm obtained by the FEA,
field monitoring and Peck calculation, respectively; the heaves on the vertical displacement curve were
Keywords:
Double-track overlapped tunnels
observed at 8 m (1.25D, where D is the diameter of the tunnel) away from the center of the tunnel and
Deformation response the curve at both sides was asymmetrical. Furthermore, the crown and bottom produce approximately
Finite element analysis (FEA) 0.38 mm and 1.26 mm of contraction, respectively. The results of the FEA of the upper and lower sections
Field monitoring demonstrated that the tunneling form has an obvious influence on the deformation response of the
Upper and lower tunneling double-track overlapped tunnel. Compared with the upper tunneling, the lower tunneling exerted
significantly less influence on the deformation response, which manifested as a smaller displacement of
the strata and deformation of the existing tunnel. The results of this study on overlapped tunnels can
provide a reference for similar projects in the future.
Ó 2022 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting by
Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/4.0/).

1. Introduction and 4 in China. In general, the environment surrounding over-

lapped tunnels tends to be complex and tunnel construction can
The unprecedented rate of metro construction has become a cause a great ground disturbance, thus endangering the safety of
highly complex network of metro lines. Overlapped tunnels have adjacent building structures. Therefore, it is necessary to under-
progressively appeared and become ever more intensive. Examples stand the deformation response induced by overlapped tunneling.
of overlapped tunnels include Beijing Metro Lines 2 and 4 between The geometric arrangement of overlapped tunnels has a sig-
Beijing and Xuanwumen stations in China, the double-track over- nificant effect on the deformation response during the tunneling
lapped tunnels of the MRT North-East Line in Singapore, the tunnel process (Chehade and Shahrour, 2008; Shahin et al., 2016; Jin et al.,
of Shenzhen Metro Shekou Line crossing right under the tunnel of 2019). Many studies have investigated the deformation response of
Luobao Line in China, and the four-track overlapped tunnels be- overlapped tunnels with different geometric arrangements. For
tween Xujiahui and Stadium stations of Shanghai Metro Lines 11 double-track overlapped tunnels, most research has focused on
three types of geometric arrangements: horizontally parallel
(Suwansawa and Einstein, 2007; Chen et al., 2011; Das et al., 2017;
* Corresponding author. School of Civil Engineering, Tianjin University, Tianjin, Zhang et al., 2019), vertically parallel (Ng et al., 2015; Fang et al.,
300350, China. 2016), and intersecting (Marshall et al., 2013; Chen et al., 2018;
E-mail address: [email protected] (H. Lei).
Qian et al., 2019; Lai et al., 2020). Many studies have focused on
Peer review under responsibility of Institute of Rock and Soil Mechanics, Chi-
nese Academy of Sciences.

https://doi.org/10.1016/j.jrmge.2021.07.003
1674-7755 Ó 2022 Institute of Rock and Soil Mechanics, Chinese Academy of Sciences. Production and hosting by Elsevier B.V. This is an open access article under the CC BY-
NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447 437

Fig. 1. Overview of the project: (a) location of M6 and (b) section of the double-track overlapped tunnel (Google Maps, annotations made by author).

horizontally parallel and intersecting arrangements, while few 2. Engineering background of the case study
have focused on vertically parallel arrangements.
Among the studies that did consider vertically parallel double- 2.1. Project overview
track overlapped tunnels, Ng et al. (2015) used centrifuge tests
and numerical modeling to analyze the deformation response The section of double-track overlapped tunnels investigated be-
during tunneling with a focus on the response of a pile group. Fang longs to the interval between Wenhuazhongxin and Leyuandao sta-
et al. (2016) investigated the ground surface settlement due to the tions of Rail Transit Line 6 (M 6) in Tianjin, China. The total length of
tunneling process and considered the influence of different geo- the interval between the two stations is 776 m and that of the double-
metric arrangements; however, the tunnel section had a horseshoe track overlapped section of the tunnels is 392 m, as shown in Fig. 1.
shape and the shallow tunneling method was used for excavation. An earth pressure balancing (EPB) shield machine with a
Few studies have considered the deformation response of vertically diameter of 6.4 m was used to excavate the tunnel, and its face
parallel double-track overlapped tunnels excavated by the shield pressure was 0.25e0.3 MPa. The shield started from the working
tunneling method and with a circular tunnel section. shaft of Wenhuazhongxin station and reached the working shaft of
The tunneling form also affects the deformation response of Leyuandao station. During the tunneling process, the shield passed
double-track overlapped tunnels during the tunneling process. Ma the adjacent existing tunnels of Tianjin Rail Transit Line Z1 (M Z1)
et al. (2018) studied a pipeline’s settlement and load-transfer to form the section of the double-track overlapped tunnels. The
mechanism induced by double-track overlapped tunneling with tunnel lining was assembled with prefabricated reinforced con-
different tunneling forms. Soomro et al. (2020) employed centri- crete segments in staggered joints, which consisted of one cap
fuge tests and numerical modeling to explore the stress-transfer block, two adjacent blocks, and three standard blocks. The strength
mechanisms and settlement of a pile group during double-track grade of the segments is C55, and the segments are connected by
overlapped tunneling with different construction sequences. bending bolts. In addition, the tunnel lining was designed to
However, the above studies focused on the response of existing pile function with an outer diameter of 6.2 m and a thickness of 0.35 m,
groups and pipelines but did not evaluate the deformation with the width of each ring being 1.5 m.
response of existing tunnels.
In this study, finite element analysis (FEA) and field monitoring 2.2. Geological conditions
were employed to investigate the deformation response on the
ground surface and new tunnel (longitudinal and transverse) dur- Tianjin is located in the east of the North China Plain and
ing upper tunneling activity for the double-track overlapped tun- downstream of the Haihe River Basin. It is an alluvial plain with flat
nels in Tianjin, China. Moreover, the FEA of the upper and lower terrain and extensive soft soil. Fig. 2 shows the illustration of the
tunneling was used to explore the influence of the tunneling form ground profile in the direction of the tunnel axis. It can be seen that
on the deformation response of the double-track overlapped tun- the main strata of the interval between Wenhuazhongxin and
nels. The displacement in the strata (vertical and horizontal) and Leyuandao stations consist of soft soils (including silty clay, sandy
deformation of the existing tunnel (longitudinal and transverse) silt and silt), and the strata crossed by the double-track overlapped
were analyzed in detail. The study of overlapped tunnels described tunnels mainly consist of silty clay and silt. In addition, the mini-
in this paper can provide a reference for similar projects in the mum cover thickness of M 6 is 6.8 m, and the minimum vertical
future. distance between M 6 and M Z1 is 7.2 m.
438 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447

Fig. 2. Illustration of the ground profile in the direction of the tunnel axis.

Fig. 3 shows the variations of the typical geotechnical properties software, ABAQUS, was used to numerically explore the distur-
of the simplified soil layers with depth. These indices include unit bance effect of the tunneling process on the double-track over-
weight, water content, void ratio, plasticity index, liquidity index, lapped tunnels (Sun, 2015; Zhang et al., 2016). This paper chooses
constrained factor, constrained modulus, cohesion, internal friction the small-strain FEA in the ‘‘Soils’’ step to calculate the deformation
angle and standard penetration test (SPT) blow counts; and the of soil, which can improve the computational efficiency (Liu et al.,
minimum, maximum and average values of each index are given. 2014).
During the construction process, the impact of unfavorable When taking into account factors such as the burial depth and
geological sections should be fully considered and the driving pa- spacing of the two tunnels, it is generally appropriate to select a
rameters should be properly set to avoid engineering risks. The calculation range that is not less than 3De4D along the tunnel
groundwater table in the site is within the range of 0.8e1.9 m. diameter in all directions, where D is the diameter of the tunnel (Sun,
2015). Therefore, the dimensions of the finite element model were
2.3. Field monitoring layout determined to be 63 m  40 m  46 m (length  width  depth), as
shown in Fig. 5. The finite element model includes five types of
Fig. 4 shows the field monitoring layout used in this study. The components (soil part, shield part, grouting part, lining part, and an
monitoring points were only arranged in the new tunnel (M 6) and existing tunnel part) and two pressure types (support pressure of the
on the ground surface to measure the deformation response of the shield tunneling face and grouting pressure); the dimensions of the
double-track overlapped tunnels to the upper tunneling process, parts and pressures are consistent with the actual project.
with a total of 5 monitoring points for vertical displacement of the In this study, the C3D8P elements (8-node trilinear displace-
ground surface, 18 monitoring points for longitudinal deformations ment and pore pressure) were selected to simulate soil strata
of the tunnel, and 36 monitoring points for transverse de- because the pore water pressure must be considered. A non-
formations of the tunnel. Fig. 4a shows the top view of the field uniform grid was used with a fine grid in the region close to the
monitoring layout, and section A-A0 in the transverse direction was tunnel and a coarser grid in the far field. The C3D8I elements (8-
chosen to show the field monitoring layout of the new tunnel, as node linear brick, incompatible mode) were employed to simulate
shown in Fig. 4b. “GS”, “LS” and “TS” represent the monitoring the shield, grouting, lining, and existing tunnels (Dong et al., 2014).
points of vertical displacement on the ground surface, the longi- The finite element model has a total of 26,688 elements. Among
tudinal deformation of the tunnel, and the transverse deformation these, the dualistic interactions between the soil and grouting
of the tunnel, respectively. The automatic leveling level DS05 parts, the soil and shield parts, the grouting and lining parts, and
(Suzhou FOIF Co., Ltd., China), the total station SET1X and prisms the soil and existing tunnel parts were connected by the tie. In
(Sokkia Co., Tokyo, Japan) and the hand-held laser rangefinder addition, the boundary conditions of the model along the X and Y
DLE50 (Robert Bosch GmbH, Stuttgart, Germany) were used in this axes separately limited “U1” and “U2” to ensure that the horizontal
study. displacement was constrained, whereas the vertical direction could
be spontaneously moved. In the Z direction, “U1”, “U2” and “U3” on
3. FEA of the double-track overlapped tunneling process the base surface were limited (both the horizontal and vertical
directions are constrained).
3.1. Finite element model Based on the upper tunneling activity, the finite element model
of the lower tunneling (the existing tunnel is above and the new
Given the engineering background of the double-track over- tunnel is below) is proposed as a basis for the discussion of the
lapped tunnels (upper tunneling), the three-dimensional FEA influence of the tunneling form on the deformation response of the
Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447 439

Fig. 3. Typical geotechnical properties and indices with depth.

440 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447

by 15% in this study (Sun, 2015). The parameters cannot be

measured for the fill, and thus silty clay was used instead.

3.3. Modeling procedure

The shield construction process is complicated, involving shield

tunneling, soil excavation, assembling lining, and shield tail
grouting. In this study, this complex process was reasonably
simplified using a “step-by-step tunneling” method (Mollon et al.,
2013). In particular, the “element death” method was adopted to
incorporate the tunneling process of the EPB shield machine, i.e. to
create the “interaction” of “mode change” in ABAQUS. Fig. 6 in-
troduces the detailed modeling procedure, which is described as
follows:

(1) Initial earth stress balance: The acceleration due to gravity

applied to the model (deactivation of the shield, grouting,
and lining) is taken as 9.8 m/s2 and then automatically
balanced; the termination condition is that the displacement
of the model is less than 1  10-6 m.
(2) Excavation 1: Reactivate the first slice shield (shield machine
length: 10.5 m, which is equivalent to seven slices of lining),
deactivate the first slice soil, and add the first support pres-
sure from the shield tunneling face.
(3) Excavation 2: Reactivate the second slice shield (simulta-
neously deactivate the first slice shield), deactivate the sec-
ond slice soil, add the second support pressure of the shield
tunneling face (simultaneously remove the first support
pressure), reactivate the first slice lining (simultaneously add
the first slice grouting pressure), and reactivate the first slice
soft-grouting.
(4) Excavation 3: Reactivate the third slice shield (simulta-
neously deactivate the second slice shield), deactivate the
third slice soil, add the third support pressure of the shield
tunneling face (simultaneously remove the second support
pressure), reactivate the second slice lining (simultaneously
add the second slice grouting pressure), and reactivate the
second slice soft-grouting (simultaneously hardening the
first slice grouting).
(5) Repeat Steps (2)e(4) until the shield has been moved out of
the hole.
Fig. 4. Field-monitoring layout: (a) top view and (b) section AeA.
4. Results and discussion

double-track overlapped tunnels. In the finite element model in the FEA and field monitoring results were compared to investigate
lower tunneling process, everything apart from the tunneling form the deformation response of the double-track overlapped tunnels
is the same as that in the upper tunneling case. during the upper tunneling process. The following aspects were
primarily analyzed: (1) vertical displacement of GS-1 during
3.2. Material properties different tunneling stages, (2) vertical displacement of the ground
surface, (3) longitudinal deformation of the new tunnel, and (4)
The material properties in this finite element model can be transverse deformation of the new tunnel. Based on this, the in-
divided into two categories: elastoplastic material (soil) and linear fluence of different tunneling forms on the double-track over-
elastic material (shield, lining, grouting, and the existing tunnel). lapped tunnels was evaluated via an FEA of the upper and lower
For the soil, the elastic part adopts the porous elastic model, tunneling.
whereas the plastic part adopts the modified Cam-Clay model. As
the materials are relatively uniform and have a high degree of 4.1. Vertical displacement of GS-1 during different tunneling stages
stiffness, linear elastic constitutive models were used for the shield,
lining, grouting, and existing tunnel (Zhang et al., 2016). The ma- Fig. 7 shows the vertical displacement of the ground surface at
terial parameters of the constitutive model used in the FEA are monitoring point GS-1 during different tunneling stages. Before GS-
listed in Table 1. The finite element model considers the grouting 1 was reached (face pre-arrival), the heave occurred, which was
hardening process, “soft” indicating that the grouting has just been observed in both the FEA and field monitoring. Liu et al. (2014)
injected and “hard” indicating that it has hardened. Furthermore, stated that the face pressure (i.e. the pressure of earth acting on
owing to the fact that the bolted connection between the linings in the tunnel face) is the main influencing factor before the tunnel
the actual project reduces its strength (Zheng et al., 2015; Zhang face is reached during the tunneling process. Therefore, it can be
et al., 2021), the Young’s modulus of the linings was discounted deduced that this heave was primarily caused by face pressure.
 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447 441

was still revealed by the two approaches, but the difference be-
tween field monitoring and FEA was greater than that for the face
arrival. The settlement at this stage may be attributed to the
grouting pressure and contraction during grouting.
In this study, the conicity of the shield machine was considered
in the field monitoring but not in the FEA. In addition, the field
monitoring indicated non-uniform contraction during grouting
(Fig. 8b), whereas the FEA revealed uniform contraction. These
explain the differences between field monitoring and FEA.
Compared with the maximum vertical displacements observed for
a single tunnel (FEA: 20 mm; field monitoring: 7.5 mm) (Liu et al.,
2014), the maximum vertical displacements in this study were
significantly smaller (FEA: 2.5 mm; field monitoring: 1.5 mm). This
may be attributed to the effect of the existing tunnel.

4.2. Vertical displacement of the ground surface

To further explore the response to the tunnel construction, the

Peck (1969) formula was used to calculate the transverse vertical
displacement of the ground surface in the tunneling process. Based
on the geological conditions of this case, Eq. (1) was used to
calculate the width of the settlement trough i (Wei, 2009):
h  4 i
i ¼ m R þ htan 45  (1)
2

where m is a coefficient in the range of 0.45e0.5, R is the tunnel

radius, h is the depth of the tunnel axis, and 4 is the internal friction
angle of the soil above the tunnel.
Fig. 9 compares the transverse vertical displacement of the
ground surface in the tunneling process (i.e. excavated to a distance
of 31.5 m) for the FEA with the field monitoring and Peck calculation.
Fig. 9a presents the tunneling face arrival stage, while Fig. 9b shows
the shield tail passage stage. It can be seen that except for the
maximum vertical displacement, the trend of the vertical displace-
ment from the center of the tunnel to the two sides is similar for the
face arrival and tail passage stages. The maximum vertical
displacement during the tail passage (FEA: 2.06 mm) is significantly
Fig. 5. Finite element model meshing of the upper tunneling: (a) whole view and (b) greater than that of the face arrival (FEA: 1.02 mm). Therefore, we
cut in half along the Y-axis. take the tail passage (Fig. 9b) as an example for analysis.
As shown in Fig. 9b, the trend of the vertical displacement ob-
tained by the FEA is similar to both the field monitoring and Peck
When GS-1 was crossed (face arrival), the settlement was observed calculation, all of which have a settlement trough. Compared with
by both approaches, but the field monitoring revealed greater the field monitoring (2.25 mm) and Peck calculation (2.39 mm),
settlement than FEA did. According to Ji et al. (2008), at this stage, FEA (2.06 mm) tends to provide a smaller maximum vertical
the conicity of the shield machine (Fig. 8a) and ovalization of the displacement under the tail passage stage. This may be because the
lining (Fig. 12) may have the greatest influence on the vertical conicity of the shield machine and non-uniform contraction during
displacement. After GS-1 was passed (tail passage), the settlement grouting are not taken into account in the FEA (Fig. 8). Furthermore,

Table 1
Material constitutive parameters used in the finite element analysis.

Soil layers Thickness (m) Unit weight (kN/m3) M l k n k (m/d) e0

Silty Clay 2.5 18.4 0.86 0.058 0.0072 0.32 0.00018 0.776
Sandy Silt 3 17.9 1.03 0.031 0.0039 0.35 0.00050 0.742
Silty Clay 10.3 18.1 0.89 0.055 0.0069 0.35 0.00031 0.764
Silt 12.2 19.8 1.37 0.020 0.0025 0.3 0.00020 0.595
Silty Clay 18 18.6 0.90 0.047 0.0059 0.35 0.00485 0.683

Materials Thickness (m) Unit weight (kN/m3) E (MPa) n

Existing tunnel 0.45 20 2104 0.3
Shield 0.1 78 2.4105 0.2
Grouting Soft 0.1 20 5 0.4
Hard 18 0.2
Lining 0.35 25 2.93104 (15% discount) 0.2

Note: E: Young’s modulus; n: Poisson’s ratio; M: Slope of the critical state line in qep0 space; l and k: Slopes of the normal compression and recompression lines in eeln p0
space; e0: Initial void ratio; k: Permeability coefficient.
442 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447

the crown and bottom of the tunnel mean that they underwent
expansion and contraction, respectively. Furthermore, the positive
value of the difference between the bottom and crown means that
the tunnel collectively moved upward.
It can be seen from Fig. 11 that the longitudinal deformation
trends of the FEA and field monitoring after tunneling are almost
identical. The trend can be divided into three stages: Stage I: the
longitudinal deformation increases and reaches the maximum
(positive) at 42.5 m (3.2D) away from the end of tunnel; Stage II: the
longitudinal deformation drops sharply and reaches the minimum
(negative) at 56.5 m (1D) away from the end of tunnel; and Stage III:
the longitudinal deformation has a slight increase. This indicates
that the tunnel deformation changed from upward movement to
downward movement at 1De3.2D away from the end of the tunnel.
Attention should be given to the tunnel deformation at 3.2D away
from the end of the tunnel for the construction of double-track
overlapped tunnels. The difference between the FEA and field
monitoring may be attributed to the difference in connections be-
tween the linings. In other words, the FEA adopted the equivalent
stiffness method, while bolted connections were used in the actual
engineering project.

4.4. Transverse deformation of the new tunnel

Fig. 12 shows the transverse deformation of the new tunnel in

the upper tunneling process. The transverse deformations obtained
by the FEA and field monitoring are essentially the same. Compared
with the pre- and post-tunneling stages, the left and right sides of
the new tunnel produce a non-uniform expansion (0.68 mm at the
left and 0.49 mm at the right), whereas the crown and bottom
Fig. 6. Finite element modeling procedure of the double-track overlapped tunnel: (a) produce 0.38 mm and 1.26 mm of contraction, respectively. Clearly,
excavation 2 and (b) excavation 3.
there is a greater deformation at the bottom of the tunnel than at
the crown. The reasons for this are the excavation of soil causing a
in the FEA and field monitoring, the vertical displacement curve is stress release and the thin overburden of the soil (6.8 m z 1.06D),
uplifted at 8 m (1.25D) away from the center of the tunnel, but this which lead to the center of the new tunnel moving upwards after
is not seen in the Peck calculation. The reason for this heave is the tunneling compared with the case prior to tunneling. This also
upward movement of the existing tunnel due to buoyancy (Zheng explains the heave of ground surface at both sides of the tunnel
et al., 2019), which requires that the release of water pressure described in Section 4.2 (Fig. 9).
during metro tunnel construction should be regulated.
Zheng et al. (2019) conducted field monitoring on the over- 4.5. Influence of tunneling form on the deformation response
lapped tunnel to study the vertical displacement of the ground
surface. The construction sequence was as follows: the down-line A comparison of the FEA results between upper and lower
tunneling followed by the up-line tunneling one year later. Fig. 10 tunneling was performed to discuss the influence of the tunneling
shows a comparison of the results between Zheng et al. (2019) form on the deformation response of the double-track overlapped
and this study during the tail passage. Clearly, the four trends of tunnels. The following aspects were analyzed: (1) vertical
the vertical displacement are similar, with a heave on the vertical displacement of the strata, (2) horizontal displacement of the
displacement curve at the left and right sides of the tunnel (in this strata, (3) longitudinal deformation of the existing tunnel, and (4)
study, field monitoring is only performed on the left of the tunnel). transverse deformation of the existing tunnel.
In addition, apart from the field monitoring in this study, the ver-
tical displacements on the left and right sides of the tunnel are 4.5.1. Vertical displacement of the strata
asymmetrical and their uplift values are different. Fig. 13 shows the vertical displacement of the ground surface
and monitoring point MD (between the new and existing tunnel, at
a depth of 16.8 m) in accordance with different tunneling forms.
4.3. Longitudinal deformation of the new tunnel
Fig. 13a shows the longitudinal direction after tunneling. On the
ground surface, the vertical displacement due to different
Fig. 11 indicates the longitudinal deformation of the new tunnel
tunneling forms (upper and lower) gradually increases with
after tunneling (i.e. excavated to a distance of 63 m) in the FEA and
tunneling. However, the increase is relatively large in the upper
field monitoring. The deformation DD can be calculated from the
tunneling process and its value ranges from 0.66 mm
difference in the longitudinal deformation between the bottom and
to 2.78 mm, whereas the increased range is relatively small dur-
crown after tunneling:
ing the lower tunneling and its value ranges from 0.66 mm
Dbottom  Dcrown to 1.11 mm. The difference between the upper and lower
DD ¼ (2) tunneling is 0 mm and 1.67 mm at 0 m and 63 m away from the
2
starting position of tunneling, respectively. At the monitoring point
where Dbottom and Dcrown are the longitudinal deformations of the MD, the vertical displacement fluctuates by 3 mm in the upper
tunnel at the bottom and crown, respectively. The positive values at tunneling process, but it increases with small fluctuations arising
 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447 443

and geometric arrangement between overlapped tunnels (Fang

et al., 2016) are critical factors influencing the horizontal
displacement of the strata during the tunneling process. We can
infer that this difference between the maximum horizontal dis-
placements in the upper and lower tunneling processes can be
attributed to the factors such as the buried depth of the new tunnel,
its deformation, and the position of the existing tunnel.
The direction of horizontal displacement was outward from the
tunnel, which is consistent with the results of Dias and Kastner
(2013). However, this can change depending on the situation. For
example, Zhang et al. (2016) observed inward movement, and
Standing and Selemetas (2013) observed both inward and outward
movements. The direction of horizontal displacement may be
related to the new tunnel deformation. In this study, the expansion
on both sides of the new tunnel (Fig. 12) led to an outward move-
ment. Furthermore, the position of the maximum horizontal
displacement has a slight upward deviation from the original po-
sition of the tunnel’s center and that upward shift is greater during
Fig. 7. Vertical displacement of monitoring point GS-1 during different tunneling upper tunneling activity. This was due to the discrepancy in the
stages.
cover thickness of the new tunnel. The lower tunnel had a cover
thickness of 20.4 m, while the upper tunnel had a cover thickness of
from lower tunneling and its value ranges from 0.39 mm only 6.8 m.
to 1.02 mm. The difference between the upper and lower
tunneling is 3.41 mm and 3.99 mm at 0 m and 63 m away from the 4.5.3. Longitudinal deformation of the existing tunnel
starting position of tunneling, respectively. Fig. 15 shows the longitudinal deformation of the existing tunnel
Fig. 13b shows the transverse direction in the upper and lower after upper and lower tunneling. Upward deformation is defined as
tunneling processes (i.e. excavated to a distance of 31.5 m). On the positive, whereas downward deformation is defined as negative. In
ground surface, the trends of vertical displacement in the upper the upper tunneling process, the variations of the longitudinal
and lower tunneling processes are essentially the same, with both deformation at the crown and bottom are roughly the same, but the
being settled. However, the settlement of the strata induced by the longitudinal deformation of the former is significantly larger than
upper tunneling (maximum value of 2.06 mm) is obviously that of the latter. The deformation values range from 2.26 mm to
greater than that by the lower tunneling (maximum value 2.38 mm at the crown and from 1.56 mm to 1.67 mm at the bottom.
of 0.37 mm). At the monitoring point MD, the trends of vertical In the lower tunneling process, the variations of the longitudinal
displacement in accordance with two tunneling forms are reversed, deformation at the crown and bottom are essentially the same,
i.e. the upper tunneling shows a heave (maximum value of 2.5 mm) with a difference of only 0.05 mm. In addition, compared with the
and the lower tunneling shows a settlement (maximum value upper tunneling process, the longitudinal deformation and differ-
of 0.11 mm). ence between the crown and bottom in the lower tunneling process
The above analysis shows that the effects of different tunneling are significantly smaller. The longitudinal deformation of the
forms on the vertical displacement of the strata are completely existing tunnel is caused by the strata displacement, especially the
different. The upper tunneling causes the strata to settle on the vertical displacement. This is consistent with the large difference
ground surface and to heave at the monitoring point MD, whereas between the two tunneling forms in terms of the vertical
the lower tunneling has little influence on the ground surface and displacement of the strata.
induces settlement in the monitoring point MD. This indicates that
during lower tunneling activity, the existing tunnel has an inhibi- 4.5.4. Transverse deformation of the existing tunnel
tory effect on the vertical displacement of the ground surface. Fig. 16 shows the transverse deformation of the existing tunnel
in the upper and lower tunneling processes. The deformation of the
existing tunnel in the transverse direction is more intuitive. In the
4.5.2. Horizontal displacement of the strata upper tunneling process, compared with pre- and post-tunneling,
Fig. 14 illustrates the transverse horizontal displacement of the the left and right sides of the existing tunnel present a contrac-
strata in the upper and lower tunneling processes under different tion of approximately 1.56 mm, whereas the crown and bottom
tunneling forms and at different stages. The trends of the transverse have upward deformations of about 1.78 mm and 1.49 mm,
horizontal displacement at two stages (face arrival and tail passage) respectively. This indicates that the existing tunnel collectively
are similar, and the difference in the magnitude, i.e. the horizontal moves up. The reason for this is that the stress release caused by the
displacement value of the tail passage, is greater than that during excavation of the new tunnel results in the upward displacement of
the face arrival. Thus, only the tail passage was selected for analysis the strata, thus affecting the existing tunnel. In the lower tunneling
in this study. process, the deformation of the existing tunnel prior to and after
The result of the tail passage indicates that under upper and tunneling is very small, i.e. only 0.25 mm. Therefore, compared
lower tunneling, the differences among the horizontal with upper tunneling, the transverse deformation of the existing
displacement-depth responses mainly occur at the location of the tunnel in the lower tunneling process is significantly smaller. This
new tunnel (upper tunneling at a depth of 10 m and lower also explains the upward deviation of the position of the maximum
tunneling at a depth of 23.6 m). Compared with the lower horizontal displacement described in Section 4.5.2 (Fig. 14). Clearly,
tunneling (1.44 mm), the maximum value of the horizontal the tunneling form has a great influence on the deformation
displacement during upper tunneling (2.63 mm) is larger. Accord- response of the existing tunnel.
ing to previous research, the cover thickness of a new tunnel The above analysis shows that compared with the lower
(Zhang et al., 2016), ovalization of the tunnel lining (Liu et al., 2014), tunneling, the upper tunneling had a greater influence on the
444 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447

Fig. 8. Influence factors: (a) conicity of the shield machine and (b) nonuniform contraction.

existing tunnel on the top during the lower tunneling, which

inhibited the deformation response of the strata and existing tun-
nel. Based on this, some suggestions for similar engineering pro-
jects in the future are given. Compared with the lower tunneling,
the upper tunneling should be given attention when the cover
thickness is only about 1D. The strata should be reinforced in the
vertical and horizontal directions to reduce the displacement. The
left and right sides of the existing tunnel should be reinforced to
mitigate the contraction, and its crown and bottom should be
reinforced to reduce the upward deformation.

5. Conclusions

A case study was conducted on double-track overlapped tunnels

in Tianjin, China using FEA and field monitoring to explore the
deformation response during upper tunneling and the influence of
the tunneling form. The main conclusions are drawn as follows:

(1) Similar results were obtained by FEA and field monitoring in

this case study. The vertical displacement at the monitoring
point GS-1 showed uplift before it was reached and settle-
ment after it was passed. The Peck formula was used to
calculate the vertical displacement of the ground surface; the
maximum vertical displacements for tail passage (FEA:
2.06 mm, field monitoring: 2.25 mm, and Peck: 2.39 mm)

Fig. 9. Transverse vertical displacement of the ground surface in the middle of the
tunnel: (a) Face arrival, and (b) Tail passage.

deformation response of the double-track overlapped tunnels. This

manifested as a greater displacement of the strata and deformation
of the existing tunnel. This may be because there was no existing
tunnel on the top during the upper tunneling and the cover Fig. 10. Comparison of the results of Zheng et al. (2019) and this study for the tail
thickness was very small (6.4 m z 1.06D). By contrast, there was an passage state.
 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447 445

Fig. 11. Longitudinal deformation of the new tunnel after tunneling completion.

Fig. 13. Vertical displacement of the stratum: (a) longitudinal (after tunneling) and (b)
transverse (in the middle of tunneling).

Fig. 12. Transverse deformation of the new tunnel in the middle of tunneling (unit:
mm). The numbers around the circle are in degree.

were significantly greater than those for face arrival (FEA:

1.02 mm, field monitoring: 1.16 mm, and Peck: 1.25 mm).
Furthermore, a heave on the vertical displacement curve was
observed at 8 m (1.25D) away from the center of the tunnel,
and the curve at both sides was asymmetrical. The results
were compared for verification. The longitudinal deforma-
tion of the new tunnel changed from upward to downward
between 1D and 3.2D from the end of the tunnel. The
transverse deformation of the new tunnel showed expansion
at the left (0.68 mm) and right (0.49 mm) sides, whereas an
upward movement was generated at the center and
contraction was produced at the crown (0.38 mm) and bot-
tom (1.26 mm).
Fig. 14. Transverse horizontal displacement of the stratum in the middle of tunneling.
(2) FEA of the upper and lower tunneling was implemented to
explore the influence of the tunneling form on the defor-
mation response of the strata and existing tunnel. The response during the tunneling of the double-track over-
tunneling form significantly influenced the deformation lapped tunnels. Regarding the vertical displacement, the
446 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447

the lower tunneling had little influence on the ground sur-

face and induced settlement at the monitoring point MD.
Regarding the horizontal displacement, the upper tunneling
caused a greater maximum displacement (2.63 mm) than the
lower tunneling (1.44 mm). In addition, the position of the
maximum displacement deviated slightly upward from the
original position at the tunnel center. For the existing tunnel,
the longitudinal deformation and difference at the crown
and bottom during the lower tunneling were significantly
smaller than that during the upper tunneling. The lower
tunneling had a significantly smaller transverse deformation
(upward deformation of 0.25 mm) than the upper tunneling
(1.56 mm contraction at the left and right sides, 1.78 mm
upward deformation at the crown, and 1.49 mm upward
deformation at the bottom).

Data availability statement

Some or all data, models, or code that support the findings of

this study are available from the corresponding author upon
reasonable request.

Declaration of competing interest

The authors declare that they have no known competing

financial interests or personal relationships that could have
appeared to influence the work reported in this paper.

Acknowledgments

The work described in this paper was financially supported by

the Open Project of the State Key Laboratory of Disaster Reduction
in Civil Engineering (Grant No. SLDRCE17-01), the National Key
Research and Development Program of China (Grant No.
2017YFC0805402), and the National Natural Science Foundation of
Fig. 15. Longitudinal deformation of the existing tunnel after tunneling: (a) upper
China (Grant No. 51808387). All the support is greatly appreciated.
tunneling; and (b) lower tunneling.
References

Chehade, F.H., Shahrour, I., 2008. Numerical analysis of the interaction between
twin-tunnels: influence of the relative position and construction procedure.
Tunn. Undergr. Space Technol. 23 (2), 210e214.
Chen, R.P., Zhu, J., Liu, W., Tang, X.W., 2011. Ground movement induced by parallel
EPB tunnels in silty soils. Tunn. Undergr. Space Technol. 26 (1), 163e171.
Chen, R.P., Lin, X.T., Kang, X., Zhong, Z.Q., Liu, Y., Zhang, P., Wu, H.N., 2018. Defor-
mation and stress characteristics of existing twin tunnels induced by close-
distance EPBS under-crossing. Tunn. Undergr. Space Technol. 82, 468e481.
Das, R., Singh, P.K., Kainthola, A., Panthee, S., Singh, T.N., 2017. Numerical analysis of
surface subsidence in asymmetric parallel highway tunnels. J. Rock Mech.
Geotech. Eng. 9 (1), 170e179.
Dias, D., Kastner, R., 2013. Movements caused by the excavation of tunnels using
face pressurized shields - analysis of monitoring and numerical modeling re-
sults. Eng. Geol. 152 (1), 17e25.
Dong, Y., Burd, H., Houlsby, G., Hou, Y., 2014. Advanced finite element analysis of a
complex deep excavation case history in Shanghai. Front. Struct. Civ. Eng. 8 (1),
93e100.
Fang, Q., Tai, Q., Zhang, D.L., Wong, L.N.Y., 2016. Ground surface settlements due to
construction of closely-spaced twin tunnels with different geometric arrange-
ments. Tunn. Undergr. Space Technol. 51, 144e151.
Ji, Q., Huang, Z., Peng, X., 2008. Analysis on influence of conicity of extra-large
diameter mixed shield machine on surface settlement. In: Complimentary
Special Issue to the 6th International Symposium on Geotechnical Aspects of
Underground Construction in Soft Ground. Shanghai, China, pp. 237e242.
Jin, D., Yuan, D., Liu, S., Li, X., Luo, W., 2019. Performance of existing subway tunnels
Fig. 16. Transverse deformation of the existing tunnel in the middle of tunneling (unit: undercrossed by four closely spaced shield tunnels. J. Perform. Constr. Facil. 33
mm). The numbers around the circle are in degree. (1), 04018099.
Lai, H.P., Zheng, H.W., Chen, R., Kang, Z., Liu, Y., 2020. Settlement behaviors of
existing tunnel caused by obliquely under-crossing shield tunneling in close
upper tunneling caused the strata to settle on the ground proximity with small intersection angle. Tunn. Undergr. Space Technol. 97,
surface and to heave at the monitoring point MD, whereas 103258.
 Y. Hu et al. / Journal of Rock Mechanics and Geotechnical Engineering 14 (2022) 436e447 447

Liu, C., Zhang, Z.X., Regueiro, R.A., 2014. Pile and pile group response to tunnelling Zhang, T.Q., Taylor, R.N., Divall, S., Zheng, G., Sun, J., Stallebrass, S.E., Goodey, R.J.,
using a large diameter slurry shield e case study in. Shanghai. Comput. Geo- 2019. Explanation for twin tunnelling-induced surface settlements by changes
tech. 59, 21e43. in soil stiffness on account of stress history. Tunn. Undergr. Space Technol. 85,
Ma, S., Liu, Y., Lv, X., Shao, Y., Feng, Y., 2018. Settlement and load transfer mechanism 160e169.
of pipeline due to twin stacked tunneling with different construction se- Zhang, Z.X., Liu, C., Huang, X., Kwok, C.Y., Teng, L., 2016. Three-dimensional finite-
quences. KSCE J. Civ. Eng. 22, 3810e3817. element analysis on ground responses during twin-tunnel construction using
Marshall, A.M., Klar, A., Mair, R.J., 2010. Tunneling beneath buried pipes: view of soil the URUP method. Tunn. Undergr. Space Technol. 58, 133e146.
strain and its effect on pipeline behavior. J. Geotech. Geoenviron. 136 (12), Zhang, J.Z., Huang, H.W., Zhang, D.M., Zhou, M.L., Tang, C., Liu, D.J., 2021. Effect of
1664e1672. ground surface surcharge on deformational performance of tunnel in spatially
Mollon, G., Dias, D., Soubra, A.H., 2013. Probabilistic analyses of tunneling-induced variable soil. Comput. Geotech. 136, 104229.
ground movements. Acta Geotechnol. 8 (2), 181e199. Zheng, G., Lu, P., Diao, Y., 2015. Advance speed-based parametric study of greenfield
Ng, C.W.W., Hong, Y., Soomro, M.A., 2015. Effects of piggyback twin tunnelling on a deformation induced by EPBM tunneling in soft ground. Comput. Geotech. 65,
pile group: 3D centrifuge tests and numerical modelling. Geotechnique 65 (1), 220e232.
38e51. Zheng, G., Fan, Q., Zhang, T., Zheng, W., Sun, J., Zhou, H., Diao, Y., 2019. Multistage
Peck, R.B., 1969. Deep excavations and tunneling in soft ground. In: Proceedings of regulation strategy as a tool to control the vertical displacement of railway
the 7th International Conference on Soil Mechanics and Foundation Engineer- tracks placed over the building site of two overlapped shield tunnels. Tunn.
ing. Mexico City, Mexico, pp. 225e290. Undergr. Space Technol. 83, 282e290.
Qian, W.Q., Qi, T.Y., Zhao, Y.J., Le, Y.Z., Yi, H.Y., 2019. Deformation characteristics and
safety assessment of a high-speed railway induced by undercutting metro
tunnel excavation. J. Rock Mech. Geotech. Eng. 11 (1), 88e98.
Shahin, H.M., Nakai, T., Ishii, K., Iwata, T., Kuroi, S., 2016. Investigation of influence of
tunneling on existing building and tunnel: model tests and numerical simula- Dr. Huayang Lei has obtained her BSc degree in Engineering
tions. Acta Geotechnol. 11, 679e692. Geology at College of Changchun Geology, China, in 1994, and
Soomro, M.A., Mangi, N., Xiong, H., Kumar, M., Mangnejo, D.A., 2020. Centrifuge and MSc and PhD degrees in Geotechnical Engineering at Jilin
numerical modelling of stress transfer mechanisms and settlement of pile University, China, in 1997 and 2001, respectively. She is pro-
group due to twin stacked tunnelling with different construction sequences. fessor of geotechnical engineering and deputy director of
Comput. Geotech. 121, 103449. Department of Civil Engineering at Tianjin University. Dr. Lei
Standing, J.R., Selemetas, D., 2013. Greenfield ground response to EPBM tunnelling has hosted a number of key projects from Chinese govern-
in London Clay. Geotechnique 63 (12), 989e1007. ment, such as National Key Research and Development Pro-
Suwansawa, S., Einstein, H.H., 2007. Describing settlement troughs over twin tun- gram of China, National Key Basic Research and Development
nels using a superposition technique. J. Geotech. Geoenviron. 133 (4), 445e468. Program of China (973 Program), and National Natural Sci-
Sun, F.M., 2015. Analysis of Interaction between Small-Spacing Parallel Shield ence Foundation of China. She has authored or co-authored
Tunnel Excavations. MSc Thesis. Zhejiang University, Hangzhou, China (in more than 130 academic papers and compiled 9 tutorial
Chinese). books and monographs. Dr. Lei’s research interests cover
Wei, G., 2009. Study on calculation for width parameter of surface settlement soft soil engineering characteristics, foundation treatment,
trough induced by shield tunnel. Ind. Constr. 39 (12), 74e79 (in Chinese). soil constitutive relationship and urban underground engineering.