Available online at www.sciencedirect.

com

ScienceDirect
Underground Space 2 (2017) 73–87
www.elsevier.com/locate/undsp

Seismic analysis of long tunnels: A review of simplified and

unified methods
Haitao Yu a,b,⇑, Yong Yuan c, Antonio Bobet d,e
a
Key Laboratory of Geotechnical and Underground Engineering of Ministry of Education, Tongji University, Shanghai 200092, China
b
State Key Laboratory for GeoMechanics and Deep Underground Engineering, China University of Mining and Technology, Beijing 100083, China
c
State Key Laboratory of Disaster Reduction in Civil Engineering, Tongji University, Shanghai 200092, China
d
High-End Foreign Expert at Tongji University, Shanghai 200092, China
e
Lyles School of Civil Engineering, Purdue University, West Lafayette, IN 47907, USA

Received 2 January 2017; received in revised form 4 May 2017; accepted 14 May 2017
Available online 19 May 2017

Abstract

Seismic analysis of long tunnels is important for safety evaluation of the tunnel structure during earthquakes. Simplified models of
long tunnels are commonly adopted in seismic design by practitioners, in which the tunnel is usually assumed as a beam supported by the
ground. These models can be conveniently used to obtain the overall response of the tunnel structure subjected to seismic loading. How-
ever, simplified methods are limited due to the assumptions that need to be made to reach the solution, e.g. shield tunnels are assembled
with segments and bolts to form a lining ring and such structural details may not be included in the simplified model. In most cases, the
design will require a numerical method that does not have the shortcomings of the analytical solutions, as it can consider the structural
details, non-linear behavior, etc. Furthermore, long tunnels have significant length and pass through different strata. All of these would
require large-scale seismic analysis of long tunnels with three-dimensional models, which is difficult due to the lack of available comput-
ing power. This paper introduces two types of methods for seismic analysis of long tunnels, namely simplified and unified methods. Sev-
eral models, including the mass-spring-beam model, and the beam-spring model and its analytical solution are presented as examples of
the simplified method. The unified method is based on a multiscale framework for long tunnels, with coarse and refined finite element
meshes, or with the discrete element method and the finite difference method to compute the overall seismic response of the tunnel while
including detailed dynamic response at positions of potential damage or of interest. A bridging scale term is introduced in the framework
so that compatibility of dynamic behavior between the macro- and meso-scale subdomains is enforced. Examples are presented to
demonstrate the applicability of the simplified and the unified methods.
Ó 2017 Tongji University and Tongji University Press. Production and hosting by Elsevier B.V. on behalf of Owner. This is an open access article
under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

Keywords: Earthquake engineering; Tunnel structure; Seismic analysis; Simplified method; Multiscale method

Contents

Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74
Long tunnels in China . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

⇑ Corresponding author at: Key Laboratory of Geotechnical and Underground Engineering of Ministry of Education, Tongji University, Shanghai
200092, China.
E-mail address: [email protected] (H. Yu).
Peer review under responsibility of Tongji University and Tongji University Press.

http://dx.doi.org/10.1016/j.undsp.2017.05.003
2467-9674/Ó 2017 Tongji University and Tongji University Press. Production and hosting by Elsevier B.V. on behalf of Owner.
This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
74 H. Yu et al. / Underground Space 2 (2017) 73–87

The HZM immersed tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

The Qingcaosha water conveyor tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 76
The Longxi tunnel . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Simplified methods . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 77
Mass-spring-beam model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Beam-spring model. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 78
Multiscale methods. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Continuum-based multiscale coupling method. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 79
Discrete-continuum multiscale coupling method . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
Examples. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Case I: Simplified seismic analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
Case II: Unified seismic analyses . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 83
Closing remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Conflict of interest . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
Acknowledgements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 86

Introduction interaction analysis using finite element methods. The input

motions in these analyses are time histories emulating
Tunnels constitute a major part of civil infrastructure design response spectra, and the input motions are applied
and serve as public transportation facilities, sanitation, irri- to the boundaries of a ‘‘soil island” to represent vertically
gation utilities and storage infrastructure (Yu, Yuan, Qiao, propagating shear waves. The second approach assumes
et al., 2013). In seismically active areas, these tunnels are that the seismic ground motions induce a pseudo-static
under earthquake-induced risks. Recent events such as loading condition to the structure. This approach allows
the Kobe Earthquake in Japan (1995), the Duzce Earth- the development of analytical relationships to evaluate
quake in Turkey (1999), the Chi-Chi Earthquake in Taiwan the magnitude of seismically induced strains in the tunnel
(1999), the Bam Earthquake in Iran (2003) and the Wench- structure (Huo, Bobet, Fernández, & Ramı́rez, 2006;
uan Earthquake in China (2008) have shown that tunnels Penzien, 2000; Penzien & Wu, 1998). These relationships
are susceptible to irrecoverable damage due to seismic are based on the premise that tunnel structures under seis-
loading (Yu, Chen, Yuan, & Zhao, 2016; Yu, Yuan, Liu, mic loading will tend to deform due to the demand from
Li, & Ji, 2013), some quite extensively (Yu, Chen, Bobet, surrounding ground, and thus the structure is designed to
& Yuan, 2016). The observed damage provides sufficient accommodate the imposed deformations without loss of
evidence to suggest that the safety of tunnels in seismically its structural integrity.
active areas is still an important issue, but not well under- Design considerations for axial and bending deforma-
stood yet, or at least not well considered during design. tions are generally taken along the tunnel axis (Wang,
Tunnels are subjected to various forms of deformation 1993). The behavior of a tunnel is sometimes approximated
under seismic loading (Wang, 1993), namely: (a) ovaling to that of an elastic beam subjected to longitudinal deforma-
or racking shear deformation of the tunnel cross section tions imposed by the surrounding ground. Closed-form
due to the shear waves propagating normal or nearly nor- solutions have been proposed to estimate the strains and
mal to the tunnel axis; (b) axial compression and extension curvature of the tunnel subjected to a harmonic motion
generated by the components of seismic waves that pro- propagating at an angle to the tunnel axis (St. John &
duce motions parallel to the axis of the tunnel; and (c) lon- Zahrah, 1987). These solutions can be conveniently used
gitudinal bending caused by the components of seismic to obtain the overall response of tunnel structures subjected
waves producing particle motions perpendicular to the lon- to seismic loading. However, the solutions ignore the inertia
gitudinal axis. forces and the interaction between the tunnel and the sur-
Shear deformation of tunnels induced bythe vertically rounding ground, and thus may overestimate or underesti-
propagating shear waves has been widely studied by a mate the structure deformations, depending on the rigidity
number of researchers (Amorosi & Boldini, 2009; Bobet, of the structure relative to the ground, which has already
2003; Kattis, Beskos, & Cheng, 2003), and it has been been noted by a number of researchers (Bobet, 2003;
known to be the critical mode of deformation for tunnels Hashash, Hook, Schmidt, & Yao, 2001; Huo et al., 2006).
under seismic loading. The tunnel lining is generally simu- Clearly, analytical formulations are limited due to the
lated as a buried structure subjected to ground deforma- assumptions that need to be made to reach the solution. In
tions under a two-dimensional plane strain condition. most cases, the design will require a numerical method that
Two basic approaches are commonly used to estimate the does not have the shortcomings of the analytical solutions,
response of tunnels under shear deformation. One as it can consider the structural details, non-linear behavior,
approach is to carry out dynamic, nonlinear soil-structure etc. For the analysis of axial and bending deformations of
 H. Yu et al. / Underground Space 2 (2017) 73–87 75

tunnels, it is most appropriate to utilize three-dimensional in China. Given that all the tunnels are in seismic areas,
(3-D) models (Hatzigeorgiou & Beskos, 2010; Li & Song, the earthquake demand on the tunnels should be evaluated
2015; Stamos & Beskos, 1996). With high-performance com- to ensure the safety of the structures.
puters and large mass storage capabilities, it is now possible
to run large-scale 3-D dynamic FEM analyses for tunnels The HZM immersed tunnel
under seismic load (Ding, Jin, Guo, & Li, 2006). However,
if the entire tunnel is modeled including structural details The Hongkong-Zhuhai-Macau linkage (HZM linkage)
such as connection bolts between liner segments and joints, is a major infrastructure project starting construction in
the finite element model will be extremely large and compu- China. Crossing the Lingding Sea, the project links Hon-
tation will be very expensive, if not impossible, even with the gkong, with its industry and tourist attractions, to Zhuhai
most advanced supercomputer. city and Macao. The main structures of the HZM linkage
The objective of the paper is to illustrate the seismic are composed of a cable stayed bridge and an immersed
analysis of long tunnels with two different methods: the tunnel via two artificial islands. The total length of the link-
simplified method and the unified method. The mass- age is 35,600 m and the immersed tunnel is 5,664 m. The
spring-beam model, and the beam-spring model and its tunnel is composed of 33 concrete tunnel elements (see
analytical solution for long tunnels are discussed, as exam- Fig. 1(a)), and each element consists of 8 concrete tunnel
ples of the simplified method. The unified method is based segments, each 22.5 m long (see Fig. 1(b)). The cross sec-
on two multiscale approaches, i.e. one is a continuum- tion of the tunnel is a rough rectangle with 37.95 m in
based multiscale method coupling both coarse and refined width and 11.4 m in height (see Fig. 1(c)) that provides
finite element meshes, and the other is a discrete-continuum two vehicle barrels, and one barrel for pipelines and venti-
multiscale method coupling the discrete element and the lation. Segments are prefabricated one by one using rein-
finite difference model, to reduce computational cost and forced concrete, and then assembled into a tunnel
improve accuracy of the solution. Examples are given to element in a dockyard before shipping and immersing them
demonstrate the applicability of the methods. at the site. The sediments at the site are of marine origin
and include silt, clay and sand above bedrock.
Long tunnels in China This tunnel is, as of today, one of the largest long-
distance immersed tunnels in the world. Given that the tun-
Three examples of tunnels in China, namely the nel area is prone to earthquake activity, seismic perfor-
Hongkong-Zhuhai-Macau (HZM) immersed tunnel, the mance of the tunnel is investigated for the 120-year
Qingcaosha water conveyor tunnel, and the Longxi tunnel, design life of the structure. Furthermore, due to the spatial
are used to demonstrate the ongoing booming construction variation of earthquake motions along long tunnels
of long tunnels in China. Note that the three tunnels are (Hwang & Lysmer, 1981; Park, Sagong, Kwak, & Jeong,
built using three different construction methods, the 2009; Yu, Yuan, Qiao, et al., 2013), the influence of non-
Immersed Tunnelling Method, the Shield Tunnelling uniform seismic excitations on the immersed tunnel needs
Method and the New Austrian Tunnelling Method, respec- to be evaluated and considered for its seismic design (Yu
tively, which are the most widely used tunnelling methods et al., 2016; Yuan, Yu, Li, Yan, & Yuan, 2016).

Fig. 1. Layout of the HZM immersed tunnel (Unit: m) (Yu et al., 2016).
76 H. Yu et al. / Underground Space 2 (2017) 73–87

for the city of Shanghai, China. The tunnel has a total

length of 14 km. Its layout can be divided into three seg-
ments: the island, cross-river and the land segment. Three
working shafts are used to connect the different tunnel seg-
ments. The general plan and location of the shield tunnel,
including the working shafts, are shown in Fig. 2. The soils
through which the tunnel is excavated are mainly com-
posed of saturated clay, silt seams and sands.
The tunnel has been constructed using the shield method
and is, as of today, one of the largest long-distance shield
tunnels with medium-size diameter in the world. The tunnel
is composed of two lines: the East-line Tunnel and the
West-line Tunnel, as shown in Fig. 2. Each ring consists
of six segments, each with a length of 1.5 m. The inner
and outer diameters of the tunnel, for the island and land
segments, are 5.84 m and 6.8 m respectively, whereas for
the cross-river segment, the diameters are 5.5 m and
6.4 m. A typical cross section is shown in Fig. 3. A
staggered-jointed assembly is adopted between segments.
Fig. 2. Map of the Qingcaosha tunnel (Yu, Yuan, Qiao, et al., 2013). Seismic demand of the tunnel needs to be evaluated to
ensure the safety of the tunnel structure under the expected
The Qingcaosha water conveyor tunnel 7-degree seismic intensity (PGA of 0.1 g) in Shanghai (Yu,
Yuan, Qiao, et al., 2013). Therefore, the performance of
Water supply is one of the key issues in Megacities such the tunnel, not only its seismic response along its full
as Shanghai, China. The Qingcaosha water-conveyance length, but also the response of the segments and joints,
tunnel is a double-line shield tunnel used for water supply should be fully evaluated.

Fig. 3. Typical cross section of the Qingcaosha tunnel (Yu, Yuan, Qiao, et al., 2013).
 H. Yu et al. / Underground Space 2 (2017) 73–87 77

The Longxi tunnel

In western areas of China, a number of tunnels have suf-

fered severe damage during the Wenchuan strong earth-
quake in 2008, some quite extensively. The Longxi road
tunnel, located 2 km from the epicenter of the Wenchuan
earthquake, suffered significant damage and collapsed
when crossing a faulted zone, as shown in Fig. 4. The total
length of the tunnel is 3658 m, with a cross section com-
posed of two parallel twin tunnels, one for each direction,
separated 30 m between axes. The cross section of each of
the twin tunnels is shown in Fig. 5. The tunnel support was
designed following the New Austrian Tunneling Method. It
consists of a primary and a secondary support, and an
impermeable lining between the primary and secondary
supports. The cross section of each of the twin tunnels
has a total width of 9.20 m and a maximum height of
8.04 m. The pavement is made of concrete with a thickness
of 0.43 m. Near the Eastern end, at a distance of 620 m Fig. 5. Typical cross section of the Longxi tunnel (units in m) (Yu, Chen,
Bobet, et al., 2016).
from the portal, the tunnel crosses a pre-existing fault
(see Fig. 4). It is a reverse fault that, at the crossing of
the tunnel, has a dip angle of 82° and has an angle with tunnel, except for ground failure (Wang et al., 2001). This
the tunnel axis of 85°, i.e. the fault is quasi-perpendicular is in contrast to the observations made after the Wenchuan
to the tunnel. Earthquake that reported damage to several tunnels that
The Longxi tunnel, one of the most damaged tunnels ranged from minor cracking to even failure (Yu, Chen,
during the 2008 Wenchuan earthquake, suffered from small Bobet, et al., 2016). This conclusion should serve as a
to heavy damage both at the portal and inside the tunnel, warning for the need of better design of tunnels and other
while sections close to the fault completely collapsed, as underground structures in seismic areas (Yu, Chen, Bobet,
seen in Fig. 6. More detailed information of the tunnel et al., 2016).
damage can be found in (Yu, Chen, Bobet, et al., 2016).
Current design codes for seismic design of mountain tun- Simplified methods
nels focus on the stability of the portals and of sections
of the tunnel near slope surfaces, but do not consider Simplified methods for the seismic design of long tunnels
potential damage that may occur at other sections of the are commonly favored by practitioners. They can be conve-

Fig. 4. Location of the Longxi tunnel and earthquake epicenter (not to scale) (Yu, Chen, Bobet, et al., 2016).
78 H. Yu et al. / Underground Space 2 (2017) 73–87

Fig. 6. Collapse of the tunnel at the crossing of the fault (Yu, Chen,
Bobet, et al., 2016).

niently used to obtain the overall response of a tunnel Fig. 7. Mass-spring model for free field (Kiyomiya, 1995).
structure subjected to seismic loading. Two simplified mod-
els, namely the mass-spring-beam model, and the beam- A new method (Li, Yuan, Yu, & Yuan, submitted for
spring model and its analytical solution for seismic analysis publication), i.e. one-dimensional multiple-degree-of-
of long tunnels are discussed. freedom (MDOF) system, based on the equivalence of
modes, has been developed to represent the one-
Mass-spring-beam model dimensional layered soil deposit. A series of one-
dimensional MDOF systems are connected to constitute
The mass-spring-beam model is widely used in seismic a two-dimensional mass-spring model of the soil. The
design and analysis of long tunnels, especially for immersed equivalent mass-spring model is then used for the longitu-
tunnels (Kiyomiya, 1995). The mass-spring model was first dinal seismic analysis of long tunnels, as shown in Fig. 8.
proposed for the earthquake response analysis of immersed The modal equivalent MDOF system has the distinction
tunnels based on shaking table model tests and earthquake and advantage that it can characterize well the one-
observations of an immersed tunnel in Japan (Okamoto & dimensional ground properties, under shaking, such as nat-
Tamura, 1973). It relies on the assumptions that (1) the ural frequency and hysteretic damping. Comparisons of the
natural period of the ground is not influenced by the exis- amplification function of the modal equivalent MDOF sys-
tence of the tunnel, and (2) the motion of the ground is tem with the theoretical solution for the one-dimensional
induced by the shear vibration of soil strata and only the wave propagation provide excellent results (Li et al.,
fundamental mode is taken into consideration. submitted for publication).
Fig. 7 is a sketch of the mass-spring model (Kiyomiya,
1995). The surface layer of the soil along the tunnel axis Beam-spring model
is divided into a number of soil slices. Each slice is repre-
sented by an equivalent mass-spring system that consists The pseudo-static or dynamic analysis based on the
of a mass, a spring and a dashpot connecting the mass to beam-spring model is a simple method for the seismic anal-
the base rock. The neighboring masses are then connected
to each other along the tunnel axis by springs and dashpots
to simulate the connection between the adjacent soil slices.
By solving the dynamic equilibrium equation of the model,
the ground displacements at the positions of the soil masses
and tunnel can be evaluated. Assuming the tunnel as a
beam supported by soil springs, the seismic response of
the tunnel subjected to the obtained ground motions can
be calculated through a dynamic time-history analysis.
Note that the mass-spring model is established with the
assumption that the ground displacement is dominated
by the fundamental shear vibration. Clearly, the ground
response calculated may not be accurate when different soil Fig. 8. Mass-spring-beam model for longitudinal seismic analysis of long
layers are contained in a soil slice. tunnels.
 H. Yu et al. / Underground Space 2 (2017) 73–87 79
(
ysis of long tunnels (Anastasopoulos et al., 2007; Hashash, 0; ðx > VtÞ
Tseng, & Krimotat, 1998). In this model, the tunnel is F ðx; tÞ ¼ 
 ð1Þ
P sin 2pX t  Vx ; ðx 6 VtÞ
assumed as a beam supported on the ground by soil springs
(and dashpots) representing the soil-structure interaction, where t, V , X and P are time, velocity, frequency and
as shown in Fig. 9. The seismic input is the free-field amplitude of the loads, respectively.
motion at the depth of the tunnel. Generally, two basic The displacement response of the tunnel is (Yu, Cai,
approaches, analytical and numerical, are employed for Guan, et al., 2016):
beams resting on elastic or viscoelastic foundations. Z þ1 Z t Z Vs 
P cos½uðr  xÞ
Numerical methods such as the finite element method are yðx; tÞ ¼ sin
useful to analyze complex structures; however, both time pqA 0 0 0 bðuÞ
h 

and spatial increments have to be considered in the r i aðtsÞ

 2pX s  e sin½bðuÞðt  sÞ drdsdu ð2Þ
dynamic algorithms, and thus these methods are time- V
consuming and the numerical accuracy depends on the
where r, u and s are the intermediate variables of the inte-
integration algorithm. Computational efficiency can be
gration transforms. The new variables a and bðuÞ are
improved if analytical solutions are available (Yu &
defined as
Yuan, 2014).
Yu, Cai, Guan, and Yuan (2016) proposed an analytical C
a¼ ð3Þ
solution for the dynamic response of long lined tunnels 2qA
subjected to travelling loads, taking into account both the sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
EI 4 K C2
inertia forces and the interaction between the soil and the bðuÞ ¼ u þ  ð4Þ
structure. Several assumptions were made: the tunnel was qA qA 4ðqAÞ2
assumed to be infinitely long with a uniform cross- The velocity vðx; tÞ and acceleration aðx; tÞ responses of
section; its behavior was linear elastic; the surrounding soil the tunnel structure can be obtained as the first and second
was assumed to be isotropic and homogeneous and behave derivatives of the displacement response with respect to
as viscoelastic; the travelling loads were plane harmonic time t. Analytical solutions for the bending moment
loads and propagated parallel to the tunnel axis. A solution Mðx; tÞ and shear force Qðx; tÞ of the tunnel can be obtained
was obtained using Fourier transform to simplify the gov- taking the second and third derivatives of the displacement
erning equation in the space domain, whereas the Laplace response with respect to the coordinate x, and each multi-
transform was employed to reduce the equation in the time plied by ðEIÞ.
domain. The governing equation, based on the integration
transform, was expressed as an algebraic equation so that
the solution was obtained in the frequency domain. Multiscale methods
Finally, the convolution theorem was employed to convert
the solution into the time domain. Simplified seismic analyses may not be able to incorpo-
Fig. 9 depicts the coordinate system and dimensions rate enough details of the structure or the damage at the
associated with an infinite long lined tunnel. The tunnel location of potential failure. To address this problem,
has constant stiffness EI and mass per unit length qA, where two multiscale approaches, namely the continuum-based
E = Young’s modulus; I = moment of inertia of the tunnel multiscale method and the discrete-continuum multiscale
cross section; q = density of the tunnel liner; and A = area method, are developed to capture not only the seismic
of the cross section of the tunnel. The tunnel is supported response along the entire tunnel length, but also detailed
by a viscoelastic foundation with constant spring stiffness structural response of the liner segments and their joints.
K and viscous damping C.
The solution is given by the function yðx; tÞ, which is the Continuum-based multiscale coupling method
vertical deflection of the tunnel and F ðx; tÞ, which is the
harmonic travelling load propagating parallel to the tunnel The continuum-based multiscale approach proposed by
axis (see the x-axis in Fig. 9). The wave-passage loads can Yu, Yuan, and Bobet (2013) couples FEM calculations
be expressed as: with coarse and fine meshes. The coarse-scale mesh is
employed to capture seismic response characteristics of
the entire system, whereas the fine-scale mesh is employed
to describe in detail the dynamic response at positions of
potential damage or interest. This multiscale method can
significantly reduce the computational load. It adequately
covers wide areas of the entire tunnel-soil system, while
including details in key locations of the tunnel. Further-
more, the total number of finite elements can be kept
Fig. 9. Long lined tunnel on a viscoelastic foundation subjected to within the range of computation capacity of the machine
dynamic loads. used for the simulations.
80 H. Yu et al. / Underground Space 2 (2017) 73–87

X 2
X 2
The multiscale method consists of two steps. In the first dI ¼ ½d iI  ¼ uCi ðXI Þ  uRi ðXI Þ
step, a regular, e.g. coarse, 3-D Finite Element Model is i i
built based on geological data, tunnel geometry, construc- " #2
X X X
tion process, etc. Non-linear soil and/or structure models ¼ N CJ ðXI ÞuCiJ  N RJ ðXI ÞuRiJ ¼0 ð6Þ
are included in the analysis as well as contact interfaces. i J J
The objective of this first step is to determine the seismic
where dI is the representation of the displacement bound-
response characteristics of the system and to identify areas
ary conditions on the overlapping subdomain; the super-
where detailed, e.g. refined, analysis is needed. In the sec-
scripts C and R correspond to the coarse and refined
ond step, the coarse mesh in those areas of interest identi-
subdomains, respectively. The refined displacements are
fied in the first step is replaced by a refined mesh and a new
required to conform to the coarse displacements at the dis-
simulation with the composite, i.e. coarse and fine, mesh is
crete nodal positions of the refined model. The constraints
carried out. This can be done while the coarse mesh still
are applied to all components of the displacements.
captures the seismic response characteristics of the system.
In the coupling method, the constraints are appended to
A difficulty that arises in multiscale couplings is that
the dynamic equations with Lagrange multipliers. The total
high frequency waves may be spuriously reflected at the
Hamiltonian can be described as
fine/coarse interface, as noted by a number of researchers X
such as Holmes and Belytschko (1976). This is because H L ¼ H þ kT d ¼ H þ kTI dI ð7Þ
the time step used for the entire domain depends on the size I

of the smallest element, and as a result one may observe where kI ¼ fkiI g is a vector of Lagrange multipliers, with
spurious reflections of high frequency waves at locations components corresponding to the components of the dis-
where there is a change of mesh density. An overlapping placement of the refined node I.
domain between coarse and refined meshes is defined to The equations of motion with the Lagrange multipliers
solve the problem. A schematic of the coupling method is are
shown in Fig. 10. The complete domain in the initial con- (
figuration is denoted by X0 . The domain is subdivided into M
I €uC ¼ f extC  f intC  f dC  f LC in XC
I I I I I 0
ð8Þ
two subdomains: the refined subdomain denoted by XR0 ,
I €uRI ¼ f extR
m I  f intR
I  f dR
I  fI
LR
in XR0
and the coarse subdomain, denoted by XC0 . The overlap int
where f ext
I and f I are the external and internal nodal forces
0 ; ‘0 denotes the
g
of these two domains is denoted by Xint
respectively, f I is the damping node force, f LI is the force
d
boundary between the refined and the overlap subdomains
and ‘g1 denotes the boundary between the overlap and the due to the constraints enforced by the Lagrange multipli-
ers; the superscripts C and R correspond to the coarse
coarse subdomains.
and refined subdomains, respectively; M I and mI are the
In the method, the total energy is taken as a linear com-
lumped mass of coarse and refined subdomains at node I,
bination of the refined and coarse mesh energies. A scaling
respectively.
parameter g is introduced in the overlapping subdomain.
The Hamiltonian for the complete domain is taken as a lin- The forces f LC I and f LRI are due to the constraints
ear combination of the refined and coarse subdomain. That enforced by the Lagrange multipliers and are given by
is, X @dJ X
I ¼
f LC kTJ C ¼ kTJ DCJI ð9Þ
@u
H ¼ ð1  gÞH R þ gH C ð5Þ J I J
X @dJ X
To keep the continuity of the two divided subdomains, I ¼
f LR kTJ R ¼ kTJ DRJI ð10Þ
J
@u I J
the refined model and the coarse model are constrained
on the overlapping subdomain Xint where N IJ ¼ N I ðXJ Þ, and
0 by:  
@dJ  C  @dJ
DJI ¼
C
¼ N JI I ; DJI ¼
R
¼ ½N RJI I ð11Þ
@uCI @uRI
where N I is the shape function.
The explicit central difference method (CDM) is adopted
to solve for the motion. First, trial displacements and
velocities are obtained neglecting the constraints. Second,
Lagrangian multipliers are obtained to satisfy the con-
straints. Finally, updated final displacements and velocities
are computed. The steps required for FEM multiscale cou-
pling are as follows:

(1) Initial geostatic stress conditions are computed using

Fig. 10. The continuum-based multiscale model for long tunnels. the dynamic relaxation method with the coarse mesh.
 H. Yu et al. / Underground Space 2 (2017) 73–87 81

(2) Seismic simulation of the coarse model is performed. particles. The meshes are employed to capture dynamic
(3) Sections of interest are identified after (2). response characteristics of the system, whereas the particles
(4) Multi-scale FEM is implemented by replacing the describe the dynamic response at the mesoscale or micro-
coarse model with the refined model in the sections scale. An overlapping domain is created between the finite
of interest. difference meshes and the discrete particles, and a bridging
(5) Numerical simulation of the multi-scale model under scale term is introduced to ensure the compatibility of
seismic loading is carried out, using the dynamic energy in the overlapping domain. In the overlapping
explicit algorithm (CDM). domain (Fig. 11), the constraints between mesh and parti-
(6) Multi-scale FEM dynamic computation ends when cle subdomains are imposed to the motion equations using
the accumulated time is greater than the maximum Lagrange multipliers, which are developed in a consistent
time tmax . manner from the energy potential. The proposed method
does not result in spurious wave reflections and does not
need additional filtering or damping in the overlapping
Discrete-continuum multiscale coupling method
domain between the FDM meshes and the DEM particles.
The equations of motion for the coupling method,
Seismic analyses of long tunnels are generally performed
including the Lagrange multipliers are (Li et al., 2015)
using continuum-based methods such as FEM or FDM
(finite difference method) because of their suitability for a
J €uJ ¼ f ext C  f int C  f LC ; in XC
M ð12Þ
J J J 0
continuous formulation of the problem considered. How-
ever, to resolve local phenomena such as crack propagation
I €dI ¼ f ext
m I
P
 f int
I
P
 f LP
I ; in X0
P
ð13Þ
and progressive failure at positions of potential damage, a
J and m
where M
I are the lumped mass of node J and par-
very refined model is desired. Furthermore, current consti-
tutive models establish stress-strain relationships at the ticle I, respectively; f ext
J
C
and f ext
I
P
are the external nodal
macroscopic scale but may lose the capability of capturing force and particle force, respectively; and f intC J and f intP
I
response at the microscale. An alternative to those methods are the internal nodal force and particle force, respectively.
is the discrete-based approach formulation, to describe the The forces f LC J and f LP
I are caused by the constraints,
mesomechanical or micromechanical response of materials enforced by Lagrange multipliers, and given by (Li et al.,
and local failure processes. However, because of the 2015)
demanding computational cost of discrete methods, it is X @g X
impossible to realistically model large engineering prob- J ¼
f LC kK K ¼ kK GCKJ ð14Þ
@uJ
lems with the current capacity. Fortunately, in many appli- K K

cations, the region of interest is localized, and complex or X @g X

I ¼
f LP kK K ¼ kK GPKI ð15Þ
highly nonlinear response is limited to well-defined areas. K
@u I K
Therefore, multiscale methods coupling discrete-based
approaches and continuum-based approaches have the where kK is a vector of Lagrange multipliers, gK is the rep-
advantage of both techniques and provide an alternative resentation of the kinematic constraints in the overlapping
solution to model problems efficiently. domain and
A novel and reliable multiscale coupling method for  
@gK @gK
dynamic simulations between continuum and discrete mod- GCKJ ¼ ¼ ½N JK I; GPKI ¼ ¼ ½dIK I ð16Þ
@uJ @uI
els is proposed by Li, Yu, Wang, Xia, and Chen (2015).
Two numerical methods, the continuum-based FDM and and N IK ¼ N I ðXK Þ is the interpolation function.
the discrete-based DEM (discrete element method), are To solve the coupled dynamic system with the Lagrange
coupled. The process starts with the discretization of the multiplier method, an explicit algorithm is developed based
entire domain with finite difference meshes and discrete on the central difference method. The following describes

Fig. 11. Bridging domain coupling discrete and continuum regions (Li et al., 2015).
82 H. Yu et al. / Underground Space 2 (2017) 73–87

the procedure to obtain displacements and velocities at propagating along the positive axis. The travelling har-
time n + 1, assuming that displacements and velocities monic load applied at time t is described in Eq. (1). The
are known at time steps n and n + 1/2. First, trial displace- parameters of the load are as follows: amplitude
ments and velocities are obtained, neglecting the con- P ¼ 21 kN; frequency X ¼ 2 Hz; and wave velocity
straints. Second, Lagrangian multipliers are obtained to V ¼ 200 m=s.
satisfy the constraints. Finally, updated final displacements The analytical solution is provided in integral form in
and velocities are computed. A user-defined subroutine is Eq. (2), which can be solved in combination with the Gaus-
developed based on the two commercial packages, fast sian integral. Fig. 12 shows the comparison of the bending
Lagrangian analysis of continua (FLAC) and particle flow moment and shear force responses at point x ¼ 500 m of
code (PFC). The advantage of the multiscale continuum- the tunnel under travelling and uniform harmonic loads.
discrete coupling approach is that it allows modeling of The figure indicates that both the bending moment and
local physical phenomena at the mesoscopic or microscopic shear force responses of the tunnel under travelling loads
scales in very large systems. are significantly larger than those under uniform loads.
This can be taken as an indication that, for long tunnels,
Examples the wave passage effect would dramatically amplify the
structural response; therefore, uniform loads should not
Several examples using the simplified and unified seismic be used in the seismic design, but rather appropriate con-
analyses of the Qingcaosha water-conveyance tunnel sideration should be given to travelling loads.
(Fig. 2) are presented to illustrate the applicability of the It is of interest to observe how the velocity and fre-
methods. More detailed information of the tunnel is pro- quency of the travelling load influence the response of the
vided in Section 2.2. To provide a better understanding tunnel. Fig. 13 shows the bending moments and the shear
and yet limit the number of analyses, the following discus- forces at point x ¼ 500 m for the travelling load, with
sion is limited to the beam-spring model and multiscale velocity 200 m/s, 400 m/s, 600 m/s, 800 m/s and constant
model, coupling coarse-refined meshes, for large-scale seis- frequency 2 Hz. It can be seen from the figure that the
mic analyses. velocity of the travelling loads has significant influence on
the response of the tunnel. In this case, the dynamic
response for the wave velocity 200 m/s is much larger than
Case I: Simplified seismic analyses the other waves, and thus it can be considered as the critical
condition for the tunnel design.
Due to the aforementioned limitations of the simplified Fig. 14 is a plot similar to Fig. 13, also showing the
method, only the cross-river segment of the tunnel (Fig. 2) bending moments and shear forces at x ¼ 500 m, for the
is considered here, which is assumed as an infinitely long travelling load with frequency 2 Hz, 4 Hz, 6 Hz, 8 Hz and
lined tunnel with external and inner diameters 6.4 m and constant wave velocity 200 m/s. The figure indicates that
5.5 m, respectively. The tunnel is assumed linear elastic, the bending moment of the tunnel structure increases with
homogeneous and isotropic, with mass density q ¼ decreasing wave frequency. This result suggests that the
2500 kg=m3 and Young’s modulus E ¼ 3:55  1010 Pa. influence of low frequency waves should be considered
The foundation is linear elastic, with spring stiffness for the seismic design of long tunnels.
K ¼ 3:5  107 N=m2 and viscous damping C ¼ 0. A travel- Regarding the advantages (and limitations) of analytical
ling harmonic load is imposed to the tunnel structure from solutions over numerical routines applied to practical engi-
the origin of the coordinate system x ¼ 0 (see in Fig. 9), neering problems, the following comments apply: although

Fig. 12. Response of the tunnel liner under travelling and uniform harmonic loads, at x ¼ 500 m: (a) bending moment; and (b) shear force.
 H. Yu et al. / Underground Space 2 (2017) 73–87 83

Fig. 13. Response of the tunnel liner for travelling harmonic waves with velocity 200 m/s, 400 m/s, 600 m/s, 800 m/s and constant frequency 2 Hz, at
x ¼ 500 m: (a) bending moment; and (b) shear force.

Fig. 14. Response of the tunnel liner for travelling harmonic waves with frequency 2 Hz, 4 Hz, 6 Hz, 8 Hz and constant velocity 200 m/s, at x ¼ 500 m: (a)
bending moment; and (b) shear force.

the assumption of linear elasticity limit the applicability of

the analytical solution presented in this article, it has the
distinct advantage of providing a rational framework to
carry out a systematic parametric analyses with the most
relevant parameters.

Case II: Unified seismic analyses

The multiscale method, coupling coarse-refined meshes,

is used to estimate the large-scale seismic response of the
Qingcaosha water-conveyance tunnel (Fig. 2) in Shanghai.
To highlight the resolution of the method, the dynamic
behavior of the tunnel segments and joints is obtained as
part of the solution. Fig. 15. Coarse-scale soil and tunnel model.
Fig. 15 depicts the coarse-scale soil and tunnel models.
The coarse-scale discretization is done with eight-node hex-
ahedral solid elements. The full size of the soil-tunnel model ments. The bottom boundary is allowed to have horizontal
is 12,660 m  2509 m  300 m (length  width  depth), displacements, but not vertical displacements; the lateral
and comprises a total of 1,323,978 nodes and 3,515,026 ele- boundaries are taken in the model as free boundaries. In
84 H. Yu et al. / Underground Space 2 (2017) 73–87

Fig. 16. Mesh of the multiscale tunnel model: (a) coarse-refined mesh; and (b) detail of the refined mesh for a liner segment and bolts.

the model, the closest distance from the tunnel to the lateral and longitudinal bolts are placed in tension. The pre-
boundary is more than ten times the spacing between the tension force is about 176 kN. The refined mesh for the
east and the west tunnels. Because the lateral boundaries liner segments is composed of eight-node hexahedral ele-
are far from the tunnel structure, the influence of wave ments, the same elements used for the coarse mesh. The
reflection caused by the free boundaries on the tunnel is neg- length of the overlap domain between the coarse and
ligible. The adequacy of the type of boundary and mesh size refined discretization is 1.5 m (Fig. 16). The prestressed
was verified by running a number of preliminary numerical bolts are simulated with two-node link elements (Fig. 16).
tests, where the lateral boundaries of the discretization were The force generated by the bolt is nonzero if and only if
placed at different distances from the tunnel. It was decided the bolt is in tension. Since the nodes of the link element
that the size of the discretization was acceptable when free- have the same degrees of freedom as the solid elements,
field conditions were recovered in the area between the tun- the connection between bolts and tunnel segments is
nel and the boundaries. consistent.
The multi-scale model is implemented by replacing the An elastic material model is assumed for both the con-
coarse model with the refined model at positions of poten- crete (liner segments) and steel (connection bolts). The
tial damage; these include the liner segments at particular nonlinear soil model (R-O model) is used for the entire
cross sections and the bolts that connect the segments ground, i.e. for both the coarse and refined meshes. The
(Fig. 16). In the refined model, each of the six segments material parameters for segments, bolts and soil properties
in each ring is connected by M36 bolts (circumferential used in the model can be found in Yu, Yuan, Bobet, 2013.
bolts), with a total of 24 bolts; the joints between the seg- The Shanghai artificial seismic wave record has been
ments of two adjacent rings have M30 bolts with a total imposed to the bottom of the discretization, which corre-
of 16 bolts (longitudinal bolts). Both the circumferential sponds to bedrock (Fig. 17). In the numerical model, the

0.10
0.55

0.50
Ratio between max. principal stress

0.05
0.45
Acceleration (g)

and tensile strength

0.40
0.00
0.35

0.30
-0.05

0.25

-0.10 0.20
0 2 4 6 8 10 12 14 16 18 20
0 5 10 15 20
Time (s) Time (s)

Fig. 17. Acceleration time history of the Shanghai artificial seismic wave. Fig. 18. Maximum tension in bolts.
 H. Yu et al. / Underground Space 2 (2017) 73–87 85

Fig. 19. Contour plots of stresses in the tunnel liner segments at time 5.2 s. (a) maximum principal stress; (b) minimum principal stress. Max. and Min.
ratio in the figures denote the maximum and minimum ratio of maximum principal stress/tensile strength or the maximum and minimum ratio of
minimum principal stress/compressive strength.

seismic input is applied to all the nodes at the bottom of the stresses, i.e. maximum tension and maximum compression,
discretization, i.e. all the nodes, at a given time, have accel- occur at the contact between segments, as shown in Fig. 19,
eration in both the longitudinal and transverse directions. which is a contour plot of the maximum and minimum
The simulations have been carried out using 64 CPUs of principal stresses. The extreme values are due to the stress
the Magic Cube supercomputer at the Shanghai Supercom- concentration induced by contact between tunnel segments.
puter Center. Each run takes about 75 h to complete. The Also, since the tensile strength of the concrete is reached
following is a discussion of the results obtained with the during the earthquake, cracking may be expected.
refined model at one of the potentially most critical sec- In reality, not only the ovaling deformation of the cross
tions along the tunnel. sections but also the deformation of joints is of great con-
Fig. 18 illustrates the ratio between maximum principal cern. Figs. 20 and 21 are plots of the maximum opening
stress and tensile strength of the bolts connecting the seg- and the maximum shearing dislocation, respectively, of
ments at the most critical section of the east-line tunnel. any of the circumferential joints. The figures show that
What is shown is the maximum stress at any given time the maximum opening of the circumferential joint is
in any of the bolts. The maximum tension, of about 2.3 mm and the maximum shear dislocation of the circum-
240 MPa, occurs at about 8.2 s within the earthquake, ferential joint is 2.4 mm. These values are smaller than the
which can be considered acceptable given that the maxi- allowable upper limits of 3–5 mm for joint opening and 4–
mum allowable stress is 480–640 MPa, depending on the 5 mm for joint dislocation, based on the Shanghai code for
grade of the bolts used (Yu, Yuan, Qiao, et al., 2013). It foundation design (DGJ08-11-1999), and so the perfor-
is interesting to note that the largest values of the principal mance of the tunnel under the seismic acceleration input
is acceptable.

3
2.5

2 2.0
circumferential joints (mm)

Maximum dislocation amount of

circumferential joints (mm)
Maximum opening of

1 1.5

1.0
0

0.5
-1
0.0

-2
-0.5

-3
-1.0
0 2 4 6 8 10 12 14 16 18 20 0 2 4 6 8 10 12 14 16 18 20
Time (s) Time (s)

Fig. 20. Maximum opening of circumferential joints. Fig. 21. Maximum shearing dislocation of circumferential joints.
86 H. Yu et al. / Underground Space 2 (2017) 73–87

Closing remarks Acknowledgements

Two different methods, the simplified and the unified The research has been supported by the National Natu-
method, are introduced, which are employed for seismic ral Science Foundation of China (Grant No. 51678438 &
analysis of long tunnels. The simplified methods such as 51478343), the Shanghai Rising-Star Program
the beam model provide effective tools for practitioners. (17QC1400500), and the Shanghai Committee of Science
They allow readily identification of the variables controlling and Technology (Grant No. 16DZ1200302 &
the magnitude of the distortions and thus provide an insight 16DZ1201904). The authors acknowledge the support from
into the behavior of the structure. Furthermore, the simpli- the Fundamental Research Funds for the State Key Labo-
fied method and its solutions are invaluable to obtain a bet- ratory for GeoMechanics and Deep Underground Engi-
ter understanding of the interplay that exists between neering, China University of Mining & Technology
dynamic loads, viscoelastic foundation and tunnel structure, (SKLGDUEK1723).
to identify what are the most critical parameters for the
problem, and to provide first estimates or even a preliminary
References
design. An added advantage is that they can be used with lit-
tle cost to conduct sensitivity analysis. However, the simpli- Amorosi, A., & Boldini, D. (2009). Numerical modeling of the transverse
fied method may not be able to capture response and damage dynamic behavior of circular tunnels in clayey soils. Soil Dynamics and
in structural details, in elements or at locations of potential Earthquake Engineering, 29, 1059–1072.
Anastasopoulos, I., Gerolymos, N., Drosos, V., Kourkoulis, R., Geor-
failure, due to the simplified assumptions taken to include garakos, T., & Gazetas, G. (2007). Nonlinear response of deep
the tunnel structure and the soil-tunnel interaction. immersed tunnel to strong seismic shaking. Journal of Geotechnical and
Thus, a novel and reliable multiscale coupling method is Geoenvironmental Engineering, 133(9), 1067–1090.
Bobet, A. (2003). Effect of pore water pressure on tunnel support during
proposed to solve the problem. The multiscale approach static and seismic loading. Tunneling and Underground Space Technol-
couples continuum-based FEM calculations with coarse ogy, 18, 377–393.
and fine meshes. The coarse-scale mesh is employed to cap- DGJ08-11-1999. Shanghai code for design of foundation, Management
Office for Shanghai Technical Standard Project: Shanghai, 1999 (in
ture seismic response characteristics of the integral system, Chinese).
whereas the fine-scale mesh describes in detail the dynamic Ding, J. H., Jin, X. L., Guo, Y. Z., & Li, G. G. (2006). Numerical
response at the locations of potential damage or of interest. simulation of large-scale seismic response analysis of immersed tunnel.
Engineering Structures, 28, 1367–1377.
An overlapping subdomain is created between the refined Hashash, Y. M. A., Tseng, W. S., & Krimotat, A. (1998). Seismic soil-
and the coarse meshes, where Lagrange multipliers are structure interaction analysis for immersed tube tunnels retrofit.
used to enforce kinematic constraints. The coupling Geotechnical Earthquake Engineering Soil Mechanics III 2, ASCE
Geotechnical Special Publication No.75 (pp. 1380-1391).
method bridging domain is implemented to minimize, or Hashash, Y. M. A., Hook, J. J., Schmidt, B., & Yao, J. I. (2001). Seismic
at least reduce, spurious wave reflections at the coarse- design and analysis of underground structures. Tunneling and Under-
fine mesh interface. The multiscale approach is further ground Space Technology, 16, 247–293.
Hatzigeorgiou, G. D., & Beskos, D. E. (2010). Soil-structure interaction
extended to couple discrete-based DEM and continuum- effects on seismic inelastic analysis of 3-D tunnels. Soil Dynamics and
based FDM for dynamic analysis. The advantage of the Earthquake Engineering, 30, 851–861.
multiscale continuum-discrete coupling approach is that Holmes, N., & Belytschko, T. (1976). Postprocessing of finite element
transient response calculations by digital filters. Computer & Struc-
it allows modeling of local physical phenomena occurring tures, 6, 211–216.
at the meso- or micro-scales in very large systems. Huo, H., Bobet, A., Fernández, G., & Ramı́rez, J. (2006). Analytical
This multiscale modeling method can significantly solution for deep rectangular structures subjected to far-field shear
stresses. Tunnelling and Underground Space Technology, 21, 613–625.
reduce the computation load. It adequately covers wide Hwang, R. N., & Lysmer, J. (1981). Response of buried structures to
simulation areas of the entire tunnel-soil system and may traveling waves. Journal of Geotechnical Engineering Division, 107(2),
also include element details in key locations of the tunnel. 183–200.
Kattis, S. E., Beskos, D. E., & Cheng, A. H. D. (2003). 2D dynamic
Furthermore, the total number of elements can be kept response of unlined and lined tunnels in poroelastic soil to harmonic
within the range of computation capacity of the supercom- body waves. Earthquake Engineering and Structure Dynamics, 32,
puter used for the simulations. Examples are presented to 97–110.
Kiyomiya, O. (1995). Earthquake-resistant design features of immersed
demonstrate the applicability of the multiscale method in tunnels in Japan. Tunneling and Underground Space Technology, 10(4),
large-scale seismic analysis of long tunnels. The success 463–475.
using the method in a long and complex tunnel reinforces Li, C., Yuan, J. Y., Yu, H. T., & Yuan, Y. (2016). Mode-based equivalent
ground model for longitudinal seismic analysis of long tunnels. Soil
the notion that full 3D simulations can be performed; with Dynamics and Earthquake Engineering (submitted for publication).
an effective selection of the areas where a refined mesh is Li, P., & Song, E. X. (2015). Three-dimensional numerical analysis for the
needed, results can be obtained even if the computational longitudinal seismic response of tunnels under an asynchronous wave
input. Computers and Geotechnics, 63, 229–243.
capacity available is limited. Li, M. G., Yu, H. T., Wang, J. H., Xia, X. H., & Chen, J. J. (2015). A
multiscale coupling approach between discrete element method and
Conflict of interest finite difference method for dynamic analysis. International Journal for
Numerical Methods in Engineering, 102, 1–21.
Okamoto, S., & Tamura, C. (1973). Behaviour of subaqueous tunnels
No potential conflict of interest was reported by the during earthquakes. Earthquake Engineering and Structural Dynamics,
authors. 1(3), 253–266.
 H. Yu et al. / Underground Space 2 (2017) 73–87 87

Park, D., Sagong, M., Kwak, D. Y., & Jeong, C. G. (2009). Simulation of Yu, H. T., Chen, J. T., Yuan, Y., & Zhao, X. (2016). Seismic damage in
tunnel response under spatially varying ground motion. Soil Dynamics mountain tunnels due to the Wenchuan strong earthquake. Journal of
and Earthquake Engineering, 29, 1417–1424. Mountain Sciences, 13(11), 1958–1972.
Penzien, J. (2000). Seismically induced raking of tunnel linings. Earth- Yu, H. T., Yuan, Y., Xu, G. P., Su, Q. K., Yan, X., & Li, C. (2016). Multi-
quake Engineering and Structure Dynamic, 29, 683–691. point shaking table test for long tunnels subjected to non-uniform
Penzien, J., & Wu, C. L. (1998). Stresses in linings of bored tunnels. seismic loadings - Part II: Application to the HZM immersed tunnel.
Earthquake Engineering and Structure Dynamics, 27, 283–300. Soil Dynamics and Earthquake Engineering. http://dx.doi.org/10.1016/
St. John, C. M., & Zahrah, T. F. (1987). Aseismic design of underground j.soildyn.2016.08.018.
structures. Tunneling and Underground Space Technology, 2(2), Yu, H. T., & Yuan, Y. (2014). Analytical solution for an infinite Euler-
165–197. Bernoulli beam on a viscoelastic foundation subjected to arbitrary
Stamos, A. A., & Beskos, D. E. (1996). 3-D seismic response analysis of dynamic loads. Journal of Engineering Mechanics, ASCE, 140(3),
long lined tunnels in half-space. Soil Dynamics and Earthquake 542–551.
Engineering, 15, 111–118. Yu, H. T., Yuan, Y., & Bobet, A. (2013). Multiscale method for long
Wang, J. N. (1993). Seismic design of tunnels: A state-of-the-art approach. tunnels subjected to seismic loading. International Journal for Numer-
New York: Parsons Brinckerhoff Quade & Douglas Inc.. ical and Analytical Methods in Geomechanics, 37(4), 374–398.
Wang, W. L., Wang, T. T., Su, J. J., Lin, C. H., Seng, C. R., & Huang, T. Yu, H. T., Yuan, Y., Liu, X., Li, Y. W., & Ji, S. W. (2013). Damages of
H. (2001). Assessment of damage in mountain tunnels due to the the Shaohuoping road tunnel near the epicentre. Structure and
taiwan chi-chi earthquake. Tunneling and Underground Space Technol- Infrastructure Engineering, 9, 935–951.
ogy, 16(3), 133–150. Yu, H. T., Yuan, Y., Qiao, Z. Z., Gu, Y., Yang, Z. H., & Li, X. D. (2013).
Yu, H. T., Cai, C., Guan, X. F., & Yuan, Y. (2016). Analytical solution Seismic analysis of a long tunnel based on multi-scale method.
for long lined tunnels subjected to travelling loads. Tunnelling and Engineering Structures, 49, 572–587.
Underground Space Technology, 58, 209–215. Yuan, Y., Yu, H. T., Li, C., Yan, X., & Yuan, J. Y. (2016). Multi-point
Yu, H. T., Chen, J. T., Bobet, A., & Yuan, Y. (2016). Damage observation shaking table test for long tunnels subjected to non-uniform seismic
and assessment of the Longxi tunnel during the Wenchuan earthquake. loadings - Part I: Theory and validation. Soil Dynamics and Earthquake
Tunnelling and Underground Space Technology, 54, 102–116. Engineering. http://dx.doi.org/10.1016/j.soildyn.2016.08.017.