International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68

Contents lists available at ScienceDirect

International Journal of
Rock Mechanics & Mining Sciences
journal homepage: www.elsevier.com/locate/ijrmms

Experimental and theoretical study of the anisotropic

properties of shale
Shuai Heng a,n, Yingtong Guo a, Chunhe Yang a, Jack J.K. Daemen b, Zhi Li a
a
State Key Laboratory of Geomechanics and Geotechnical Engineering, Institute of Rock and Soil Mechanics, Chinese Academy of Sciences,
Hubei 430071, Wuhan, China
b
Mackay School of Earth Sciences and Engineering, University of Nevada, Reno, NV, USA

art ic l e i nf o a b s t r a c t

Article history: The effect of bedding plane orientations on shear strength and failure mechanisms of shale is investigated,
Received 3 April 2014 with an emphasis on the shear parameters of the bedding planes. This study is based on direct shear tests
Received in revised form conducted on directional shale samples obtained from formations which exhibit well-developed stratifica-
22 December 2014
tion. The tested cylindrical specimens were prepared by coring at 301 intervals from the direction of the
Accepted 2 January 2015
bedding planes. The results indicate that the anisotropy of the shear strength of shale can be readily
Available online 22 January 2015
observed due to the anisotropic effects of bedding plane orientation and normal stress. The shear
Keywords: parameters, such as strength, cohesion and internal friction angle reach their maximum and the minimum
Anisotropy values at β ¼601 and β ¼ 01, respectively, where β is the angle between the bedding planes and the coring
Bedding planes
orientation, which is also the direction of normal stress during the laboratory tests. It is demonstrated that
Direct shear test
the bedding planes are weak not only with regards to cementing strength, but also with regards to friction.
Shear strength
Failure mechanism Three different shear failure modes were built up depending on the shearing angle β: (1) sliding failure
Shear stress concentration factor across the bedding planes, (2) sliding failure along the bedding planes and (3) sliding failure across the
bedding planes combined with tensile splitting along the bedding planes. An expression of the shear stress
concentration factor of the direct shear model is derived to evaluate the anisotropy of shear strength.
& 2015 Elsevier Ltd. All rights reserved.

1. Introduction assumption of isotropy. Erroneous results may be produced when

Anisotropy is one of the most distinct features that must be paid
due attention in rock engineering whether it is in civil engineering,
foundations, slopes or mining, ground excavations, or petroleum
engineering. The anisotropy of the properties of rocks is mainly due
to the presences of cleavage, foliation, bedding planes, schistosity,
joints and micro or macro fissures [1]. These structural character-
istics make the mechanical, seismic, hydraulic properties and
thermal properties of rocks vary with orientation [2–4]. Hence, the
anisotropic discontinuities and the degree of their influence must be
analyzed carefully in rock mechanics and engineering.
Rock anisotropy is of importance for in-situ stress measurements
and for stability analysis of underground excavations in mining
engineering and underground constructions [5–8]. Rock anisotropy
affects TBM (tunnel boring machine) rock-cutting performance
[9,10]. In petroleum engineering, rock anisotropy is a critical factor
in controlling borehole stability, deformation and failure. It also
affects fracturing and fracture propagation [11–14]. But up to now,
much of available engineering methodology is built on the
n
Corresponding author. Tel.: þ 86 13628659132.
E-mail address: shheng@yeah.net (S. Heng).
ignoring the anisotropy for some rocks [12,15]. As the stress
distribution around the borehole in shale formation is determined
by assuming that the rock formation is elastically isotropic, but only
the anisotropic rock strength characteristic is took into account. That
is, only the anisotropy of failure strength is considered in these
studies. Therefore, it is essential to be aware of the degree of
mechanical anisotropy to decide whether it is relevant and necessary
to consider anisotropy before certain operations begin.
Most foliated metamorphic rocks, such as schist, slates, gneisses,
and phyllites contain fabric with preferentially parallel arrangements
of flat or elongated minerals. Metamorphism changes the initial fabric
of rocks with the directional structure. Foliation induced by the non-
random orientation of macroscopic minerals, parallel fractures or
microscopic mineral plates, such as fracture cleavage, slaty cleavage,
bedding cleavage, lepidoblastic schistosity, nematoblastic schistosity or
lineation lead to rock properties that are highly direction-dep-
endent [16,17]. As a result, such rocks display strongly anisotropic
behavior in response to load orientation with respect to weak planes.
The anisotropic behavior can also be found in bedded sedimentary
rocks, such as siltstone, sandstone, shale or sandstone–shale [18]. This
anisotropy originates from the bedding planes that have formed
during the sedimentation phase [19]. These metamorphic and sedi-
mentary rocks are inherently anisotropic, called transversely isotropic

http://dx.doi.org/10.1016/j.ijrmms.2015.01.003
1365-1609/& 2015 Elsevier Ltd. All rights reserved.
 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68
rocks. Transverse isotropy implies that at each point in the rock there
is an axis of rotational symmetry and the rock has isotropic properties
in the plane normal to that axis, this plane being the plane of trans-
verse isotropy [20].
Many investigators have devoted considerable efforts to study-
ing the strength and failure modes of anisotropic rocks in uniaxial
and triaxial conditions [21–34]. An overall review and analysis
of the experimental results exhibit that the maximum failure
strength is either at δ ¼01 or δ ¼901, and the minimum value
usually is around δ ¼301, more precisely at (451  φ/2), where δ is
the angle between the structural planes and the direction of major
principal stress and φ is the internal friction angle along the plane
of weakness [29,35]. With increasing confining pressure in triaxial
compression tests, the anisotropic rocks become more ductile, and
the effect of the strength anisotropy usually reduced. But the
degree of mechanical anisotropy, which is generally defined by the
elastic modulus ratio, customarily increases [33,36]. The results
also show that the failure modes of anisotropic rocks in uniaxial
and triaxial compression tests are influenced by the loading
orientation, as well as by the confining pressure. The failure
modes can be divided into two categories: (1) sliding mode, in
which the plane discontinuity predominated and (2) non-sliding
mode, in which the material strength dominated.
Many attempts have been made to describe the strength and
failure anisotropy of inherently anisotropic rocks. A series of works
have considered the effect of weak planes on the failure strength
of anisotropic rocks. Most of them focused on the influence of the
weak planes on the compressive and tensile strength. This effect is
generally considered in the stability analysis of engineering rock
masses in drilling, slopes, tunnels, and large caverns where shale
formations are often encountered. Stress analysis assuming iso-
tropy can be inaccurate and often underestimate fracturing pres-
sure [11,13,14,37]. Different failure modes may occur under a
foundation at slopes depending on boundary condition and the
inclination of weak planes for inherently anisotropic rocks [38].
Bedding plane orientation changes the rockburst behavior in
mining and tunneling construction [39]. Shear stress concentra-
tion usually exists near the interface between weak planes and the
matrix of the rocks, and this may easily result in shear slippage or
damage [40]. However, most of these attempts have considered
the bedding orientation on the failure strength and failure modes
in uniaxial and triaxial conditions and Brazilian test conditions.
But the failure strength and failure modes of anisotropy rocks in
shearing are also necessary for us to analyze the propagation of
shear fractures in drilling, tunnels and hydraulic fracturing, espe-
cially exploitation of shale gas [41–43].
Hydraulic Fracturing by Stimulated Reservoir Volume (SRV) [44] is
a major technology to achieve commercial development of shale gas.
However, the formation mechanism of effective fracture network has
not been well understood. The mechanism of fracture propagation in
shale with hydraulic fracturing needs to be further explored, in order
to realize the control on morphology of fracture network with SRV
and increase the single well production of shale gas. The key factors
affecting the morphology of post-fracturing network include hor-
izontal in-situ stress difference, rock brittleness and natural fracture
system (sedimentary bedding) [45–47]. The hydraulic fracturing
simulation experiments, conducted in a large-scale true triaxial
physical simulation test system, show that the hydraulic fracture
morphology is strongly influenced by the development degree of
bedding planes [48]. Thus, bedding planes have a great influence on
the propagation of fractures approaching bedding planes. The
propagation mechanism of fractures approaching bedding planes is
an important part of the in-depth understanding of the formation
mechanism of fracture network of shale gas reservoirs.
There are two types of fractures, which are shear fractures and
tensile fractures, intersect with bedding planes with different
59
approach angles in shale gas reservoir. For the shear fractures,
the relevant variables are stress traction vector with one normal
component and one tangential in a simple two-dimensional stress
field (or two tangential in 3D), and the conjugate “strain” variables
are the corresponding relative displacements [49]. However, such
stress conditions can be only realized in direct shear tests, which
are, therefore, of urgent need to reproduce slip between rock mass
either side of the fracture. Hence, the direct shear tests are conducted
on shale specimens with different bedding orientations, to investigate
the shear strength and shear fracture propagation when intersecting
with bedding planes with different approach angles.
It is recognized that the direct shear test has several inherent
defects such as the principal stress rotation, stress non-uniformity
(stress concentration on the shear plane), and the failure plane
definition during shearing [50,51]. Despite these problems, the test
is still used as one of the methods to evaluate the shear strength of
certain planes or certain orientations of anisotropic rocks, as the
failure plane is defined in the plane between the upper and lower
shear boxes during shearing. Thus, it is an effective approach to
appraise shear strength parameters of bedding planes and matrix,
even the shear strength and shear fracture propagation of shale with
different approach angles between shear direction and bedding
planes. Consequently, direct shear tests, conducted in shale speci-
mens with different bedding orientations, are to investigate the
shear strength and shear fracture propagation when shear fracture
intersects with bedding planes with different approach angles.
In this study, cylindrical shale specimens with different shearing
angles are used to carry out direct shear tests. The mechanical
properties of bedding planes and shear strength of shale with
different bedding orientations are determined based on the experi-
mental data. Finally, the anisotropy of shear strength is analyzed
based on the shear stress concentration factor, which enables a
detailed theoretical interpretation of the results and a specification
of the factors that influence the anisotropy of shear strength.
2. Direct shear tests on anisotropic shale specimens
2.1. Sample preparation
As relatively intact core samples could not be obtained from shale
gas wells deep over 2 km in the shale gas blocks of Pengshui in
Chongqing City of China, the anisotropic shale specimens tested were
prepared from the outcrops of the Longmaxi Formation in Shizhu
County, which is the natural extension of the formations in shale gas
blocks of Pengshui (Fig. 1). The shale formation in the outcrops is a
black-dark carbonaceous shale, with alternating thin and thick layers.
The cohesive strength between layers is so small that it is sensitive to
weathering and easily splits into pieces. The dip angle of the
formations is about 701. Well-developed bedding planes are
Fig. 1. The shale outcrops of the Longmaxi formation where shale specimens tested
were collected.
60 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68
exhibited, and wave ripples are developed in stratification faces,
where abundant graptolites and radiolarite are detected. Fissure veins
are filled with scattered pyrite, pyrite nodules, quartz and calcite.
In order to analyze the mechanical properties of bedding planes
and the anisotropy of shear strength, cylindrical specimens with
shearing angles of β ¼ 01, 301, 601 and 901 were cored out of large
blocks of the rock (Fig. 2(a)). The prepared specimens all have
a length-to-diameter ratio of 2:1 (or slightly different), with a
diameter of 50 mm (or slightly smaller). The errors of length and
diameter of specimens are within70.5 mm. The parallelism of speci-
men ends is within70.02 mm.
2.2. Experimental devices and test principle
The direct shear tests were carried out on RMT-150C, which is
self-developed by Wuhan Institute of Rock and Soil Mechanics
(WHRSM) and can be used for various mechanical experiments of
rocks and concrete. The force diagram of specimens in direct shear
tests is shown in Fig. 3. The test processes are: first, put the
specimen in the shear box; second, apply the normal force on the
top; finally, exert shear force on the specimen until it fails.
Fig. 2. Directional coring diagram of specimens with different bedding orientations
(a) Directional coring diagram for shale sample preparation (b) Diagram of the
angles α, β and δ (when the coring direction is the same as the direction of the
major principal stress, δ ¼β).
N
shearing plane
Q
Q
specimen bedding planes
N
Fig. 3. Force diagram of specimens in direct shear test.
In the direct shear tests, α is the angle between the orientation
of bedding planes and the shear force, and it should never be
larger than 901, that is, the shear force is applied with an acute
angle or right angle relative to the bedding orientation in the tests.
The relationship between α and β is as follows:
α ¼ 901  β ð1Þ
A diagram of the angles α, β and δ is shown in Fig. 2(b).
The normal stress σn acting on the shear plane and the shear
stress τ in the predetermined shearing plane can be calculated as
follows:
(
σ n ¼ NA
ð2Þ
τ ¼ QA
where N is the normal force, Q is the shearing force, and A is the
nominal area of the specimen along the shear direction.
The relationship between shear strength τ and normal stress σn
is approximately linear when the normal stress is relatively small
according to the Mohr–Coulomb criterion, which can be expressed
as follows:
τ ¼ σ n tan φ þc ð3Þ
where c and φ are the cohesion and internal friction angle of the
shear failure plane, respectively.
The Mohr–Coulomb diagram representing the state of stress on
the shear area is shown in Fig. 4, where σ 1 and σ 2 are two different
normal stresses, and τ1 and τ 2 are the corresponding shear strengths.
The curvature of the envelope can be disregarded when the normal
stress is relatively small, so the envelope can be considered a straight
line, as shown in Fig. 4. In such case two points are sufficient to
calculate the parameters c and φ for a given β.
2.3. Experimental procedure
Considering that the uniaxial compressive strength of shale
specimens prepared from the outcrops of the Longmaxi Formation
for β ¼301 is 47.22 MPa, much less than that for β ¼01, 601 and 901
(121.64, 108.76 and 124.26 MPa, respectively), due to the easy sliding
along bedding planes for that angle, under uniaxial compression
tests [52]. The normal stresses applied for testing of specimens with
β ¼ 301 are divided into four levels (5, 12.5, 20 and 27.5 MPa), which
can reduce the failure possibility of the specimens when normal
force is loaded. For specimens with β ¼01, 601 and 901, the normal
stresses are 12.5, 25, 37.5 and 50 MPa.
When testing, the normal force is loaded at the rate of 1 kN/s to
the predetermined value. Then, keeping the normal stress constant,
the shearing force is applied under the horizontal shear displace-
ment control mode, at the rate of 0.002 mm/s. The test terminates
when the shear stress reaches residual shear strength.
Fig. 4. Mohr–Coulomb envelopes
 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68 61

3. Experimental results and analysis of the tests 60

12.5MPa
3.1. Relationship between shear stress and shear displacement 50 25MPa
37.5MPa
The shear stress–shear displacement curves of shale specimens 50MPa

shear stress(MPa)
40
with various orientations under different normal stresses are plotted
in Fig. 5. Significant slip-weakening of shear strength is indicated in 30
shear stress–shear displacement of the shale, as shown in Fig. 5. The
same features shown in these curves for β ¼ 01, 301 and 601 are as 20
follows:
10
i) At the initial loading stage, the slope of the curve is relatively
small. The shear stress increases slowly but the shear displace- 0
0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00
ment rapidly. This is mainly due to the closure of interspaces
shear displacement(mm)
between the shear box and specimens and the closure of
fractures or micro-fractures of specimens.
ii) When the shear displacement exceeds a certain value, which is 35
5MPa
a turning point in the shear stress–shear displacement curve,
12.5MPa
the slope of the curve increases rapidly. At this stage micro- 30
20MPa
cracks are created in the specimens due to tensile effect of 27.5MPa
25

shear stress(MPa)
shear stresses on bedding planes with the increase of shear
force. These fractures are probably along the bedding planes 20
rather than along the intended shear plane. When the shear
15
stress is near the peak, the slope of the curve changes from
steep to gentle. Distinct brittleness is present at this stage 10
owing to the extremely small plastic deformation.
iii) At the peak, there is an obvious inflexion in the curve. The 5
shear stress drops sharply as the shear displacement continues
0
to increase. The slope of the curve changes from positive to 0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1 2.4
negative. This stage is the so-called shear stress drop. The shear displacement(mm)
common features of slip-weakening of shear strength of rocks
are shown at this stage. But the difference of this stage is also 70
clear for different values of β. 12.5MPa
iv) After the shear stress decreases to a certain degree, the slope 60 25MPa
37.5MPa
of the curve becomes very small, and the shear stress reaches 50MPa
50
shear stress(MPa)

residual shear strength.

40

For the specimens with β ¼ 901, an obviously distinct trend

30
especially the post-peak stage is shown in the shear stress–shear
displacement curves as then the predetermined shear plane is a 20
bedding plane. When the normal stress is 12.5 MPa, the whole
curve is relatively smoother, and the shear strength is slightly 10
larger than the sliding friction of interlaminations under the
0
relatively lower normal stress, which is mainly due to the bedding 0.00 0.25 0.50 0.75 1.00 1.25 1.50 1.75 2.00 2.25
planes are weak planes and shear strength of bedding planes is shear displacement(mm)
small. Hence a not clear shear stress drop can be seen in the shear
stress–shear displacement curves. But an abrupt shear stress drop 45
occurs after the peak strength when the normal stresses are 25,
40 12.5MPa
37.5 and 50 MPa, even the drop directly from peak to residual 25MPa
strength can be observed under the normal stress of 50 MPa. This 35
37.5MPa
shear stress(MPa)

phenomenon illustrates that the sudden release of interlaminated 30 50MPa

cohesion when the shear plane begins to slide, it further indicates
25
the weakness of cohesion strength of bedding planes. And a larger
momentary shear stress drop can be observed with the larger 20

normal stress. As the intended shear plane is a bedding plane for 15

the specimens with β ¼901, the shear stress drop is close to a 10
vertical drop, different from for that with β ¼ 01, 601 and 901.
Moreover, the change of the residue stage with β ¼901 is gentler. 5

0
0.0 0.3 0.6 0.9 1.2 1.5 1.8 2.1

3.2. Anisotropic shear strength shear displacement(mm)

Fig. 5. Plots of shear stress versus. shear displacement of shale specimens under
Table 1 presents the direct shear tests results of shale speci- different normal stresses at orientations of: (a) β¼ 01 (b) β ¼ 301 (c) β ¼ 601 and
mens with different orientations under different normal stresses. (d) β ¼901.
62 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68

Table 1 isotropic rock materials, the fracture pattern is more complicated

Results of direct shear tests of shale specimens for different orientations under and composed of mixed-mode fractures along the weak planes
different normal stresses.
and the matrix, which is different from the isotropic rocks. This is
β Specimen Normal Normal Normal Shear Shear Residual the main reason for the maximum failure strength is not obtained
(1) no. shear force N stress σn force stress shear in the direction perpendicular to the bedding planes for transver-
area A (kN) (MPa) peak Q peak τ stress τd sely isotropy rocks [53,54].
(mm2) (kN) (MPa) (MPa) The shear strength parameters of this shale for different angles
0 Y0-15 1923.761 24.047 12.5 51.815 26.934 14.256
β are displayed in Table 2. For β ¼901, specimens were sheared
Y0-24 1922.76 48.069 25 59.913 31.16 19.353 almost purely along the bedding planes. In this case, the shear
Y0-5 1897.93 71.172 37.5 79.220 41.74 31.352 strength parameters tested are the values of bedding planes. The
Y0-9 1907.222 95.361 50 100.987 52.95 33.649 cohesion and internal friction angles of bedding planes are all
30 Y3-17 1912.426 9.562 5 24.766 12.95 5.350 significantly less than those for the other values of β. This further
Y3-16 1907.248 23.841 12.5 45.831 24.03 14.997 shows that the bedding planes are weak planes in shale forma-
Y3-10 1908.773 38.175 20 49.800 26.09 13.838 tions, and their failure tends to precede failure of the matrix due to
Y3-15 1916.537 52.705 27.5 63.706 33.24 14.362
the much weaker cementation of the bedding planes.
60 Y6-19 1919.352 23.992 12.5 55.865 29.106 21.028 The internal friction angle increases in order of β ¼901, 01, 301
and 601, as presented in Table 2. The values for β ¼01, 301 and 601
Y6-16 1908.773 47.719 25 82.793 43.375 32.126
Y6-17 1907.222 71.521 37.5 99.576 52.21 39.226
Y6-18 1908.773 95.439 50 127.831 66.97 50.853 are obviously larger than those of weak planes, which illustrates
that the bedding planes are not only weak on the cohesion but
90 Y9-24 1924.243 24.053 12.5 31.329 16.281 12.108
Y9-23 1922.426 48.061 25 49.401 25.697 13.658
also on the internal friction angle. For cohesion of shale, the value
Y9-21 1916.634 71.874 37.5 61.511 32.093 17.069 for β ¼301 is close to that of the bedding planes mainly due to the
Y9-20 1914.983 95.749 50 77.384 40.41 27.176 splitting effect of shear stress on weak planes, which is signifi-
cantly less than the maximum value for β ¼601.
Fig. 7 indicates that the anisotropy of residual shear strength of
shale is also distinct. The effect of normal stress on residual shear
strength can be clearly seen. The residual shear strength for β ¼901
is the minimum mainly due to the shearing almost pure on the
weak planes and the friction surface is much smoother than the
others. The residual shear strength for specimens with β ¼ 601 is
maximum for all the normal stresses, which is the same as shear
strength observed in Fig. 6. The reason for this may be the friction
surface is much rougher than the others, due to the splitting along
bedding planes when shearing. To a great extent, the variation
tendency of the residual shear strength is the same as the shear
strength with the change of β, owing to the residual friction
bearing capacity mainly depending on the friction coefficient of
the fractured surface formed when shearing, according to Amon-
tons' friction law [55], yet the fractured surface shapes and shear
strength are all dominated by the failure mechanisms.
Fig. 6. Mohr–Coulomb envelopes of shale for different orientations of β.

Plots of peak shear strengths versus normal stresses, for different Table 2
Shear strength parameters of shale for different orientations of β.
orientations of β, are shown in Fig. 6. The anisotropy of the shear
strength is pronounced, resulting from the presence of bedding β (1) Fitted envelopes R2 φ (1) c (MPa)
planes. Shear strength for β ¼901 is significantly less than that for
other β values at the same normal stress. The shear strength of 0 τ ¼ 0:709σ n þ 16:039 0.949 35.337 16.039
bedding planes is lower compared with those of the other orien- 30 τ ¼ 0:839σ n þ 10:434 0.902 39.996 10.434
60 τ ¼ 0:979σ n þ 17:309 0.987 44.421 17.309
tations, due to the weak cohensive strength of bedding planes. 90 τ ¼ 0:630σ n þ 8:92 0.992 32.211 8.92
Fig. 5 also indicates that the shear strength for β ¼ 601 is the
maximum, not in the orientation of β ¼01,which may be mainly
due to not a pure shear type of failure occurs in shearing, as the
failure is a combination of shear sliding across bedding planes
accompanied by tensile splitting along bedding planes, even
splitting of bedding planes throughout the entire specimen can
be observed in the faiured specimens. The relatively complex
failure patterns result in higher failure strength. But for the
specimens of β ¼01, only shear sliding across beding planes occurs
in shearing. For β ¼301, the shear strength is less than that of β ¼01
when the normal stresses are less than 40 MPa. The main reason
for this is the same as the specimens of β ¼601, due to the mixed
failure mode and complex fracture mechanisms along the bedding
planes and matrix. But the conclusions are the opposite when the
normal stresses are greater than 40 MPa, the main reason for this
is the more easily occurring of tensile splitting of bedding planes
in shearing as the lower confinement of normal stress, which
induces smaller shear strength. In many cases, for transversely Fig. 7. Residual shear strength of shale with respect to the variations of σn and β.
 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68 63

12.5 MPa 50 MPa

12.5 MPa 25 MPa

12.5 MPa 50 MPa

12.5 MPa 50 MPa

Fig. 8. Failure patterns of shale for different orientations of β under a series of normal stresses in direct shear tests: (a) at β¼01 under the normal stress of 12.5 and 50 MPa (b) at
β¼ 301 under the normal stress of 12.5 and 25 MPa (c) at β¼ 601 under the normal stress of 12.5 and 50 MPa and (d) at β¼ 901 under the normal stress of 12.5 and 50 MPa.
64 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68

3.3. Anisotropic failure mechanisms

Failure patterns of shale specimens with different β after direct

shear tests are shown in Fig. 8, which indicates that the shear failure
patterns of shale specimens with different β show obvious aniso-
tropy. Under different normal stresses, distinct fractured surfaces are
also observed on the samples with the same orientation. The detailed
descriptions of shear fractured surface features are as follows:

(a) For β ¼01, 301 and 601, the failure surface changes from rough to
smooth as the normal stress increases, but for β ¼ 901, the
fracture surface is almost completely along the bedding planes;
and graptolites and radiolarites can be observed on the extre-
mely smooth fractured face.
(b) Splitting of beding planes can hardly be observed for β ¼ 01,
but bedded structure can easily be seen from the rough failure
surface when the normal stress is 12.5 MPa. Local scratch
traces on the failure surface are particularly pronounced at the
normal stress of 50 MPa.
(c) Splitting of bedding planes can be seen for β ¼301 under every
tested normal stress, but the phenomenon weakens when the
normal stress increases, which illustrates that the normal stress
weakens the bedding spliting effect induced by the shear stress.
(d) More obvious splitting of bedding planes can be noticed for
β ¼601 and splitting throughout the specimen can be observed
under all levels of normal stresses, which may be one of the
reasons why the maximum shear strength is reached in this
orientation.

There are plenty of particles and powders around the shear
plane after failure. The shear fractured surface is not just a plane
but a fracture zone. A typical shear fractured surface is shown in
Fig. 9. The apparently splitting of bedding planes can be seen after
the fragments and particles were cleared, and the fractured sur-
face is still not a plane. What is more, the layered sedimentary
structure is shown, which is similar to the shale formations as
shown in Fig. 1. This illustrates that the shear fractures across the
matrix accompanied by splitting of bedding planes occur in
shearing, and the particles and fragments are formed mainly in
the process of shear sliding.
From the analysis of the relationship between shear failure
surface, bedding planes and the loading orientation, the shear
failure mechanisms of shale can be divided into following three
types (Fig. 10):
(1) Shear strength for β ¼01 is dominantly controlled by the matrix
as shearing direction is perpendicular to the bedding planes, and
failure mechanism is shear sliding across weak planes.
(2) Shear strength for β ¼301 and 601 is controlled by the matrix and
bedding planes as tensile stresses are induced on the bedding
planes which tend the weak planes to split and detach; and the
failure mode is a combination of shear sliding across weak planes
accompanied by tensile splitting along weak planes.
Fig. 10. Typical failure mechanisms in the direct shear test. (a) Shear sliding across
the weak planes for β ¼ 01. (b) Tensile splitting along the weak planes accompanied
by sliding across the weak planes for β ¼ 601.Bedding splitting can be observed
throughout the specimen. (c) Sliding along the weak planes for β¼ 901.
(3) Shear strength for β ¼ 901 is mainly controlled by the bedding
planes as shearing planes are almost purely along the bedding
planes, and failure mode follows sliding along weak planes.
The layered sedimentary structure and the weak cementing of
bedding planes are the main causes of the anisotropy of failure
mechanisms, and the anisotropic shear strength is dominated by
the anisotropy of failure mechanisms.
Based on the anisotropy of the failure mechanisms and the
relative relationship between the shear directions and bedding
planes, the shear strength parameters of matrix and bedding planes
of shale are obtained respectively from the specimens with β ¼01
and β ¼901. The shear strength parameters of matrix and bedding
planes of shale are as follows: the cohesion of matrix and bedding
planes are 16.039 MPa and 8.92 MPa. 35.3371 and 32.2111 are the
internal friction angle of matrix and bedding planes, respectively.
Consequently, weak cohesion is pronounced for bedding planes
of shale compared with the matrix, and similar conclusions could
be obtained for the internal frictional angle even though it is not
particularly obvious, that is, the weak planes in inherently aniso-
tropic shale are weak not only on the cementing strength, but also
on the friction. That is the reason why the lowest shear strength
and residual shear strength are observed for the bedding planes of
shale (Fig. 7).

4. Theoretical analysis of the anisotropic shear strength

Three simplified assumptions are used in the direct shear model

to analyze the anisotropic shear strength of shale, which are as
follows: (1) The moment caused by eccentric force when shearing is
ignored. (2) Only shear force is applied in the direction of intended
Fig. 9. Typical shear fractured surface of shale for β ¼601. shearing in the shear layer. (3) The shear stress in the shear layer is
 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68 65

proportional to the relative displacement between the upper and Substitution of Eq. (9) into Eq. (11), and using Eq. (5) yields the
lower non-shear layers, that is, the stresses in the shear layer are in following expression:
elastic state when analyzed. 2  
d T1 1 1 kP
For the sake of simplicity, only applied forces in the x-direction are  k þ T1 þ ¼0 ð12Þ
shown in Fig. 11 to find out the shear stress distribution in the shear dx2 E 1 t 1 E2 t 2 E2 t 2
layer. The equilibrium relations of an infinitesimal element in the The general solution of differential Eq. (12) can be expressed as:
upper and lower non-shear layers in the x-direction are as follows:
PE1 t 1
9 T 1 ¼ C 1 sinh λx þ C 2 cosh λx þ ð13Þ
 dT 1
þτ ¼ 0 = E1 t 1 þ E2 t 2
dx
ð4Þ rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
 
 dT 2
τ ¼ 0 ;
dx where λ ¼ k E11t1 þ E21t2 , “sinh”, “cosh” and the following “coth”
where T 1 and T 2 are the shear force of the infinitesimal element in in Eq. (20) are the symbols of hyperbolic functions in mathematics
the upper and lower non-shear layers, E1 and E2 , t1 and t2 are Young's [56]. The integration constants C 1 and C 2 can be obtained from the
moduli parallel to the x-direction and thickness of the upper and lower boundary conditions:
non-shear layers, respectively, P is the shear force of per unit width, h
l l
and l are the thickness and length of the shear layer, respectively, and x¼ ; T 1 ¼ P; x ¼  ; T1 ¼ 0 ð14Þ
2 2
G is the shear modulus in the xy plane of the shear layer.
The force equation in an arbitrary cross section perpendicular The integration constants C 1 and C 2 are as follows:
to the shear direction can be easily found P PðE2 t 2 E1 t 1 Þ
C1 ¼ ; C2 ¼ ð15Þ
T1 þT2 ¼ P ð5Þ 2 sinh λ2l 2 cosh λ2lðE1 t 1 þ E2 t 2 Þ

Based on the assumption (3), the shear stress can be given by: Then the forces T 1 , T 2 and τ can be calculated as follows:
!
τ ¼ kðu2  u1 Þ ð6Þ P sinh λx E2 t 2  E1 t 1 cosh λx 2E1 t 1
T1 ¼ þ þ ð16Þ
2 sinh λ2l E1 t 1 þ E2 t 2 cosh λ2l E1 t 1 þ E2 t 2
where k is a scale factor, u1 and u2 are the displacement of upper
and lower non-shear layers, respectively. According to the con- !
stitutive law and geometric deformation of the shear layer, the P sinh λx E2 t 2  E1 t 1 cosh λx 2E2 t 2
T2 ¼   þ ð17Þ
shear stress can be written in another form: 2 sinh λ2l E1 t 1 þ E2 t 2 cosh λ2l E1 t 1 þ E2 t 2
ðu2 u1 Þ !
τ ¼ Gγ ¼ G ð7Þ λP cosh λx E2 t 2  E1 t 1 sinh λx
h τ¼ þ ð18Þ
2 sinh λ2l E1 t 1 þ E2 t 2 cosh λ2l
where γ is the shear strain of the shear layer. Hence, k can be
derived from Eqs. (6) and (7): In the direct shear tests, E1 ¼ E2 ¼ E, t 1 ¼ t 2 ¼ t, then Eq. (18)
G can be simplified as follows:
k¼ ð8Þ
h λP cosh λx
τ¼ ð19Þ
Assume that ε1 and ε2 are the strain of the upper and lower 2 sinh λ2l
qffiffiffiffi qffiffiffiffiffiffi
non-shear layers in the x-direction. From Hooke's law, ε1 and ε2 where λ ¼ 2k Et ¼
2G
hEt
is the simplified expression.
can be obtained as: Consider one condition for the transversely isotropic shale,
T1 du1 T2 du2 according to the corresponding test results (for β ¼301, G E7 GPa,
ε1 ¼  ¼ ; ε2 ¼  ¼ ð9Þ
hE1–4 mm, E E22 GPa, tE48–49.5 mm and l¼50 cm). From Eq.
E1 t 1 dx E2 t 2 dx
(19), the distribution of the normalized shear stress in the shear
Following equation is derived from the differentiation of Eq. (4): direction with different thicknesses of the shear layer is plotted in
d T 1 dτ
2 Fig. 12.
¼ ð10Þ As can be seen from Fig. 12, the shear stress at the specimen end is
dx2 dx
significantly greater than that in the middle of the specimen. The
The substitution of Eq. (6) into Eq. (10) gives: shear stress is concentrated at the end of the specimen in the shear
2   layer. The relative magnitude of the shear stress in the middle of
d T1 du2 du1
¼ k  ð11Þ specimen compared to the area near the end is more sensitive to the
dx2 dx dx
thickness of the shear layer. The normalized shear stress in the middle
of specimens increases from about 10% to 50%, as the thickness of
the shear layer increases from 1.0 mm to 4.0 mm. In addition, the
increasing rate gradually reduces as the increase of the thickness of
the shear layer. For the tested shale specimens, the thickness of the
shear layer is about from 2 mm to 4 mm, that is, the normalized shear
stress in the middle of specimens is about from 30% to 50% of that in
the area near the end. The shear stress concentration is lower when
the thickness of the shear layer is larger.
The shear stress at the specimen end usually reaches the
critical failure strength firstly, and then shear failure occurs at
the end of the specimen when direct shearing, due to the shear
stress concentration at the specimen end in the shear layer. Once
shear failure occurs at the end of the specimen, the shear cracks
propagate from the end to the middle in the shear layer. When the
shear cracks intersect with a bedding plane, shear sliding and
Fig. 11. Shear layer and its mechanical analysis in x-direction for an arbitrary β. tensile splitting along the bedding plane are easier to occur, owing
66 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68

Fig. 6, is interpreted by the anisotropy of the shear stress con-

Fig. 12. Normalized shear stress in the shear direction with different thicknesses of
the shear layer.
to the lower shear strength and lower tensile strength of bedding
planes. The shear fractures branching and re-orientating at the
bedding planes can induce complicated fracture propagation
patens, which is the main reason for the complex failure mechan-
isms of specimens with β ¼301 and 601. The branching and re-
orientating shear fractures at bedding planes are also the reason
for the larger thickness of the shear layer of specimens with β ¼ 01,
301 and 601. The complicated failure modes induced by the
complex crack propagating patens can be observed in Figs. 8–10.
Consequently, the high shear stress concentration at the specimen
end is closely related to the shear strength.
As can be seen from Fig. 12, the maximum shear stress is in the
position of x ¼ 7 2l , and its magnitude is
centration factor.
The variation of the shear stress concentration factor K t versus
λ is presented in Fig. 13. As can be noted, Kt is positively correlated
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
with λ. Considering that λ ¼ 2G=ðhEtÞ, and in direct shear tests of
directional shale specimens, G is a constant, t is approximately the
same for different β, hence, the shear stress concentration factor
K t is only related to Young's modulus in the shear direction and
the
pffiffiffiffiffiffithickness of shear layer, that is, Kt is inversely correlated with
Eh.
The variation of Young's modulus of shale in the shear direction
with α is plotted in Fig. 14 according to the uniaxial compression
pffiffiffiffiffiffi of β. From Fig. 15 and the inverse
test results for different values
correlation between Kt and Eh, it can be noted that K t is probably
the maximum for β ¼901 as h is much smaller than for the other
β's in spite of Young's modulus being a maximum. Therefore, the
shear strength is minimum for β ¼901 at the same normal stress.
For β ¼ 01, 301 and 601, the difference of h is small compared with
the height of the specimens before and after the tests, whereas E
for β ¼01 is smaller than that of β ¼301 and 601, thus K t for β ¼01 is

λP λl
τmax ¼ coth ð20Þ
2 2
The mean shear stress of the shear layer is
Z l
1 2 P
τm ¼ τdx ¼ ð21Þ
l  2l l Fig. 13. Variation of the shear stress concentration factor Kt with λ.

Defining the shear stress concentration factor K t as follows:

26
τmax λl λl
Kt ¼ ¼ coth ð22Þ
τm 2 2 24

When direct shearing, the yield stage can be ignored for hard 22
brittle shale as shown in Fig. 4. Consequently, as the shear load
E(GPa)

increases, shear cracks occur first in the regions of stress concen- 20

tration (both ends of the shear layers), when the shear stress
18
reaches the shear strength. Then the specimen fails as the cracks
expand further. That is, the larger the shear stress concentration 16
factor, the easier shear failure occurs in the shear stress concen-
tration region. Therefore, to some extent, the shear stress con- 14
centration factor can reflect the shear strength of rock when
0 15 30 45 60 75 90
shearing, and it can be used to analyze the anisotropy of shear
α (°)
strength of rock.
As the shear strength is concentrated at the end of the speci- Fig. 14. Young's modulus E versus. α in uniaxial compression tests.
men in the shear layer, the shear failure will firstly occur at the
specimen end with the increase of the shear force at a certain
normal force. The specimen with a higher shear stress concentra-
tion factor usually has a larger shear stress at the specimen end
with the same shear force. Thus, the higher degree of shear stress
concentration at the specimen end in the shear direction is the
main reason for the lower shear strength, as cracks, fractures or
other defects are easily occurs in the region of the higher shear
stress concentration, which induces lower shear strength. The real
failure strength of a material, especially for rock, concrete and
other brittle materials is always lower than the theoretical value β = 0° β = 30° β = 60° β = 90°
because the concentrated stress. The anisotropy of shear strength Fig. 15. Schematic diagram of the thickness of the shear layers for different β in direct
of shale specimens with different bedding orientations, shown in shear tests (the thickness of the shear layer is the distance between the red lines).
 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68
higher than that of β ¼301 and 601. Consequently, shear strength is
lower for β ¼ 01 at the same normal stress. In the four tested
directional specimens, for β ¼601, shear strength is the maximum
because K t is smallest.
But for β ¼301, shear strength is less than that for β ¼01 when
normal stress is smaller than 40 MPa, which is counter to the
theoretical analysis results. The reason may be that failure tends to
take place along bedding planes and is dominated by splitting of
bedding planes. And the enhanced effect of confining stresses
results in a shear strength that is greater than that for β ¼01 when
the normal stress is greater than 40 MPa.
5. Conclusions
Based on the experimental and theoretical analysis presented,
some probably useful conclusions can be drawn as follows:
(1) The anisotropic effect of bedding planes on shear strength is
easily observed. This effect decreases for the same β when the
normal stress increases because of the enhanced confining
stress. This weakening effect increases with increase of angle
β, for the orientation up to 901, due to the sliding failure or
tensile splitting of the specimens along bedding planes.
(2) The shear strength, cohesion and internal friction angle of
bedding planes are the minimum for shale. But the maximum
values of these mechanical parameters are not obtained for
β ¼01 but for β ¼601. This may be due to the complicated
fracture patterns and composed of mixed-mode fractures
along the weak planes and the matrix in shearing, even the
phenomenon of the splitting of bedding planes throughout the
entire specimen can be observed in some faiured specimens.
(3) Weak cohesion is pronounced for bedding planes of shale
compared with that of the matrix, and the internal frictional
angle of bedding planes is also smaller than that of the matrix.
This further demonstrates that the weak planes in this
inherently anisotropic rock are weak not only on the cement-
ing strength but also on the friction.
(4) The shear stress–shear displacement curves display the char-
acteristic of slip-weakening of shear strength for all the tested
specimens (β ¼ 01, 301, 601 and 901). But a distinct trend is
shown for the bedding planes especially when the normal
stress is higher. This also implies that the bedding planes of
shale are weak both on the cohesion and the friction.
(5) For specimens with β ¼01, shear sliding across the weak planes
is dominant. For specimens with β ¼ 301 and 601, shear
strength is controlled both by the matrix and by the weak
planes, and the failure mode is a combination of shear sliding
across the weak planes accompanied by tensile splitting along
the weak planes. For β ¼901, shearing occurs with sliding only
along the weak planes.
(6) An expression of the shear stress concentration factor is
derived. It can be used to analyze the anisotropy of shear
strength of shale. The shear stress concentration factor is
related only to the elastic modulus in the shear direction and
the thickness of the shear layers. For shale specimens with
β ¼01, 301, 601 and 901, the shear stress concentration factor is
the maximum for β ¼901, and the minimum for β ¼601. Hence,
for the shear strength, it is the minimum for β ¼901 and the
maximum for β ¼601 under the same normal stress.
(7) The weak cohesion of bedding planes may lead to serious bore-
hole instability when drilling horizontal wells in shale formations.
In hydraulic fracturing, fractures may prefer to extend along the
bedding planes, which will restrain the formation of fracture
networks when hydraulic fractures initiating from the bedding
planes, and thus reduce the efficiency of hydraulic fracturing.
67
Although this study has yielded some facts on the response of
shale specimens subjected to direct shear loading, the tests are
just a special case where the shear direction is perpendicular to
the intersection lines between weak planes and specimen ends. As
for an arbitrary angle, more researches are needed to throw light
on the effect of the orientations of weak planes on the shear
strength of inherently anisotropic rocks.
Acknowledgment
The authors acknowledge the financial support from National
Basic Research Program of Science Foundation of China (973 Program)
(No.2010CB226701) and National Natural Science Foundation of China
(No. 51104144).
References
[1] Al-Harthi AA. Effect of planar structures on the anisotropy of Ranyeh
sandstone, Saudi Arabia. Eng Geol 1998;50:49–57.
[2] Kohl T, Evansi KF, Hopkirk RJ, et al. Coupled hydraulic, thermal and
mechanical considerations for the simulation of hot dry rock reservoirs.
Geothermics 1995;24(3):345–59.
[3] Hudson JA, Stephansson O, Andersson J. Guidance on numerical modelling of
thermo-hydro-mechanical coupled processes for performance assessment of
radioactive waste repositories. Int J Rock Mech Min Sci 2005;42(5):850–70.
[4] Kim H, Cho JW, Song I, et al. Anisotropy of elastic moduli, P-wave velocities,
and thermal conductivities of Asan Gneiss, Boryeong Shale, and Yeoncheon
Schist in Korea. Eng Geol 2012;147:68–77.
[5] Hakala M, Kuula H, Hudson JA. Estimating the transversely isotropic elastic
intact rock properties for in situ stress measurement data reduction: a case
study of the Olkiluoto mica gneiss Finland. Int J Rock Mech Min Sci 2007;44(1):
14–46.
[6] Min KB, Lee CI, Choi HM.. An experimental and numerical study of the in-situ
stress measurement on transversely isotropic rock by overcoring method. In:
Proceedings of the third international symposium on rock stress. Kumamoto,
Japan; 2003. p. 89–95.
[7] Everitt RA, Lajtai EZ. The influence of rock fabric on excavation damage in the
Lac du Bonnett granite. Int J Rock Mech Min Sci 2004;41(8):1277–303.
[8] Song I, Suh M, Woo YK, et al. Determination of the elastic modulus set of
foliated rocks from ultrasonic velocity measurements. Eng Geol 2004;72(3):
293–308.
[9] Gong QM, Zhao J, Jiao YY. Numerical modeling of the effects of joint
orientation on rock fragmentation by TBM cutters. Tunn Undergr Space Tech
2005;20(2):183–91.
[10] Sanio HP. Prediction of the performance of disc cutters in anisotropic rock. Int
J Rock Mech Min Sci 1985;22(3):153–61.
[11] Ong SH, Roegiers JC. Influence of anisotropies in borehole stability. Int J Rock
Mech Min Sci 1993;30(7):1069–75.
[12] Zhang J. Borehole stability analysis accounting for anisotropies in drilling to
weak bedding planes. Int J Rock Mech Min Sci 2013;60:160–70.
[13] Ong SH. Borehole stability. Norman, Oklahoma: University of Oklahoma; 1994
(Ph.D. thesis).
[14] Ong SH, Roegiers JC. Fracture initiation from inclined wellbores in anisotropic
formations. J Pet Tech 1996;48(7):612–9.
[15] Lee H, Ong SH, Azeemuddin M, et al. A wellbore stability model for formations
with anisotropic rock strengths. J Pet Sci Eng 2012;96:109–19.
[16] Tien YM, Kuo MC. A failure criterion for transversely isotropic rocks. Int J Rock
Mech Min Sci 2001;38(3):399–412.
[17] Matsukura Y, Hashizume K, Oguchi CT. Effect of microstructure and weath-
ering on the strength anisotropy of porous rhyolite. Eng Geol 2002;63(1):
39–47.
[18] Ajalloeian R, Lashkaripour GR. Strength anisotropies in mudrocks. Bull Eng
Geol Environ 2000;59(3):195–9.
[19] Ramamurthy T. Strength and modulus response of anisotropic rocks. Com-
prehensive rock engineering, 1. Oxford: Pergamon; 1993. p. 313–29 (In:).
[20] Amadei B. Importance of anisotropy when estimating and measuring in situ
stresses in rock. Int J Rock Mech Min Sci 1996;33(3):293–325.
[21] Chenevert ME, Gatlin C. Mechanical anisotropies of laminated sedimentary
rocks. Soc Pet Eng J 1965;5:67–77.
[22] McLamore R, Gray KE. The mechanical behavior of anisotropic sedimentary
rocks. J Eng Ind 1967;89(1):62–73.
[23] Attewell PB, Sandford MR. Intrinsic shear strength of a brittle, anisotropic
rock. I: experimental and mechanical interpretation. Int J Rock Mech Min Sci
1974;11(11):423–30.
[24] McCabe MW, Koerner RM. High pressure shear strength investigation of an
anisotropic mica schist rock. Int J Rock Mech Min Sci 1975;12(8):219–28.
[25] Nasseri MH, Rao KS, Ramamurthy T. Failure mechanism in schistose rocks. Int J
Rock Mech Min Sci 1997;34(3):219.e1–15.
68 S. Heng et al. / International Journal of Rock Mechanics & Mining Sciences 74 (2015) 58–68
[26] Horino FG, Ellickson ML. A method for estimating strength of rock containing
planes of weakness [M]. US Dept. of Interior, Bureau of Mines, 1970.
[27] Ramamurthy T, Rao GV, Singh J. A strength criterion for anisotropic rocks. In:
Proceedings of the fifth Australia–New Zealand conference on geomechanics.
(Australia): Institution of Engineers; 1988. p. 253.
[28] Tien YM, Tsao PF. Preparation and mechanical properties of artificial trans-
versely isotropic rock. Int J Rock Mech Min Sci 2000;37:1001–12.
[29] Nasseri MHB, Rao KS, Ramamurthy T. Anisotropic strength and deformational
behavior of Himalayan schists. Int J Rock Mech Min Sci 2003;40(1):3–23.
[30] Cho JW, Kim H, Jeon S, et al. Deformation and strength anisotropy of Asan
gneiss, Boryeong shale, and Yeoncheon schist. Int J Rock Mech Min Sci
2012;50:158–69.
[31] Niandou H, Shao JF, Henry JP, et al. Laboratory investigation of the mechanical
behaviour of tournemire shale. Int J Rock Mech Min Sci 1997;34(1):3–16.
[32] Tien YM, Kuo MC, Juang CH. An experimental investigation of the failure
mechanism of simulated transversely isotropic rocks. Int J Rock Mech Min Sci
2006;43(8):1163–81.
[33] Gatelier N, Pellet F, Loret B. Mechanical damage of an anisotropic porous rock
in cyclic triaxial tests. Int J Rock Mech Min Sci 2002;39(3):335–54.
[34] Li D, Wong LNY, Liu G, et al. Influence of water content and anisotropy on the
strength and deformability of low porosity meta-sedimentary rocks under
triaxial compression. Eng Geol 2012;126:46–66.
[35] Ghazvinian A, Hadei MR. Effect of discontinuity orientation and confinement
on the strength of jointed anisotropic rocks. Int J Rock Mech Min Sci
2012;55:117–24.
[36] Duveau G, Shao JF. A modified single plane of weakness theory for the failure
of highly stratified rocks. Int J Rock Mech Min Sci 1998;35(6):807–13.
[37] Suarez-Rivera R, Green SJ, McLennan J., et al.. Effect of layered heterogeneity
on fracture initiation in tight gas shales. Paper SPE-103327. 2006.
[38] Ghazvinian A, Vaneghi RG, Hadei MR, et al. Shear behavior of inherently
anisotropic rocks. Int J Rock Mech Min Sci 2013;61:96–110.
[39] He MC, Nie W, Zhao ZY, et al. Experimental investigation of bedding plane
orientation on the rockburst behavior of sandstone. Rock Mech Rock Eng
2012;45(3):311–26.
[40] Huang X, Xiong J. Numerical simulation of gas leakage in bedded salt rock
storage cavern. Procedia Eng 2011;12:254–9.
[41] Torsaeter M, Vullum PE, Nes OM.. Nanostructure vs. macroscopic properties of
Mancos Shale. In: Proceeding of the SPE Canadian unconventional resources
conference, paper SPE-162737. 2012.
[42] Harris NB, Miskimins JL, Mnich CA. Mechanical anisotropy in the Woodford
Shale, Permian Basin: Origin, magnitude, and scale. Lead Edge 2011;30(3):
284–91.
[43] Curtis ME, Ambrose RJ, Sondergeld CH.. Structural characterization of gas
shales on the micro-and nano-scales. In: Proceedings of the Canadian
unconventional resources and international petroleum conference, paper
SPE-137693. 2010.
[44] Cipolla CL, Warpinski NR, Mayerhofer MJ., et al.. The relationship between
fracture complexity reservoir properties and fracture treatment design. In:
Proceedings of the SPE annual technical conference and exhibition, paper SPE-
115769. 2008.
[45] King GE.. Thirty years of gas shale fracturing: what have we learned? In:
Proceedings of the SPE annual technical conference and exhibition, paper SPE-
133456. 2010.
[46] Britt LK, Schoeffler J.. The geomechanics of a shale play: what makes a shale
prospective. In: Proceedings of the SPE eastern regional meeting, paper SPE-
125525. 2009.
[47] Rickman R, Mullen M, Petre JE., et al.. A practical use of shale petrophysics for
stimulation design optimization: All shale plays are not clones of the Barnett
Shale. In: Proceedings of the SPE annual technical conference and exhibition,
paper SPE-115258. 2008.
[48] Heng S, Yang CH, Zeng YJ, et al. Experimental study on hydraulic fracture
geometry of shale. Chin J Geot Eng 2014;36(7):1243–51.
[49] Popp T, Salzer K, Minkley W. Influence of bedding planes to EDZ-evolution
and the coupled HM properties of Opalinus clay. Phys Chem Earth 2008;33:
S374–S387.
[50] Annual book of standards. Soil and rock (I), 04.08. West Conshohocken, Penn:
ASTM International; 2011.
[51] Kim BS, Shibuya S, Park SW, et al. Suction stress and its application on
unsaturated direct shear test under constant volume condition. Eng Geol
2013;155:10–8.
[52] Jia CG, Chen JH, Guo YT, et al. Research on the mechanical behaviors and
failure modes of layer shale. Rock Soil Mech 2013;34(2):57–61.
[53] Tavallali A, Vervoort A. Effect of layer orientation on the failure of layered
sandstone under Brazilian test conditions. Int J Rock Mech Min Sci 2010;47(2):
313–22.
[54] Vervoort A, Min KB, Konietzky H, et al. Failure of transversely isotropic rock
under Brazilian test conditions. Int J Rock Mech Min Sci 2014;70:343–52.
[55] Jaeger JC, Cook NGW, Zimmerman RW. Fundamentals of rock mechanics.
4th ed.. New York: Wiley; 2009.
[56] Bradley RE, D'Antonio LA, Sandifer CE. Euler at 300: an appreciation.
Mathematical Association of America 2007 (100 pp.).