Bull Eng Geol Environ (2019) 78:431–444

DOI 10.1007/s10064-017-1155-8

ORIGINAL PAPER

Mechanisms of large deformation in soft rock tunnels: a case

study of Huangjiazhai Tunnel
Kang Bian 1 & Jian Liu 1 & Zhenping Liu 1 & Shangge Liu 1 & Fei Ai 1 & Xiaoqing Zheng 1 &
Shaohu Ni 2 & Wei Zhang 3

Received: 12 December 2016 / Accepted: 28 August 2017 / Published online: 13 September 2017
# Springer-Verlag GmbH Germany 2017

Abstract Typical large-deformation phenomena of soft rock deformations in the Huangjiazhai Tunnel: the plastic flow
occurred frequently during the tunneling process in caused by tunnel excavation under high geostress and low
Huangjiazhai Tunnel, Hubei Province, China, including extru- rock strength and a hydrated-mechanical coupling process be-
sion at the tunnel wall and severe damage of the primary tween the shales and water.
support and secondary lining. To reveal the mechanisms of
these anomalies, comprehensive investigations have been un- Keywords Large deformation . Soft rock tunnels . Immersion
dertaken combining engineering, laboratory, and microscopic time . Microstructure . Hydrated-mechanical coupling
analyses. Since the monitoring results show that there might
be a close relationship between the large deformation phe-
nomena and water infiltration into the tunnel, the water–rock Introduction
interaction is the research focus in the investigations. The
experimental results reveal that the deforming resistance and Highway transportation in mountainous areas is of great im-
strength of shales exposed at the excavation face weaken portance for local economic development and Tunneling en-
quickly in the first 20 days after the shales have contact with gineering plays a significant role in highway construction in
water. The results obtained by mineral composition detection these mountainous regions.
and microstructure observation illustrate the microscopic rea- Large deformations of surrounding rock are often encoun-
sons for the decreasing strength and deforming resistance of tered during tunnel excavation in rock masses with squeezing
the tunnel surrounding rock after water infiltration. The results characteristics. Generally, the deformation usually lasts from
from in situ geostress tests indicate that as a result of high several weeks to several months (Aydan et al. 1996; Brox et al.
tectonic stress and low rock strength, the field of 1999; Hoek 2001; Meng et al. 2013) or even up to more than
Huangjiazhai Tunnel is an extremely high geostress area. By 1 year in soft rock tunnels (Yassaghia and Salari-Rad 2005;
combining analyses from the mechanical and geological per- Wang et al. 2015), often accompanied by the lining cracking
spectives, there are two main reasons for the large and even collapsing. These phenomena are widely observed at
relatively shallow depths in soft rocks such as phyllite, mud-
stone, shale, and siltstone (Aydan et al. 1996). When a tunnel
* Kang Bian is excavated in a soft squeezing rock, the ground advances
biankang2002@163.com slowly into the opening, without visible fracturing or loss of
continuity. The continuous inward deformation of the rock
1
State Key Laboratory of Geomechanics and Geotechnical
may occur together with the separation of rock fragments, or
Engineering, Institute of Rock and Soil Mechanics, Chinese blocks, from the roof and walls of the excavation (Gioda and
Academy of Science, Wuhan, Hubei 430072, China Cividini 1996). Because of the universality and pernicious-
2
Power China Huadong Engineering Corporation Limited, ness of large deformations in the construction of soft rock
Zhejiang, Hangzhou 310014, China tunnels, the squeezing behavior and its mechanisms during
3
College of Water Conservancy and Civil Engineering, South China tunneling in soft rock layers have aroused much interest
Agricultural University, Guangzhou 510642, China among a number of researchers in the past (Wood 1972;
432 K. Bian et al.

Nakano 1979; Tan et al. 1980; Aydan et al. 1995, 1996; Barla 2007). Adopting higher-capacity segmental lining systems
et al. 1995; Kalamaris and Bienawski 1995; Gioda et al. 1996; will extend the operational possibilities for Tunnel Boring
Kovfiri and Staus 1996; Steiner 1996). Wood (1972) proposed Machine (TBM) in squeezing rock (Mezger et al. 2017).
a competency factor, which was defined as the ratio of uniax- Water-weakening effects are widely believed to be an im-
ial strength of rock to overburden stress to assess the stability portant factor contributing to large deformations in soft rock
of tunnels. Nakano (1979) later used this parameter to recog- tunnels, and most of the past relevant research (Van Eeckhout
nize the squeezing potential of soft-rock tunneling in Japan. 1976; Alm 1982; Hadizadeh and Law 1991; Erguler et al.
Aydan et al. (1995) and Barla (1995) stated that squeezing was 2009; Yilmaz 2010; Duda and Renner 2012; Jiang et al.
a large, time-dependent deformation that occurred around 2014) has concentrated on the mesoscopic mechanisms and
tunnels and was essentially associated with creep caused by mechanical properties of soft rock influenced by water.
exceeding a shear stress limit. On the basis of surveying and However, little research has investigated the relationship be-
analyzing the similarity of more than 20 squeezing tunnels, tween the mesoscopic mechanisms of soft rock and the mac-
Aydan et al. (1993, 1996) proposed that squeezing-type fail- roscopic deformation behavior of a tunnel with water infiltra-
ures tended to occur in soft rock tunnels when several condi- tion. In this paper, we study the interaction between the hy-
tions were satisfied, such as a competency factor less than 2 dration effect of shales and the deformation behavior of tunnel
and a tangential strain of the tunnel wall greater than 1%. With rock. The coupling process continuously cycles and finally
an increasing number of soft rock tunnels being constructed, induces the large deformation and even failure of the tunnel.
researchers studied large-deformation mechanisms of practi- In this paper, to reveal the mechanism of large-deformation
cal tunnel projects (Khanlari et al. 2012; Meng et al. 2013; phenomena during tunneling in Huangjiazhai Tunnel, com-
Agan 2016). Among them, Meng et al. (2013) summarized prehensive investigations have been undertaken by combining
three factors inducing large deformations in Zhegu Tunnel, engineering, laboratory, and microscopic analyses. The large
situated on the Sichuan-Tibet Highway: the plastic flow of deformation in Huangjiazhai Tunnel has two main reasons:
soft rock, shear sliding of wedges, and bending of thin- the plastic flow caused by tunnel excavation under high
layered soft rock. geostress and low rock strength and the hydrated-mechanical
Over the past several decades, a number of researchers coupling process between shales and water.
have studied the large deformations in squeezing tunnels, es-
pecially the extent of the plasticity zone (Anagnostou 1993;
Duncan 1993; Yassaghia et al. 2005; Sterpi et al. 2009; Meng Topographic and geological conditions
et al. 2013; Wang et al. 2015; Agan 2016). Many studies
(Brox et al. 1999; Dalgic et al. 2002; Bizjak and Petkovsek Location and topographic conditions
2004; Khanlari et al. 2012; Wang et al. 2015) have focused on
the large-deformation mechanisms from the perspective of Huangjiazhai Tunnel is one of the 29 tunnels of the Macheng-
either mechanics or geology. To reveal the mechanisms of Zhuxi Highway, which is located in the north-west of Hubei
large deformations in a squeezing tunnel, integrated analyses Province in Central China. The owner of Huangjiazhai Tunnel
of the mechanical and geological methods are usually essen- is Hubei Provincial Communications Investment Group Co.,
tial. However, investigations of the causes triggering squeez- Ltd., a wholly state-owned company funded by the
ing in soft rock tunnels are normally not comprehensive in the Government of Hubei Province. The tunnel was constructed
existing research; few studies have researched this problem by between 29 May 2013, and 25 December 2015, and the con-
combining engineering, laboratory, and microscopic analyses. struction period lasted for nearly 2.5 years. Designed as a
Because of the various geological conditions in different tun- long, separated tunnel, Huangjiazhai Tunnel has a maximum
nels, each tunnel has distinct reasons underlying the observed buried depth of 270 m and a length of 1437 m. Both tunnels
squeezing problems. To intrinsically reveal the causes of large are of 12.4 m width and 9.9 m height, and an end-wall tunnel
deformations for a specific tunnel, comprehensive investiga- portal was chosen. The tunnel sites are mainly hilly areas, and
tions must be carried out. It has been demonstrated that there are also some valleys and terraces.
squeezing can be controlled effectively by installing yielding
supports (Hoek and Guevara 2009; Cantieni and Anagnostou Lithology
2009; Barla et al. 2011; Aksoy et al. 2012), and the use of
simple sliding joints in steel sets has proven to be a very The principal stratum at the construction site of the
effective means for providing emergency support in the event Huangjiazhai Tunnel is Silurian shale belonging to the Sllm
of a tunnel collapse close to the face (Hoek et al. 2009). If the Group (S1l). Quaternary residual gravelly soil (Q4el+dl) consti-
permissible deformations are known beforehand, measures tutes the covering layer of this area. Geological explorations
such as the design of steel fiber reinforced shotcrete (SFRS) via boreholes drilled at the tunnel route show that three layers
may be used to solve the squeezing problem (Singh et al. divide the ground surface to the bottom: gravelly soil, heavily
Mechanisms of large deformation in soft rock tunnels: a case study of Huangjiazhai Tunnel 433

weathered shale and moderately weathered shale. The topo- Quaternary, is supplied by the infiltration of atmospheric pre-
graphic map and geological profile of Huangjiazhai Tunnel cipitation. The annual average rainfall is 800–1200 mm at the
are shown in Figs. 1 and 2. The estimated rock mass classifi- tunnel site, mostly in summer, when the south-east monsoon
cation systems of Huangjiazhai Tunnel are shown in Table 1. carries rainwater into the inland provinces.
The gravelly soil, which is made up of shale fragments and
small amounts of clay, displays a yellow to brown color. The
diameter of these shale pieces ranges from 2 to 8 cm, and the
soil thickness at the entrance to the tunnel varies between 1.7 Large-deformation phenomena during tunneling
and 14.8 m. The heavily weathered shale has a pelitic texture, and monitoring situation
with thin- and medium-layered structures and well-developed
joints. The layer thickness of heavily weathered shale usually Large-deformation phenomena during tunneling
ranges from 10.2 to 26.4 m, and the cores are mostly fragmen-
tized. The moderately weathered shale is medium-jointed, The New Austrian Tunneling Method (NATM) was applied in
with a depth generally from 10 to 20 m. The cores drilled from the construction of Huangjiazhai Tunnel. The bench-cut meth-
the boreholes of the three layers are shown in Fig. 3. od was utilized at Chainage K60 + 860–K60 + 980, K61 +
140–K61 + 240 and K61 + 320–K61 + 440 in the right tunnel,
and it was applied at Chainage ZK60 + 830–ZK60 + 945,
Geological structure
ZK61 + 166–ZK61 + 196 and ZK61 + 310–ZK61 + 603 in
the left tunnel, with a total length of 778 m. In the residual
According to data on the regional structure of Hubei Province,
sections where the rock masses were considered soft and frag-
Huangjiazhai Tunnel is located at the intersection of the third
mentized, the ring cut method (Hisatake et al. 2009; Hisatake
uplifted zone of a Neocathaysian system and the west reflex-
et al. 2012) was adopted. The initial supports in the tunnel
arc of the Huaiyang epsilon-type structure. A world-class
were combinative technologies, including ductile grouting,
fault, Qingfeng Fault, is located in the north of Huangjiazhai
steel arch, reinforcing mesh, and shotcrete. Reinforced con-
Tunnel. East-west folded belts and fractured zones of the west
crete was adopted as the secondary lining. The details of the
reflex-arc of the Huaiyang epsilon-type structure run through
construction procedure and supports for the ring cut method
the tunnel sites. A north–north-east (NNE) tectonic belt and a
are shown in Fig. 4.
north–north-west (NNW) fractured zone, both of which be-
Continuously exposed at the excavation surface during the
long to the Neocathaysian system, overlay the epsilon-type
tunneling process, most of the rock masses were fragmentized,
structure.
thin-layered, and jointed (Fig. 5). The groundwater developed
at several sections in Huangjiazhai Tunnel (e.g., Chainage
Hydrogeology ZK61 + 977–ZK62 + 010, ZK61 + 603–ZK61 + 624,
ZK61 + 203–ZK61 + 196, ZK61 + 166–ZK61 + 156,
The groundwater at the tunnel site, which is composed of ZK61 + 122–ZK61 + 112, ZK61 + 085–ZK61 + 052,
fissure water at bedrock and pore water at the loose bed of ZK61 + 052–ZK61 + 036, ZK60 + 690.5–ZK60 + 707,

Fig. 1 Topographic map of Huangjiazhai Tunnel

434 K. Bian et al.

Fig. 2 Geological profile of Huangjiazhai Tunnel

ZK60 + 945–ZK60 + 980.6, and ZK60 + 726.5–ZK60 + surface was halted. The steel arches between Chainage K60 +
747.1 in the left tunnel, and Chainage K60 + 803–K60 + 733 and K60 + 734.2 had to be replaced at that time; the
784, K61 + 132–K61 + 112, K61 + 072–K61 + 055, K60 + primary supports of these sections were chiseled off by the
720–K60 + 745, and K60 + 745–K60 + 763 in the right workers. First, some of the rock fragments were seen to slip
tunnel). Among these tunnel sections, the groundwater at from the left sidewall and crown; as time passed, the number
Chainage K61 + 132–K61 + 112 and K61 + 072–K61 + 055 of rock fragments increased rapidly, and collapse at the exca-
was the most abundant, as shown in Fig. 6. Collapse occurred vation surface was induced afterwards. Typical large-
occasionally in Huangjiazhai Tunnel. For instance, on 1 deformation phenomena of soft rock occurred frequently, such
September 2015, a collapse occurred at the excavation surface as extrusion of the tunnel wall (Chainage K60 + 745–K60 +
at Chainage K60 + 762.8 in the right tunnel, as shown in Fig. 763, Fig. 7b), damage of the primary supports (Chainage
7a. When the collapse occurred, construction at the excavation ZK60 + 980.6–ZK60 + 945, Fig. 7c), and surface peeling of
the secondary lining (Chainage K61 + 026–K61 + 014,
Fig. 7d). According to the statistics, a special alteration of rock
Table 1 The estimated rock mass classification systems in
Huangjiazhai Tunnel mass grade occurred 27 times in constructing Huangjiazhai
Tunnel, resulting in a dramatic increase in the project costs
Chainage Length /m RMR and a delay in the tunnel construction.
K60 + 666–K60 + 750 84.5 16
ZK60 + 636–ZK60 + 720 84
Monitoring situation
K60 + 750–K60 + 860 110 31
ZK60 + 720–ZK60 + 830 110
K60 + 860–K60 + 980 120 44
The monitoring results provide quantitative data to analyze the
ZK60 + 830–ZK60 + 980 150 real situation during the tunneling process. The horizontal
K60 + 980–K61 + 060 80 32 displacement of Huangjiazhai Tunnel was measured with tape
ZK60 + 980–ZK61 + 010 30 extensometers, which measure the distance between reference
K61 + 060–K61 + 140 80 17 points (hooks) bolted on the tunnel wall. The accuracy of such
ZK61 + 010–ZK61 + 140 130 measurements is typically ±0.2 mm for distances up to 10–
K61 + 140–K61 + 170 30 35 15 m (Kavvadas 2005; Walton et al. 2014).
ZK61 + 140–ZK61 + 170 30
The vertical displacement of Huangjiazhai Tunnel was
K61 + 170–K61 + 240 70 32
measured with total station. Total station surveying uses opti-
ZK61 + 170–ZK61 + 240 70
cal reflectors that are installed around the tunnel profile (usu-
K61 + 240–K61 + 320 80 27
ZK61 + 240–ZK61 + 310 70 ally at five to seven points) and a stable reference point that is
K61 + 320–K61 + 440 120 43 outside the zone of convergence to measure deformations. The
ZK61 + 310–ZK61 + 430 120 total station must be moved progressively forward from the
K61 + 440–K61 + 700 260 34 location of the stable reference points to the tunnel excavation
ZK61 + 430–ZK61 + 720 290 surface as the tunneling progresses. Total stations have an
K61 + 700–K61 + 890 190 31 accuracy of about ±2.5 mm over 100 m for tunneling applica-
ZK61 + 720–ZK61 + 900 180 tions (Kavvadas 2005; Walton et al. 2014).
K61 + 890–K62 + 020 130 33 Figure 8 shows the monitoring results of displacement at
ZK61 + 900–ZK62 + 000 100
two typical sections of Huangjiazhai Tunnel. During the mon-
K62 + 020–K62 + 101 81 15
ZK62 + 000–ZK62 + 084 84
itoring period of 27 days, the maximum accumulative horizon-
tal displacement exceeded 500 mm, and the maximum velocity
RMR rock mass rating of horizontal convergence reached 60.6 mm/day. Compared
Mechanisms of large deformation in soft rock tunnels: a case study of Huangjiazhai Tunnel 435

Fig. 3 Cores drilled from boreholes of three strata at Huangjiazhai Tunnel: (a) gravelly soil; (b) heavily weathered shale; and (c) moderately weathered shale

with the horizontal convergence, the vertical displacement was month; thereafter, the velocity of pressure increment gradually
slightly smaller; however, the maximum accumulative vertical decreased and became stable. The pressure at the crown
displacement reached 304 mm. After monitoring for 5 or remained unchanged from early September 2014, and the value
6 days, the deformation velocity decreased compared with that at the tunnel wall maintained a continuously increasing trend
in the earlier days; nevertheless, the velocity was still great. until the middle of October 2014. In general, the pressure was
Throughout the monitoring period, both the horizontal and ver- higher at the tunnel wall than at the crown, and the maximum
tical deformations at the monitored sections showed no possi- contact pressure occurred at the right tunnel wall, which was in
bility of convergence. The deformation in Huangjiazhai Tunnel excess of 0.68 MPa.
had the properties of large magnitude, high velocity, and long According to former research (Zhu et al. 2007; Yang
duration. Since the ground displacements that occurred imme- et al. 2014), shale is considered a type of rock with typical
diately after tunnel excavation could not be monitored, the laser rheological properties. The experimental results from the
measuring instrument was also utilized in this tunnel section. direct-shear creep tests showed that the shear modulus,
The results showed that the actual horizontal deformation of the shear strain, and shear stress all have non-linear relation-
sections exceeded 800 mm. In this case, the originally designed ships with time, and a rheological model has been devel-
contour lines of the tunnel had been encroached. The primary oped to describe the rheological characteristics of shales
supports at these large-deformation parts, including the (Yang et al. 2014). Other experimental research showed
distorted steel arches and cracked shotcretes, together with that the axial and lateral creep curves of shale specimens
the squeezing rock masses, were chiseled off meter by meter. exhibit the attenuation creep stage under a relatively low
Immediately after chiseling, new steel arches were installed and stress state. The rheological constitutive equation was ob-
the new shotcrete-bolting-mesh supports were applied. The tained via the parametric regression of an Hooke-Kelvin
above cycles were repeated until all of the large deformations (H-K) three-component model (Zhu et al. 2007). The
were repaired. above observed deformation characteristics and contact
Figure 9 shows the monitoring results of contact pressure pressure seem to be explained by the typical rheological
between the primary support and secondary lining at Chainage property of the shale in Huangjiazhai Tunnel. However,
ZK61 + 746 in Huangjiazhai Tunnel. During the observation considering the buried depth of the large-deformation re-
period, the contact pressure increased quickly in the initial gions (generally approximately 150–200 m), it is still

Fig. 4 Details of tunnel construction procedure and supports for the ring cut method
436 K. Bian et al.

are important factors inducing large-deformation phenomena

in tunnels. Previous studies (Tan and Kang 1980; Anagnostou
1993; Erguler and Ulusay 2009; Meng et al. 2013) have indi-
cated that rock swelling due to clay material encountering
groundwater is one of the most important factors in large
deformation during tunneling. Therefore, the mineral compo-
sition of rock masses and groundwater are also potential fac-
tors that trigger squeezing in the tunnel. Since the monitoring
results show that there might be a close relationship between
the large deformation and water infiltration, the water–rock
interaction was the research focus of the investigations.
Fig. 5 Rock masses at the excavation surface at Chainage K61 + 931 in Therefore, in addition to a geostress test, a series of laboratory
Huangjiazhai Tunnel experiments was carried out, including uniaxial compression
tests, microstructure observations of shale under different im-
mersion times, and mineral composition detection, to reveal
difficult to understand such large deformations and high
the mechanisms of such anomalies during the construction of
contact pressures. It was found that the tunnel deformed
Huangjiazhai Tunnel.
intensively during continuous rainy days; on the days with-
out rain, the occurrence of severe deformations decreased
Uniaxial compression tests for different immersion times
dramatically. Consequently, it appeared that there might be
close relationships between the large deformation and wa-
Rock samples obtained from the intact shale inside
ter infiltration into the tunnel.
H u a n g j i a z h a i Tu n n e l w e r e p r o c e s s e d i n t o Φ 5 0
mm × 100 mm cylindrical specimens. Most of the samples
were prepared according to the saturated standard, which
Investigations of the mechanism of large was based on BGB/T 50266–2013 (2013) (Standard for test
deformations in Huangjiazhai Tunnel methods of engineering rock mass)^, a Chinese national
standard, in which the procedure to saturate the rock spec-
Large-deformation phenomena in a soft rock tunnel could be imens was described. The immersion time was 2, 15, 30,
induced by several potential causes. The geological conditions 50, 70, and 90 days (‘immersion time’ refers to the time
are considered to be essential causes inducing large- elapsed since immersion of the specimens). First, the spec-
deformation problems in tunnels (Brox and Hagedorn 1999; imens were put into a trough, and water was injected into
Yassaghia et al. 2005; Sterpi and Gioda 2009; Khanlari et al. one-quarter of the specimen’s height. After 2 h, water was
2012; Meng et al. 2013; Agan 2016). The strength–stress ratio injected into half of the specimen’s height, and another 2 h
(SSR), which can be defined as the ratio between the uniaxial later, to three-quarters of the specimen’s height. After 6 h,
compressive strength of rock mass σcm and the in situ stress the specimens were fully submerged. The uniaxial com-
P0, is widely recognized as a factor that describes the extent of pression tests of the dry and saturated specimens were con-
rock squeezing (Hoek et al. 2000; Hsiao et al. 2009). The ducted on the Rock Mass Testing System (RMT), and the
mechanical properties of rock masses and geostress levels mechanical parameters of rock specimens with different

Fig. 6 Groundwater in
Huangjiazhai Tunnel
Mechanisms of large deformation in soft rock tunnels: a case study of Huangjiazhai Tunnel 437

Fig. 7 Anomalies in a b
Huangjiazhai Tunnel during the
tunneling process: (a) collapse
inside the tunnel; (b) squeezing of
the sidewall; (c) damage of the
primary support; and (d) surface
peeling of the secondary lining

Extrusion of the tunnel wall

c d

Surface concrete peeling off and

reinforcement mesh exposed

immersion times are shown in Table 2. Figure 10 shows the The relationship between the uniaxial compressive
fitting curves of the elastic modulus and immersion time of strength and immersion time is described by the following
rock specimens, and Fig. 11 shows the curves of the uni- expression (Eq. 1):
axial compressive strength and immersion time. The uni-
σc ¼ 60:45−8:82lnt ðt > 1Þ ð1Þ
axial compressive strength and elasticity modulus of the
rock samples gradually declined along with the increase where σc is the uniaxial compressive strength of shale
of immersion time, which indicated that the deforming re- (MPa) and t is the immersion time (day).
sistance and strength of shales weaken with an increase of Equation 1 shows that the uniaxial compressive strength
time in a wet environment. After the saturated status had had a negative logarithmic relationship with immersion time.
persisted for 90 days, the uniaxial compressive strength The uniaxial compressive strength of shale declined quickly in
and elasticity modulus of the rock samples were reduced the initial 20 days. With an increase in the immersion time, the
by 62% and 56% compared with the values in the dry state. rate of reduction decreased. When the saturated status had

vertical displacement at Section ZK61+605

vertical displacement at Section ZK61+600
600.0 horizontal displacement at Section ZK61+600
accumulative displacement (mm)

horizontal displacement at Section ZK61+605

500.0

400.0

300.0

200.0

100.0

0.0
8/19/2014

8/24/2014

8/29/2014

9/3/2014

9/8/2014

9/13/2014

9/18/2014

9/23/2014

Date
Fig. 8 Curves of displacement and time at two sections of the tunnel Fig. 9 Curves of contact pressure and time at Chainage ZK61 + 746
438 K. Bian et al.

Table 2 Mechanical parameters under different immersion times of the

rock specimens

Immersion time Elastic modulus Possion’s Uniaxial

(days) (GPa) ratio compressive
strength (MPa)

0 27.38 0.295 56.85

2 25.67 0.312 52.66
15 20.90 0.343 41.32
30 17.59 0.355 30.23
50 13.55 0.388 23.15
70 12.36 0.392 22.25
90 11.40 0.412 21.54
Fig. 11 Curves of uniaxial compressive strength and immersion time for
rock specimens

persisted for 50 days, the uniaxial compressive strength had

become approximately stable. the tests were conducted on a D8 Advanced X-ray diffractom-
Similarly, the relationship between the elasticity modulus eter. The shale samples, taken from eight different positions at
and immersion time could be described by the following Huangjiazhai Tunnel, were ground to powder. The scanning
fitting expression (Eq. 2): angle was between 3 ° and 50 °, and the scanning velocity was
5 °/min. The voltage and current were 40 kV and 80 mA,
E ¼ 29:46−3:87lnt ðt > 1Þ ð2Þ
respectively. The results showed that the shale samples were
where E is the elasticity modulus of shale (GPa) and t is the composed primarily of chlorite, quartz, muscovite, and albite,
immersion time (day). with a small amount of calcite, as shown in Table 3 and
It is important to note that to achieve the experimental Fig. 12. Among all of the components, the chlorite accounted
standard, the rock samples were collected from the shale with for the highest amount, ranging from 31.12% to 42.66%,
the best integrity and highest strength. However, most of the followed by the albite, with a content of 11.35–36.28%. The
rock masses in Huangjiazhai Tunnel were actually fragmen- content of muscovite was relatively fixed and was approxi-
tized, thin-layered, and jointed. Thus, the mechanical param- mately 20%.
eters in Table 2 are generally much larger than those of the
rock masses exposed at the excavation surface of tunnel. Microstructure observation of shale for different
immersion times

Mineral composition detection The microstructural features of the different minerals

contained in rock samples, including the type of minerals
The mineral compositions of the rock samples can be quanti- and crystal morphology, can be observed using scanning elec-
tatively identified by X-ray diffraction (XRD) analysis, and tron microscopy (SEM) techniques. A micrograph of a dry

Table 3 Mineral content of shale in Huangjiazhai Tunnel

Sample number Mineral content (%)

Calcite Albite Quartz Muscovite Chlorite

1 5.80 12.36 31.25 17.03 33.56

2 4.02 12.27 33.10 19.48 31.12
3 3.63 14.93 20.37 22.04 39.02
4 0.74 12.06 21.67 23.75 41.78
5 3.28 21.70 11.35 21.02 42.66
6 – 16.77 30.78 19.09 33.36
7 0.84 17.30 34.72 14.91 32.23
Fig. 10 Curves of elastic modulus and immersion time for rock 8 0.96 10.01 36.28 17.72 35.03
specimens
Mechanisms of large deformation in soft rock tunnels: a case study of Huangjiazhai Tunnel 439

Fig. 12 Mineral compositions of

the shale samples from
Huangjiazhai Tunnel

sample is shown in Fig. 13a, and the microstructures of shale of hydration, with a diameter of approximately 8 μm for a
for different immersion times are listed in Fig. 13b–d. peeling particle. At this stage, the face-to-face contact
Micrographs with different magnifications are shown in remained the primary style; however, as more mineral
Fig. 14, with an immersion time of 50 days. particles separated from the laminar structure, the number
In dry conditions, the chlorite crystals presented laminar- of point-to-face contacts increased. The compact laminar
structure aggregate morphologies, with a diameter of approx- structures became loose, and the edges of the structures
imately 50 μm. The microstructures were compact, with few gradually fragmentized. A large number of micro-pores
micro-pores and micro-cracks. The mineral particles were appeared among the mineral particles.
mostly bonded in a face-to-face contact style. After the saturated status had lasted for 20 days, the
When the samples had been saturated for 2 days, the mineral particles increasingly peeled off, and the cementa-
laminar structures of chlorite began peeling off as a result tion among mineral particles was weakened. In this case,

Fig. 13 Micrographs of shale for

dry and saturated conditions
(magnification: 2000): (a) dry; (b)
saturated for 2 days; (c) saturated
for 20 days; (d) saturated for 50
days
440 K. Bian et al.

Fig. 14 Microstructures of shale

with different magnifications
(saturated for 50 days): (a)
magnification: 800; and (b)
magnification: 5000

the microstructures of the samples were less compact, and geostress measurement, with a depth of 50 m, was located at
the scales of micro-pores and micro-cracks inside the shale Chainage K61 + 195 in the right tunnel (Fig. 15), and the buried
quickly expanded. The peeling particles scattered into depth of measurement was 260 m. The hydraulic fracturing
smaller ones, with a diameter of approximately 3 μm, fill- method, which has been widely used to measure geostress
ing in the micro-pores and micro-cracks. The flocculent (Matsunaga et al. 1989; Kuriyagawa et al. 1989; Hayashi and
microstructures emerged in this stage, making the shale Ito 1993; Haimson and Cornet 2003), was applied at
much more fragmentized. The dominant combination style Huangjiazhai Tunnel. The in situ test results are listed in
gradually changed from face-to-face to point-to-face Table 4. The hydraulic fracturing test procedure is as follows:
contact.
After the saturated status had lasted for 50 days, an increas-
(1) The borehole with a depth of 50 m along the vertical
ing number of flocculent microstructures had appeared due to
direction was drilled, and continuous cores were
continuous hydration, leading to increased damage to the orig-
extracted.
inal cementation among mineral particles. At this stage, the
(2) The test interval was sealed off by positioning the strad-
compact laminar structures mostly became porous loose ones,
dle packer at the planned depth, and the packers were
and the dominant combination style became point-to-face.
pressurized to a typical level of 3 MPa.
This analysis illustrates the reasons for the decrease of
(3) The test interval was pressurized for an initial qualitative
strength and deforming resistance of tunnel rock after water
permeability inspection (slug test).
infiltration from the micro perspective.
(4) The interval pressure was raised to maintain a constant,
predetermined flow rate.
Geostress test (5) Upon reaching breakdown pressure, pumping was
stopped.
The geostress test was carried out to assess the geostress level at (6) After the pore pressure had reached its original value,
the field of Huangjiazhai Tunnel. The vertical borehole for i.e., after waiting for a few minutes, the above

Fig. 15 Hydraulic fracturing test

at Section K61 + 195 of
Huangjiazhai Tunnel
Mechanisms of large deformation in soft rock tunnels: a case study of Huangjiazhai Tunnel 441

pressurization was repeated and cycled three times, using magnitude of geostress and rock strength. A quantitative eval-
the same flow rate. uation of the rock strength and geostress magnitude around the
(7) Finally, both the test interval and the packer pressure in situ test hole at Huangjiazhai Tunnel is given in this section,
were vented to allow the packers to deflate. utilizing the loading factors, e.g., σcm/P0 and Rc/σmax.
The rock mass rating (RMR) system was initially devel-
oped by Bieniawski (1974, 1989) on the basis of his experi-
As shown in Table 4, within the depth range of geostress ences in shallow tunnels. According to the geological explo-
measurement, the maximum horizontal principal stress at the ration data shown in Table 1, the value of RMR at the
borehole was 14.60 MPa. The results indicated that the max- geostress test areas of Huangjiazhai Tunnel (Chainage
imum principal stress and the second principal stress were K61 + 195) is 32.
both horizontal stresses and that only the minimum principal Over the past several decades, based on the rock mass
stress was in the vertical direction. The lateral pressure coef- classification system, many empirical equations for calculat-
ficient, i.e., the ratio of the maximum horizontal principal ing the strength of rock mass (σcm) have been proposed. The
stress and the vertical stress, reached nearly 1.8, indicating most widely used equations and the results of σcm at the
that the tectonic stress was the primary composition of geostress test areas are listed in Table 5. In these equations,
geostress at the tunnel site. Consequently, it was concluded σci, the strength of intact rock (MPa), is 56.85 MPa at the
that the site had experienced intense tectonic movements in geostress test areas, as shown in Fig. 11.
geological history, which could also be verified by the littery On the basis of the equations in Table 5, the average value
attitude of exposed rock masses at the excavation surface. of σcm at the drilling hole is 3.2 MPa. The in situ stress P0 at
the geostress test area is 14.6 MPa. The SSR, defined as σcm/
P0 (Wood 1972; Nakano 1979; Hoek and Marinos 2000), is
Mechanism of large deformation in Huangjiazhai 0.22 at Chainage K61 + 195, which is less than the cut-off
Tunnel value of σcm/P0 (0.25; between moderately squeezing rock
and severely squeezing rock) (Hsiao et al. 2009).
The mechanisms of large deformation in Huangjiazhai Tunnel Consequently, the rock masses at the abovementioned area
can be attributed to a number of reasons, both geological and of Huangjiazhai Tunnel are considered to be severely squeez-
mechanical. The main inducing factors are summarized in the ing rock.
following sections. In BGB/T 50218–2014 (2014) (Standard for Engineering
Classification of Rock Mass)^, a Chinese national standard
Plastic flow of rock masses induced by tunnel excavation describing the classification of engineering rock masses, the
under high geostress and low rock strength factor of Rc/σmax is adopted to evaluate the grade of the initial
stress field. Rc/σmax is defined as the ratio of uniaxial compres-
According to previous research (Anagnostou 1993; Meng et al. sive strength of saturated rock to maximum principal stress.
2013), excavation-induced plastic deformation is considered Similar to choosing 0.25 as the cut-off value of σcm/P0 to
one of the most important factors resulting in squeezing in soft differentiate moderately squeezing rock and severely squeez-
rock tunnels. The deformation in plastic flow will appear to ing rock (Hsiao et al. 2009), 4 is chosen to be the cut-off value
different extents, moderate or severe, depending on the of Rc/σmax to differentiate high geostress from extremely high

Table 4 Results of the geostress

test in vertical boreholes at Depth Maximum horizontal Minimum horizontal Vertical Direction of maximum
Huangjiazhai Tunnel (m) principal stress principal stress stress horizontal principal stress
(σH [MPa]) (σh [MPa]) (σv
[MPa])

18 2.44 2.14 7.23 /

22 6.41 4.30 7.33 /
26 13.20 7.01 7.44 /
30 12.61 6.56 7.54 /
37 11.90 6.56 7.72 /
39 14.33 7.85 7.77 NE8°
41 14.60 7.86 7.83 NE10°

NE north-east
442 K. Bian et al.

Table 5 The empirical equations and results of strength of rock mass 0.285% after saturation, and the increased proportion of ex-
(σcm) at the geostress test areas
pansive force reaches 124.4%; for chlorite, the respective
Researcher Equation σcm values are 0.047% and 111.8%. Another study (Zhu 1996)
(MPa) showed that the water molecules are adsorbed on the surface
of the chlorite particles, after which the polarized layers of
Kalamaris and RMR−100 3.34
Bieniawski σcm ¼ σci exp 24 ðMPaÞ water molecules form. Through continuous water absorption,
(1995) the layers of water molecules gradually thicken, causing the
Sheorey (1997)   1.90 volume expansion of chlorite. Since the volume expansion is
σcm ¼ σci exp RMR−100
20 ðMPaÞ
heterogeneous, differential expansion stresses appear inside
Trueman (1998) σcm = 0.5 exp(0.06RMR)(MPa) 3.41
the chlorite, inducing the dissolution of argillaceous cementa-
Aydan and 4.13
Dalgic (1998) σcm ¼ RMRþ6RMR
ð100−RMRÞ σci ðMPaÞ
tion; consequently, the shales gradually fragment.
The hydrated-mechanical coupling process in rock
masses is shown in Fig. 16. Due to the effects of dilatancy
geostress by summarizing the geological data from a number and unloading rebound of rock masses in the process of
of projects. An area is considered to bear extremely high stress tunnel excavation, the number of rock fractures increases,
when Rc/σmax is less than 4. and the fracture apertures increase in size. Groundwater
Based on the results of geostress tests and laboratory ex- continuously infiltrates into the rock mass along the rock
periments, the maximum value of Rc/σmax is 3.89 at the bot- fractures, thus accelerating the hydration process inside the
tom of the geostress test hole. Consequently, it indicates that shales. As time passes, the chlorite and muscovite at the
although the maximum buried depth (270 m) of the tunnel is surface of shales gradually swell and peel off, and more
not too large, the geostress test area of Huangjiazhai Tunnel is porous and flocculent microstructures appear at the peeling
still considered an extremely high stress area due to the high particles, making the structures increasingly incompact and
tectonic stress and low rock strength. fragmentized. The cementations among mineral particles
According to the Mohr-Coulomb criterion, the yielding of are gradually dissolved, and the cohesion and stiffness of
rock mass has close relationships with both the deviator stress shales decrease, thereby inducing larger deformations of
and the rock strength. The adjusted deviator stress after exca- rock masses after excavation. As a result, more rock frac-
vation in a high stress case is obviously higher than that in a tures expand, and more surfaces contacting water are ex-
low stress case, which leads to the rock fracture propagation posed inside the rock masses, thus aggravating the hydra-
and further accelerates the deformation and failure of rock tion action. The above cycles repeat and finally induce the
masses. Compared with the case of high rock strength, rock large deformation and even failure of the tunnel.
masses with low strength yield more easily, resulting in a
larger deformation of plastic flow. High geostress and low
rock strength are important factors in the large deformation
at Huangjiazhai Tunnel.
Unloading Deformation increase Dilatation
rebound of tunnel rock
Hydrated-mechanical coupling process in rock masses
Increase of rock Expansion of rock
The results of laboratory experiments show that the uni- fractures’ amount fractures’ opening
axial compressive strength and elasticity modulus of the
rock samples are gradually reduced along with an increase
in the immersion time, following a negative logarithmic Acceleration of water
Peeling off of the infiltration into rock Weakening of
relationship. This result indicates that the deforming resis- fractures cementations among
mineral particles
tance and strength of shales exposed at the excavation mineral particles
surface are quickly weakened in the initial 20 days after Aggravation of
the shales contact water. From the mechanical perspective, hydration inside shales

this is also why Huangjiazhai Tunnel deforms intensively

Generation of porous Structure becoming
during the continuous rainy days. and flocculent incompact and
The results of XRD analysis indicate that the chlorite and microstructures fractured
Reduction of strength
muscovite account for a large proportion of the mineral com- and stiffness of tunnel
position of shale. According to previous research (Teng et al. rock

2010), both have the characteristics of hydro-expansiveness. Fig. 16 Hydrated-mechanical coupled process in rock masses of the
For muscovite, the spacing of the crystal layer increases by Huangjiazhai Tunnel
Mechanisms of large deformation in soft rock tunnels: a case study of Huangjiazhai Tunnel 443

Conclusions temperature. Proceedings of the 23rd Symposium on Rock

Mechanics, University of California, Berkeley, California, August
25-27. New York: Society of Mining Engineers of the American
Severe deformation phenomena occurred frequently during Institute of Mining, Metallurgical and Petroleum Engineers, pp.
tunneling in Huangjiazhai Tunnel in Hubei Province, China. 261–269
To determine the mechanisms of such anomalies, comprehen- Anagnostou G (1993) A model for swelling rock in tunnelling. Rock
sive engineering, laboratory, and microscopic investigations Mech Rock Eng 26(4):307–331
Aydan O, Dalgic S (1998) Prediction of deformation behavior of 3 lanes
were conducted, including geostress tests, uniaxial compres- Bolu tunnels through squeezing rocks of North Anotolian Fault
sion tests, microstructure observations of shale with different Zone (NAFZ). Proceedings of the Regional Symposium on
immersion times, and mineral composition detection. Sedimentary Rock Engineering, Taipei, pp. 228–233.
The experimental results show that the uniaxial compres- Aydan O, Akagi T, Kawamoto T (1993) The squeezing potential of rocks
sive strength and elasticity modulus of the rock samples de- around tunnels: theory and prediction. Rock Mech Rock Eng 26(2):
137–163
crease gradually along with an increase in the immersion time, Aydan O, Akagi T, Ito T, Ito J, Sato J (1995) Prediction of deformation
following a negative logarithmic relationship. When the satu- behaviour of a tunnel in squeezing rock with time dependent char-
rated status lasts for 90 days, the uniaxial compressive acteristics. In: Proceedings of numerical models in geomechanics
strength and elasticity modulus of the rock samples are re- NUMOG V, pp. 463–469
Aydan O, Akagi T, Kawamoto T (1996) The squeezing potential of rock
duced by 62% and 56% compared with the values under dry
around tunnels: theory and prediction with examples taken from
conditions. This indicates that the deforming resistance and Japan. Rock Mech Rock Eng 29:125–143
strength of shales exposed at the excavation surface of Barla G (1995) Squeezing rocks in tunnels. ISRM News J II(3–4):44–49
Huangjiazhai Tunnel are quickly weakened in the initial Barla G, Bonini M, Semeraro M (2011) Analysis of the behaviour of a
20 days after the shales contact water. The mineral composi- yield-control support system in squeezing rock. Tunn Undergr
Space Technol 26:146–154
tion detection and microstructure observation results illustrate
Bieniawski ZT (1974) Geomechanics classification of rock masses and its
the reasons for the decrease in strength and deforming resis- application in tunneling. Proceedings of the Third International
tance of tunnel rock masses after water infiltration. Congress on Rock Mechanics, Vol. 11A. Denver: International
According to the results of in situ geostress tests, the max- Society of Rock Mechanics, pp. 27–32
imum horizontal principal stress at the borehole in Bieniawski ZT (1989) Engineering rock mass classifications. Wiley, New
York, p 251
Huangjiazhai Tunnel is 14.60 MPa. Although the maximum
Bizjak KF, Petkovsek B (2004) Displacement analysis of tunnel support
buried depth (270 m) of the tunnel is not too large, considering in soft rock around a shallow highway tunnel at Golovec. Eng Geol
the high tectonic stress and low rock strength, the geostress 75:89–106
test area of Huangjiazhai Tunnel can be considered an ex- Brox D, Hagedorn H (1999) Extreme deformation and damage during the
tremely high stress area. construction of large tunnels. Tunn Undergr Space Technol 14(1):
23–28
By combining engineering, laboratory, and microscopic
Cantieni L, Anagnostou G (2009) The interaction between yielding sup-
analyses, the large deformation at Huangjiazhai Tunnel can ports and squeezing ground. Tunn Undergr Space Technol 24:309–
be attributed to two primary triggering factors: the plastic flow 322
induced by tunnel excavation under high geostress and low Dalgic S (2002) Tunneling in squeezing rock, the Bolu tunnel, Anatolian
rock strength and the hydrated-mechanical coupling process motorway, Turkey. Eng Geol 67:73–96
Duda M, Renner J (2012) The weakening effect of water on the brittle
in the shales. failure strength of sandstone. Geophys J Int 192(3):1091–1108
Duncan FME (1993) Numerical modeling of yield zones in weak rocks.In
Acknowledgements The authors gratefully acknowledge the support comprehensive rock engineering. Rock Mech Min Sci 36(6):777–
by the National Key Research and Development Program of China 809
(Grant No. 2016YFC0401802), the State Key Program of National Erguler ZA, Ulusay R (2009) Water-induced variations in mechanical
Natural Science of China (Grant No. 51539002), the National Natural properties of clay-bearing rocks. Int J Rock Mech Min Sci 46(2):
Science Foundation of China (Nos. 51209198, 51204158, and 355–370
51409265), and the Natural Science Foundation of Zhejiang Province GB/T 50266-2013 (2013) Standard for test methods of engineering rock
(Grant No. LY13E090003). mass, Beijing: China Planning Press
GB/T 50218-2014 (2014) Standard for Engineering Classification of
Rock Mass. Beijing: China Planning Press
Gioda G, Cividini A (1996) Numerical methods for the analysis of tunnel
References performance in squeezing rocks. Rock Mech Rock Eng 29(4):171–
193
Agan C (2016) Prediction of squeezing potential of rock masses around Hadizadeh J, Law RD (1991) Water-weakening of sandstone and quartz-
the Suruc water tunnel. Bull Eng Geol Environ 75:451–468 ite deformed atvarious stress and strain rates. Int J Rock Mech Min
Aksoy CO, Ogul K, Topal I, Ozer SC, Ozacar V, Posluk E (2012) Sci Geomech Abstr 28(5):431–439
Numerical modeling of non-deformable support in swelling and Haimson BC, Cornet FH (2003) ISRM suggested methods for rock stress
squeezing rock. Int J Rock Mech Min Sci 52:61–70 estimation—part 3: hydraulic fracturing (HF) and/or hydraulic test-
Alm O (1982) The effect of water on the mechanical properties and ing of pre-existing fractures (HTPF). Int J Rock Mech Min Sci 40(7–
microstructures of granitic rock at high pressure and high 8):1011–1020
444 K. Bian et al.

Hayashi K, Ito T (1993) In situ stress measurement by hydraulic fractur- Mezger F, Ramoni M, Anagnostou G (2017) Evaluation of higher capac-
ing at the Kamaishi mine. Int J Rock Mech Min Sci Geomech Abstr ity segmental lining systems when tunneling in squeezing rock.
30(7):951–957 Tunn Undergr Space Technol 65:200–214
Hisatake M, Ohno S, Katayama T, Ohmae Y (2012) Effects of the ring- Nakano R (1979) Geotechnical properties of mudstone of neogene tertia-
cut method as a settlement deterrent in a soft ground tunnel. Tunn ry in Japan. Int Syrup Soil Mech Oaxaca 1:75–92
Undergr Space Technol 28:90–97 Sheorey PR (1997) Empirical Rock Failure Criteria. Balkema, Rotterdam
Hisatake M, Ohno S, Katayama T, Ohmae Y, Sano S (2009) Effects of the Singh M, Singh B, Choudhari J (2007) Critical strain and squeezing of
ring-cut excavation method on the restraint of displacements ahead rock mass in tunnels. Tunn Undergr Space Technol 22:343–350
of a tunnel face. Tunn Undergr Space Technol 24:547–554 Steiner W (1996) Tunnelling in squeezing rocks: case histories. Rock
Hsiao FY, Wang CL, Chern JC (2009) Numerical simulation of rock Mech Rock Eng 29(4):211–246
deformation for support design in tunnel intersection area. Tunn Sterpi D, Gioda G (2009) Visco-plastic behaviour around advancing tun-
Undergr Space Technol 24:14–21 nels in squeezing rock. Rock Mech Rock Eng 42:319–339
Hoek E (2001) Big tunnels in bad rock. J Geotech Geoenviron 127(9): Tan JK, Kang WF (1980) Locked in stresses, creep and dilatancy of
726–740 rocks, and constitutive equations. Rock Mech 13:5–22
Hoek E, Guevara R (2009) Overcoming squeezing in the Yacambu- Teng HW, Ren S, Jiang DY, Yang CH (2010) Experimental study of
Quibor tunnel, Venezuela. Rock Mech Rock Eng 42:389–418 mechanical properties of water-saturated weaken shale in Gonghe
Hoek E, Marinos P (2000) Predicting tunnel squeezing problems in weak tunnel. Chin J Rock Mech Eng 29(S1):2657–2662
heterogeneous rock masses. Tunnels Tunnelling Int 32(11):45–51
Trueman R (1998) An evaluation of strata support techniques in dual life
Jiang Q, Jiang Y, Cui J, Feng X (2014) Application of computerized
gateroads. Ph.D Thesis. University of Wales, Cardiff
tomographic scanning to the study of water-induced weakening of
Van Eeckhout EM (1976) The mechanisms of strength reduction due to
mudstone. Bull Eng Geol Environ 73:1293–1301
moisture in coal mine shales. Int J Rock Mech Min Sci Geomech
Kalamaris GS, Bieniawski ZT (1995) A rock mass strength concept for
Abstr 13(2):A22
coal incorporating the effect of time. Proceedings of the Eighth
International Congress on Rock Mechanics. Vol. 1. Rotterdam: Walton G, Delaloye D, Diederichs MS (2014) Development of an ellip-
Balkema, pp. 295–302 tical fitting algorithm to improve change detection capabilities with
Kavvadas M (2005) Monitoring ground deformation in tunnelling: cur- applications for deformation monitoring in circular tunnels and
rent practice in transportation tunnels. Eng Geol 79(1–2):93–113 shafts. Tunn Undergr Space Technol 43:336–349
Khanlari G, Meybodi RG, Mokhtari E (2012) Engineering geological Wang MY, Zhang N, Li J, Ma LJ, Fan PX (2015) Computational method
study of the second part of water supply Karaj to Tehran tunnel with of large deformation and its application in deep mining tunnel. Tunn
emphasis on squeezing problems. Engineering Geology 145– Undergr Space Technol 50:47–53
146(3): 9–17 Wood AM (1972) Tunnels for road and motorways. Q J Eng Geol 5:119–
Kovfiri K, Staus J (1996) Basic considerations on tunnelling in squeezing 120
ground. Rock Mech Rock Eng 29(4):203–210 Yang TH, Xu T, Liu HY (2014) Rheological characteristics of weak rock
Kuriyagawa M, Kobayashi H, Matsunaga I, Yamaguchi T, Hibiya K mass and effects on the long-term stability of slopes. Rock Mech
(1989) Application of hydraulic fracturing to three-dimensional in Rock Eng 47:2253–2263
situ stress measurement. Int J Rock Mech Min Sci Geomech Abstr Yassaghia A, Salari-Rad H (2005) Squeezing rock conditions at an igne-
26(6):587–593 ous contact zone in the Taloun tunnels, Tehran-Shomal freeway,
Matsunaga I, Kuriyagawa M, Sasaki S (1989) In situ stress measurements Iran: a case study. Int J Rock Mech Min Sci 42:95–108
by the hydraulic fracturing method at Imaichi pumped storage pow- Yilmaz I (2010) Influence of water content on the strength and
er plant, Tochigi, Japan. Int J Rock Mech Min Sci Geomech Abstr deformability of gypsum. Int J Rock Mech Min Sci 47(2):342–347
26(3–4):203–209 Zhu JB, Wang B, Yang HP, Hu JM (2007) Experimental study on rheo-
Meng LB, Li TB, Jiang Y, Wang R, Li YR (2013) Characteristics and logical mechanical properties of shale under unloading condition.
mechanisms of large deformation in the Zhegu mountain tunnel on Chin J Rock Mech Eng 26(Supp.2):4552–4556
the Sichuan-Tibet highway. Tunn Undergr Space Technol 37:157– Zhu XJ (1996) Characteristics of soft rocks interacting with water.
164 Scientific Technol Min 4(3):46–50