Engineering Geology: Yusheng Shen, Bo Gao, Xiaoming Yang, Shuangjiang Tao
Engineering Geology: Yusheng Shen, Bo Gao, Xiaoming Yang, Shuangjiang Tao
Engineering Geology: Yusheng Shen, Bo Gao, Xiaoming Yang, Shuangjiang Tao
Engineering Geology
journal homepage: www.elsevier.com/locate/enggeo
a r t i c l e i n f o a b s t r a c t
Article history: After the 2008 Wenchuan earthquake (Mw = 8.3), many mountain highway tunnels in the affected area were
Accepted 20 July 2014 severely damaged. In order to improve the current knowledge about the seismic performance of mountain
Available online 1 August 2014 tunnels under strong earthquakes, a forensic investigation was conducted on 52 tunnels after the earthquake.
Based on the seismic damage data, the damage level of the inspected tunnels was classified using a new damage
Keywords:
evaluation criterion. Subsequently, typical seismic damage characteristics and mechanisms of mountain tunnels
Wenchuan earthquake
Seismic damage mechanism
were analyzed based on three different damage patterns: damage to shallow tunnel, damage to deep-buried
Relative displacement tunnel structure and damage to pavement. Finally, based on the relative displacement (RD) method, a three-
Mountain tunnel dimensional numerical model was developed to investigate the deformation characteristics of the portal struc-
ture under a strong earthquake. The numerical results showed that the tunnel lining cross section exhibited an
alternate tension–compression deformation at the diagonal directions, and the relative displacement value of
the upper-structure (above the sidewall) was generally more than that of the lower-structure (under the side-
wall). Three stress states of the lining were deduced from the numerical model, which provided a good interpre-
tation for the seismic tunnel damages observed after the Wenchuan earthquake.
© 2014 Elsevier B.V. All rights reserved.
http://dx.doi.org/10.1016/j.enggeo.2014.07.017
0013-7952/© 2014 Elsevier B.V. All rights reserved.
86 Y. Shen et al. / Engineering Geology 180 (2014) 85–98
characteristics of the tunnel portal structures under a strong earthquake portals within 20 km from the epicenter were buried or damaged due
action. to slope failure or landslide, but some tunnels being in operation also
were damaged to some extent even though those were more than
2. Field investigation of tunnel damage 150 km from the epicenter, as shown in Table 1.
Fig. 2 presents the typical cross section of a tunnel structure in the
The 2008 Wenchuan earthquake occurred in Wenchuan County, different region. The support systems of tunnels for provincial highways
Sichuan, China. The epicenter was at Yingxiu town with a depth usually consist of primary liner, mesh, piper roof, waterproof layer, and
14 km. The maximum peak ground acceleration (PGA) exceeded 1.0 g. secondary lining (Figure 2a). For local highways, the support systems
The earthquake caused about 320 km rupture along the Longmenshan only consist of primary lining, waterproof layer and secondary lining
Fault Zone by thrusting with a dextral component (Li, 2008). (Figure 2b).
After the earthquake, 52 tunnels in the affected area (Figure 1) were Currently, no standard tunnel damage classification is available,
inspected of which 35 tunnels were in operation and 17 tunnels were although some criteria were introduced in literatures. Dowding and
under construction when the earthquake happened. Most tunnel Rozen (1978) considered three damage classes (i.e., no damage, minor
Qingchuan-Shazhou
road one tunnel
Langmusi-Chuanzhu Jiuzhaigou-Gansu
si road four tunnels road two tunnels
Jianqing road
Beichuan-Qingchuan
Wenchuan-Maerkan one tunnel
road one tunnel
g road five tunnels
Mian-Guang expressway
two tunnels
Yingxiu-Wenchuan
road seven tunnels Wenchuan-Maoxian fault
Yingxiu
Yingxiu-Rilong
road two tunnels
Du-Long road
Dujiangyan-yinxiu one tunnel
road three tunnels
Du-Ying expressway
Four tunnels
Yingxiu
Longmenshan fault
Kangding-Yaan
road seven tunnels
Pubogou road
five tunnels
Hanyuan-Baixionggou
Hanyuan road road three tunnels
four tunnels
No. Tunnel name Length (m) Distance to Tunneling Fault amount Operation/construction Portal Rock Headwall Pavement Lining Lining Cable Water Damage Geology
epicenter (km) method failure falls damage uplift crack dislocation trench leakage level condition
Note: √ = damage; × = no damage; / = no detail; HR = hard rock; SR = soft rock; BTM = benching tunneling method; NATM = New Austrian Tunneling method; FFM = full face method; PEM = partial excavation method.
87
88 Y. Shen et al. / Engineering Geology 180 (2014) 85–98
25
Pipe roof 22
Primary lining (C20, 20cm)
20
Waterproof layer
Tunnel counts
Secondary lining (C25, 50cm) 15
11
10
10
6
5 3
0
ES S M SL N
Damage degree of tunnel
Pavement
Backfill (C10)
Fig. 3. Seismic damage distribution of tunnel.
Table 2
Damage classification and description for tunnels.
3. Mechanism and typical seismic damage analysis of 3.1. Damage to shallow tunnels
tunnel structure
3.1.1. Tunnel portals
Based on the damage level and the mechanism of the tunnel damage For tunnel portal, the typical seismic damage includes inclined crack,
during the earthquake, the damage patterns of tunnel structures were portal wall damage and portal rupture. In some occasions, the tunnel
classified into three types (shallow tunnel, deep-buried tunnel struc- portal was even completely buried, and the retaining wall developed
ture, and pavement) in order to better explain the seismic damage rupture and other structural failures, as shown in Fig. 4.
mechanism of tunnel engineering in Wenchuan earthquake. A similar Generally, most portals are located in the highly weathered strata,
classification method was used by Yashiro et al. (2007). which easily causes secondary disasters during the strong earthquake.
The secondary disasters may be one of the main causes for the portal
damage. For example, at the Longdongzi tunnel, the right tunnel exit
was buried completely, and the left tunnel exit had some rockfalls due
to slope failure during the earthquake (Figure 4a). The Taoguan tunnel
portal developed severe fracture due to the large seismic inertia force
Secondary lining
collapse zone
Buried
portal
rockfall
Portal
rupture
Reinforcing steel
bar distortion
b) Portal rupture of the Taoguan tunnel
b) Reinforcing steel bar distortion
Structure
punctured
(Figure 4b), and the portal of the Gengda tunnel was punctured by the 3.1.2. Portal structure
rockfall during the main shock (Figure 4c). For the existing tunnel lining, typical seismic damage of lining
Portals which suffered extremely severe damage are generally locat- includes various kinds of lining cracks (longitudinal, transverse and cir-
ed in the high intensity area and near the earthquake epicenter (less cumferential cracks), spalling, collapse, rupture, dislocation, water leak-
than 30 km). The regional seismic intensity of some portals reached IX age, reinforcing steel bar distortion, and so on. In the Longxi tunnel, the
degree, which exceeded the original design intensity of VII degree. In secondary lining without reinforcing steel bar collapsed and a large
addition, at the portal, the geological condition is generally a relatively quantity cracking (1 mm ~ 40 mm in width) appeared in the tunnel
loose quaternary formation or highly weathered strata. The ground lining surface, as shown in Fig. 5.
motion may be amplified at the portal section, which can result in the For under-construction tunnels, the typical lining damage includes
larger seismic inertia force and the slope instability. Moreover, second- large deformation, shotcrete falling, wire-mesh exposure, steel arches
ary disasters have crucial influence on the stability and safety of portals distortion, and work face collapse. Because the primary lining with a
under the main shock and aftershock actions. All these factors contrib- flexible support could tightly contact with surrounding rock, it seldom
ute to the portal damage during the earthquake. collapsed during the earthquake. One exception is the Longxi tunnel
where about 100 m long lining collapsed with the steel arches severely
distorted (Figure 6). The main reason may be that the Longxi tunnel is
Steel arches distortion only 0.5 km from the epicenter, and the seismic intensity reached up
and lining spalling to XI degree.
It was observed that portal structures in the hard rock were seldom
damaged during this earthquake, while those in the soft rock were se-
verely damaged. This trend indicates that the seismic inertia force to
the tunnel structure is minor in hard rock compared to that in soft
rock. Portal structures in the bedrock are considered to have a good
earthquake resistance because the bed rock has a good intactness,
high strength and stiffness, and a high load bearing capacity. Moreover,
the amplification factor of PGA for the soft rock is generally higher as
compared to that for hard rock at the portal section.
The instability of surrounding rock is mainly caused by fault disloca-
tion, sand liquefaction, and slope failure. When a tunnel passes through
strata with poor geological conditions, the surrounding rock becomes
a) Steel arches distortion in crown instable during the strong earthquake, especially at the interface be-
tween the cover soil and the bedrock or near the fault fracture zone
(Figure 7a). In this case the portal structure has to carry the additional
seismic load and displacement caused by the surrounding rock instabil-
ity during the earthquake.
Soft
ground
Steel arches
distortion
Earthquake
Hard ground
Displacement caused
by earthquake
In addition, the relative movement or differential formation at the 3.2. Damage to deep-buried tunnel structure
portal section is also considered one of the main causes for the structure
failure. It is crucial to mitigate the relative movement by improving the For deep-buried tunnels, the typical seismic damage includes lining
stiffness mismatch between the structure and the surrounding forma- spalling, collapse, rupture, dislocation and the reinforcing rebar distor-
tion in a strong earthquake area. tion, and cracks (10 cm ~ 60 cm in width). Some tunnels near the epi-
Tunnels with a small overburden at the portal are often located at center developed lining shear rupture near the fault fracture zone,
the interface between soft and hard rocks. There tunnels often experi- especially at the Longxi tunnel and the Baiyunding tunnel, as shown in
enced an amplified ground motion (Miyabayashi et al., 2008). The Fig. 8.
earth pressure generated by amplified ground deformation in a shallow Generally, it was observed that the deep-buried tunnel structure lo-
tunnel is exerted on the tunnel lining as a horizontal force and causes cated in poor geological conditions or faults was vulnerable to seismic
bending tensile stresses at the arch shoulders, as shown in Fig. 7b. damages, especially at fracture zone, active fault, interface between
Under strong earthquake, the additional horizontal seismic force has soft and hard rocks, and so on, as illustrated in Fig. 9. At these locations,
great effect on the bending moment of tunnel structure, and some ten- damages to tunnel structures resulted mainly from the relative motion
sile or compressive cracks occur at the shoulder part of arches, where of surrounding rock during the strong earthquake.
there has larger bending moments. Therefore, the upper-structure of a At the fault fracture zone or transition zones of different rock quali-
tunnel is the most vulnerable to damage due to the squeezing action ties, the two sides of strata may develop relative movement under the
or horizontal seismic force. If the squeezing stress exceeds the shear force action (Figure 9d). The tunnel structure suffers shear failure
maximum load bearing capacity of structure, tunnel lining may crack or dislocation under the strong seismic motion. The cavity existing be-
or develop localized concrete collapses or spalling. hind the lining, the loose zone of surrounding rock, and the poor
d=20cm~30cm
Steel bar d
distortion
d
Rupture d
a) Lining rupture b) Lining dislocation
h=20cm
a) Fault fracture zone
a) Longitudinal dislocation
h=30cm~60cm
b) Change zones between soft and hard rock mass
h
b) Longitudinal uplift
Tensile force d
Tensile force
d
d=20cm~30cm
of the uplifted pavement in left tunnel and the right tunnel were about
geometry of tunnel cross-section are the main causes of the tunnel 1290 m and 1183 m, which accounted for 87% and 81% of the total
structure damage at deep-buried sections. length of the tunnel, respectively. Other tunnel pavements without
the center drainage were seldom damaged, which indicated that the in-
3.3. Damage to tunnel pavement tact bottom structure of the tunnel could mitigate earthquake-related
damage effectively. The transverse shear fracture width of the pave-
The typical seismic damage of tunnel pavements includes pavement ment at the Longdongzi tunnel ranged from 2 cm to 20 cm with a
dislocation, uplift, shear fracture, center drainage ditch damage, and so large number of cracking (3–4 cm in width) and dislocation (2 cm in
on. For example, at the Longxi tunnel, the pavement suffered severe lon- height), as shown in Fig. 10c.
gitudinal and transverse dislocations and uplifts with more than 30 cm, The shear force concentration at the invert bottom and the stiffness
and the maximum height of pavement dislocation and uplift reached difference between different materials are considered the main causes
20 cm and 120 cm, respectively, as shown in Fig. 10a and b. The lengths of the longitudinal pavement dislocations or uplifts (Figure 11a).
Y. Shen et al. / Engineering Geology 180 (2014) 85–98 93
Tunnel portal
Squeezing
action Q4el+dl
Squeezing
Squeezing
action
action
highly weathered mudstone YK207+710R13 Lugu
25.59m
Dislocation
Pavement
Q4ml
Stress Stress
8.50 7.20
concentration concentration
15.80
one
udst 21.80
dm tone
25.59
t here ilts
wea tic s
Shear force kly peli
wea ed
ther ered
wea wea
th
a) Sketch of invert damage wea
kly kly
wea stone
mud
tunnel centre line
gravel backfill
Primary lining (C20, 24cm)
C15/C10
Waterproof layer
R25 C15/C10
60
invert 25
20
C25 concrete
surrounding rock
concrete concrete
C25 concrete
surrounding rock From the above analysis of seismic damage analysis for mountain
tunnel, the portal structure is the most vulnerable to earthquake dam-
c) Local profile without drainage ditch ages as compared to other parts of a tunnel. Therefore, a numerical anal-
ysis was performed to investigate earthquake damage patterns and
Fig. 11. Damage and profile of tunnel invert and pavement (unit: cm). a) Sketch of invert deformation characteristics of tunnel portals for a mountain highway
damage. b) Local profile with drainage ditch. c) Local profile without drainage ditch. tunnel in high intensity earthquake area.
V grade rock
and soil
100
118
120
IV grade rock
mass
b) Cross-section model
Fig. 13. Tunnel computation model. a) Computation model (unit :m). b) Cross-section model.
Analysis of Continua in Three-Dimensional), developed by Itasca Con- model (Figure 13a). The cross section of the tunnel structure model is
sulting Group, Inc. Visco-elastic boundary conditions were applied to shown in Fig. 13b.
the model boundaries in order to reduce the reflection of seismic Based on a previous study by Lysmer and Kuhlemeyer (1969), the el-
waves. The Rayleigh damping was adopted in the analysis since the ement size of the numerical model should not exceed 1/10–1/8 of the
model was elastic, and the critical damping ratio and center frequency smallest wavelength of seismic wave 1/10–1/8. In this study, 1/10 of
were 5% and 5 Hz in the computational model, respectively. More de- the minimum length of shear wave was about 4.97 m, whereas the di-
tails about the Rayleigh damping could be found in Itasca (2009). Con- mension of the numerical elements along the propagation direction of
sidering the boundary effect of the seismic wave propagation, the seismic wave ranged from 0.25 m to 4 m, which was expected to be
dimension of the model was 120 m wide in the horizontal direction, adequate.
118 m long along the tunnel axis, and 100 m in height. The maximum Material properties of the tunnel lining and the surrounding rock in
and minimum buried depths of the tunnel were 42 m and 12 m respec- the numerical model are presented in Table 3. Rock properties in the
tively. A total of 110,046 solid elements were used in the numerical model are based on the report of the geotechnical investigation. In this
Table 3
Physical mechanics parameters of tunnel lining and surrounding rock.
Item Parameter
Density (kg/m3) Elastic modulus (GPa) Poisson's ratio Cohesion (kPa) Internal friction angle (°)
500
400
300
Acceleration (cm/s2)
200
100
0
-100
-200
-300
-400
-500
0 3 6 9 12 15 a) Distribution of monitoring point and measuring line
Time (s)
a) Time-history curve of accelerations
400 surrounding
rock
350 monitoring
300
section
Amplitude (cm/s)
250
A
200
tunnel
A
150
bedrock
100
b) Monitoring section position for portal structure
50
Fig. 15. Distribution of research objective. a) Distribution of monitoring point and measur-
0 ing line. b) Monitoring section position for portal structure.
0 4 8 12 16 20
Frequency (Hz)
0 4 8 12 16
b) Fourier spectrum of accelerations Time,(sec.)
20
Relative displacement,(mm)
AB BD CE AC
Fig. 14. History curve and Fourier spectrum of seismic wave accelerations. a) Time-history
curve of accelerations. b) Fourier spectrum of accelerations. 10 P
O BD
0 M
paper, based on the response spectrum of the tunnel site conditions, the
seismic wave adopted is a synthesis artificial wave from Earthquake
AC
Administration of Sichuan Province, China, as shown in Fig. 14a. The -10
corresponding acceleration Fourier spectrum of the seismic wave is AB Q
shown in Fig. 14b, the principal frequency is around 2.93 Hz, the maxi- 3.24s 6.38s 7.88s 12.5s
mum amplitude of seismic acceleration is 0.42 g, and the duration time -20
is 15 s. The seismic input wave is applied on the bottom boundary of the a) Four measuring lines of upper-structure
computational model, and the direction of incident shear wave is
upwards. 0 4 8 12 16
Time,(sec.)
12
Relative displacement,(mm)
DF NF EG NG
4.2. Deformation analysis method 8
DF
4 NF
Generally speaking, it is difficult to reflect the tensile and compres- K
sive deformation state of the lining during the earthquake using the ab- 0
solute displacement of tunnel structure due to the small relative L
deformation between structure and the surrounding rock. Therefore, -4
this paper adopts a maximum allowable relative displacement (RD) NG
-8 EG
method to evaluate the dynamic deformation characteristics of tunnel
3.24s 6.38s 7.88s 12.5s
structure, namely, the difference of absolute displacement value of the -12
adjacent monitoring points on the tunnel lining. Eight monitoring b) Four measuring lines of lower-structure
points are selected along the lining to evaluate the deformation charac-
teristic of structure under the seismic load. Monitoring points were con- Fig. 16. Relative displacement time-history curve of various measuring lines. a) Four mea-
nected to each other by measuring lines as shown in Fig. 15a. A positive suring lines of upper-structure. b) Four measuring lines of lower-structure.
96 Y. Shen et al. / Engineering Geology 180 (2014) 85–98
Relative displacement,(mm)
CF BG
Since the tunnel portal is the most vulnerable to earthquake dam- 10 Line BG
ages, only the deformation characteristic of a near-portal cross section
O
(as shown in Figure 15b) was analyzed in the numerical analysis. 0
-10
4.3. Relative displacement analysis of portal structure
H
12.5s
4.3.1. Relative displacement characteristic analysis -20
3.24s 6.38s 7.88s
Based on Fig. 16a, The RD curves of measuring lines at the same Line CF
height are symmetrically distributed along a certain line with an angle -30
against horizontal line, such as line OMP, line OMQ, and line KL
Fig. 17. RD time-history curve of line CF and BG.
(Figure 16a). It is shown that the RD values of measuring lines AB and
AC near the crown are always negative after the first peak value,
which indicates that the crown was mainly under compression during compression, and the lining collapse or spalling at the crown was prob-
the seismic motion. Similarly, the RD values of measuring lines BD and ably caused by the larger RD of the upper-structure during the strong
CE near the springline are always positive after the second peak value, earthquake.
which indicates that the structure near the springline was mainly
governed by tension. Fig. 16 also shows that the maximum RD values
of the upper-structure (above the sidewall) are more than that of the 4.3.2. Deformation mechanism of tunnel structure at the diagonal
lower-structure (under the sidewall) of the tunnel. The maximum RD directions
values at the upper-structure are about + 15.3 mm at line CE and Fig. 17 presents the RD time-history of diagonal measuring lines
− 15.10 mm at line AB (at 12.5 s), respectively, while those values at from the numerical model. It is shown that the peak and trough values
lower-structure are about + 7.8 mm at line EG and − 8.6 mm at line of RD curves at line BG and CF are symmetrically distributed along the
DF (at 3.24 s), respectively. Theoretically speaking, the RD value of the line OH. The RD peak value of the measuring line BG reaches
lining should return to zero after the earthquake. However, the RD − 20.2 mm, + 3.86 mm, − 12.2 mm and + 2.0 mm at 3.24 s, 6.38 s,
values of measuring lines AB and AC reach −12.9 mm and −12.2 mm 7.88 s and 12.5 s, respectively. This indicates that the tunnel lining at
at 15 s, which indicates that the tunnel structure developed residual BG diagonal direction experienced a cyclic stress state of compres-
strain or permanent deformation during the seismic motion. A Similar sion–tension–compression–tension during the earthquake. However,
phenomenon was observed in an experimental study by Lanzano et al. tunnel lining at CF diagonal direction is mainly in compression after 4
(2012). s and its maximum RD peak value reaches − 28.6 mm (compressive
The RD peak values of different measuring lines at different time pe- state) at 12.5 s.
riods are listed in Table 4. The signs of the RD values demonstrate that Based on Fig. 17 and Table 4, the deformation state of the lining cross
the lower left sidewall (line DF and NF) experienced a cyclic stress of section is sketched at 3.24 s and 12.5 s seismic wave, respectively, as
compression–tension–compression–tension, whereas the lower right shown in Fig. 18. From Fig. 18a, the RDs of the four measuring lines
sidewall (line EG) experienced an opposite cyclic stress. So the lower (AB, DB, EG, and NG) near the diagonal line BG direction are all negative
portions of both sidewalls were controlled alternately by the tensile values, On the contrary, the RDs of measuring lines AC, CE, DF, and FN
and compressive stress during the earthquake. Before the earthquake, are all positive. So the tunnel structure at CF and BG diagonal direction
both sidewall bottoms of tunnel structure were prone to the stress con- is perhaps controlled by the tension and compression at the peak time
centration issue due to those abrupt changes for cross-section shape. 3.24 s, respectively. Based on Fig. 18b, the RD of the tunnel structure
When the earthquake happened, the additional tensile and compressive at time 12.5 s is opposite to that of peak 12.5 s. In addition, the RD values
stress could be applied to both sidewall bottoms. This trend could of line AC and CE are − 6.7 mm (compressive state) and + 15.4 mm
possibly explain why the bottom of mountain tunnel often suffered (tensile state), respectively. If the monitoring point A and E are assumed
the severe damage or collapse in the Wenchuan earthquake. Measuring to be still, the point C moves about 22.1 mm towards point A. So the in-
lines NF and NG also exhibited similar deformations during the earth- side of the lining is the most vulnerable to tensile fracture component
quake, which indicated the cause of the shear fracture or dislocation at under the seismic squeezing effect.
the tunnel invert in the Wenchuan earthquake. Considering the deformation characteristic of the structure
In addition, the maximum and minimum horizontal RD reached (Figure 18) and RD data of all measuring lines (Table 4), the stress
−25.1 mm at line BC and −6.16 mm and line FG at 15 s, respectively. state of the lining being deduced was categorized into three types dur-
This indicates that the whole tunnel structure was mainly under ing the seismic motion of the 15-s period, as shown in Fig. 19. It should
Table 4
RD peak values of measuring lines for portal structure (unit: mm).
Time Item
AB BD DF NF NG EG CE AC BC DE FG
3.24 s +2.80 +10.90 −8.40 −3.70 +2.80 +7.60 −7.90 −7.50 −4.70 −1.70 −0.90
6.38 s −6.78 −2.31 +5.17 +0.95 −4.16 −5.15 +7.33 −1.43 −8.21 −3.19 −3.21
7.88 s −4.55 +7.01 −1.56 −2.12 −1.52 +1.23 +0.85 −7.28 −11.83 −3.97 −3.64
12.50 s −15.00 −0.20 +7.10 +1.10 −6.00 −8.30 +15.40 −6.70 −21.70 −6.10 −4.90
15.0 s −12.87 +7.69 +1.36 −2.09 −4.07 −2.86 +9.69 −12.18 −25.05 −7.66 −6.16
Y. Shen et al. / Engineering Geology 180 (2014) 85–98 97
Fig. 18. Deformation characteristic map of portal lining at peak time. a) RD sketch at 3.24 s. B C
b) RD sketch at 12.50 s.
tunnel during
D wave motion E
be noted that the mechanism of tunnel ovalization has also been docu-
mented by previous researchers. Owen and Scholl (1981) and Hashash G
F
et al. (2001) both discussed the mechanism of tunnel ovalization with N
alternative tensile-compressive deformation based on a circular tunnel (compressive stress)
section. The tunnel section modeled in this study is located at the portal
(shallow tunnel) with an oval cross-section which consists of four arcs. c) At 7.88s
Therefore the numerical analysis is complement of the existing tunnel
deformation theories. Fig. 19. Schematic drawing of stress behavior state for tunnel structure during the earth-
Under the influence of tensile and compressive stresses, the upper- quake. a) At 3.24 s. b) At 6.38 s and 12.50 s. c) At 7.88 s.
ditch is generally caused by the large shear force of invert and the stiff- He, W.,Wu, Z.S.,Kojima, Y., 2009. Failure mechanism of deformed concrete tunnels subject
to diagonally concentrated loads. Comput. Aided Civ. Infrastruct. Eng. 24, 416–431.
ness mismatch among various structures. Huang, T.H., Ho, T.Y., Chang, C.T., Yao, X.L., Chang, Q.D., Lee, H.C., 1999. Quick investigation
• Using the relative displacement (RD) method, the RD values among and assessment on tunnel structure after earthquake, and the relevant reinforced
the selected monitoring points show a regular alternation of tensile- methods. Report for Public Construction Commission, Taipei, Taiwan.
Itasca, 2009. Software Manual of FLAC3D. Itasca Consulting Group, Inc., Minneapolis, MN.
compressive deformations at the tunnel diagonal directions during Konagai, K.,Johansson, J.,Zafeirakos, A.,Numada, M.,Sadr, A.A., 2005. Damage to tunnels in
the earthquake. The RD values of the upper-structure of a tunnel the October 23, 2004 Chuetsu Earthquake. JSCE J. Earthq. Eng. 1–8.
were generally larger than that of the lower-structure. The final RD Lanzano, G.,Bilotta, E.,Russo, G., 2008. Tunnels under seismic loading: a review of damage
case histories and protection method. Presented at: In Workshop of Mitigation of the
value among monitoring points does not return to zero under the ad-
Earthquake Effects in Towns and in Industrial Regional Districts. Italy.
ditional strain and permanent deformation caused by the strong Lanzano, G.,Bilotta, E.,Russo, G.,Silvestri, F.,Madabhushi, G.S.P., 2012. Centrifuge modeling
earthquake. of seismic loading on tunnels in sand. Geotech. Test. J. 35 (6), 1–16.
Li, T.B., 2008. Failure characteristics and influence factor analysis of mountain tunnels at
• Three stress states of the lining were deduced from the deformation
epicenter zones of great Wenchuan earthquake. J. Eng. Geol. 16 (6), 742–750.
characteristics of tunnel structure during the seismic motion action, Lin, M.L.,Chung, C.F.,Jeng, F.S.,Yao, T.C., 2007. The deformation of overburden soil induced
which can provide a good interpretation for the seismic damage ob- by thrust faulting and its impact on underground tunnels. Eng. Geol. 92, 110–132.
served in the Wenchuan earthquake. Lysmer, J., Kuhlemeyer, R.L., 1969. Finite dynamic model for infinite media. J. Eng. Mech.
Div. Proc. ASCE 95 (EM4), 859–877.
Miyabayashi, H.,Tosaka, T.,Isogai, A., 2008. Basic studies on earthquake damage to shallow
Acknowledgements mountain tunnels. World Tunnel Congress 2008 — Underground Facilities for Better
Environment and Safety — India, pp. 406–417.
Owen, G.N., Scholl, R.E., 1981. Earthquake engineering of large underground structures.
The authors appreciate the support from the National Natural Report No. FHWA/RD-80/195Federal highway administration and national science
Science Foundation of China (No. 41102220 and No. 51178398), the foundation.
Fundamental Research Funds for the Central Universities (No. Qian, Q.H., He, C., Yan, Q.X., 2009. Analysis on seismic damages and dynamic response
characteristic of tunnel engineering subjected to Wenchuan earthquake. Study on
SWJTUCX014) and the Development Plan for Ministry of Education In- Engineering Damage in Wenchuan Earthquake (Special Issue)pp. 608–618.
novation team (No. IRT0955). Shen, Y.S.,Gao, B.,Wang, Y.X., 2009. Analyses on tunnel failure mode and mechanism dur-
ing Wenchuan earthquake. International Conference on Earthquake Engineering —
The 1st Anniversary of Wenchuan Earthquake, pp. 568–573.
References Shen, Y.S., Gao, B., Hu, B., Wen, Y.M., 2011. Study on the interval grouting pre-
reinforcement scheme of mountain tunnel surrounding rock in high-intensity earth-
Alija, S.,Torrijo, F.J., Ferreira, M.Q., 2013. Geological engineering problems associated with
quake zone. China Civ. Eng. J. 44 (Supp. l), 186–192.
tunnel construction in karst rock masses: the case of Gavarres tunnel (Spain). Eng.
Sun, Y.W., Li, P., Bo, J.S., Liu, B., Qi, W.H., 2010. Damage investigation and cause analysis of
Geol. 157, 103–111.
Longxi tunnel in Wenchuan great earthquake. World Earthq. Eng. l26 (Supp. l),
Asakura, T., Tsukanda, K., Matsunaga, T., Matsuoka, S., Yashiro, K., Shiba, Y., Oya, T., 2000.
187–192.
Damage to mountain tunnels by earthquake and its mechanism. Proc. JSCE (Jpn
Wang, Z.Z., Zhang, Z., 2013. Seismic damage classification and risk assessment of moun-
Soc. Civ. Eng.) 659, 27–38.
tain tunnels with a validation for the 2008 Wenchuan earthquake. Soil Dyn. Earthq.
Communication Surveying and Design Institute of Sichuan Provincial Communication
Eng. 45, 45–55.
Transport Department CSDISPCTD, 2008. Investigation, Detection, and Evaluation of
Wang, W.L.,Wang, T.T.,Su, J.J.,Lin, C.H.,Seng, C.R.,Huang, T.H., 2001. Assessment of damage
the Restoration and Reconstruction Project of Highways, National and Provincial
in mountain tunnels due to the Taiwan Chi-Chi earthquake. Tunn. Undergr. Space
Trunk Roads in Disaster Areas Induced by Wenchuan earthquake. Communication
Technol. 16, 133–150.
Surveying and Design Institute of Sichuan Provincial Communication Transport
Wang, Z.Z., Zhang, Z., Gao, B., Shen, Y.S., 2012. Factors of seismic damage and fuzzy syn-
Department, Chengdu, China.
thetic evaluation on seismic risk of mountain tunnel portals. J. Cent. South Univ.
Cui, G.Y., Wang, M.N., Lin, G.J., Zhang, W.Q., Wang, W.J., 2011. Study of the earthquake
(Sci. Technol.) 43 (3), 1122–1130.
damage mechanism and aseismatic countermeasure of a highway tunnel portal
Yashiro, K.,Kojima, Y.,Shimizu, M., 2007. Historical earthquake damage to tunnels in Japan
section in the Wenchuan seismic disaster area. Mod. Tunn. Technol. 06, 6–10.
and case studies of railway tunnels in the 2004 Niigataken-Chuetsu earthquake. Q.
Dowding, C.H., Rozen, A., 1978. Damage to rock tunnels from earthquake shaking. Am.
Rep. RTRI 48 (3), 136–141.
Soc. Civ. Eng. J. Geotech. Eng. Div. 104, 175–191.
Gao, B., Wang, Z.Z., Yuan, S., Shen, Y.S., 2009. Lessons learnt from damage of highway
tunnels in Wenchuan earthquake. J. Southwest Jiaotong Univ. 44 (3), 336–343.
Hashash, Y.M.A.,Hook, J.J.,Schmidt, B.,Yao, J.I., 2001. Seismic design and analysis of under-
ground structures. Tunn. Undergr. Space Technol. 16 (4), 247–293.