Int J Fract (2011) 168:227250
DOI 10.1007/s10704-010-9576-4
ORIGINAL PAPER
Strength failure and crack coalescence behavior of brittle
sandstone samples containing a single fissure under uniaxial
compression
Sheng-Qi Yang Hong-Wen Jing
Received: 17 April 2010 / Accepted: 15 November 2010 / Published online: 2 December 2010
Springer Science+Business Media B.V. 2010
Abstract Uniaxial compression experiments were
performed for brittle sandstone samples containing a
single fissure by a rock mechanics servo-controlled
testing system. Based on the experimental results of
axial stress-axial strain curves, the influence of sin-
gle fissure geometry on the strength and deformation
behavior of sandstone samples is analyzed in detail.
Compared with the intact sandstone sample, the sand-
stone samples containing a single fissure show the
localization deformation failure. The uniaxial compres-
sive strength, Youngs modulus and peak axial strain of
sandstone samples with pre-existing single fissure are
all lower than that of intact sandstone sample, which is
closely related to the fissure length and fissure angle.
The crack coalescence was observed and character-
ized from tips of pre-existing single fissure in brit-
tle sandstone sample. Nine different crack types are
identified based on their geometry and crack propaga-
tion mechanism (tensile, shear, lateral crack, far-field
crack and surface spalling) for single fissure, which
can be used to analyze the failure mode and cracking
process of sandstone sample containing a single fis-
sure under uniaxial compression. To confirm the sub-
sequence of crack coalescence in sandstone sample, the
photographic monitoring and acoustic emission (AE)
S.-Q. Yang (B) H.-W. Jing
State Key Laboratory for Geomechanics and Deep
Underground Engineering, School of Mechanics and Civil
Engineering, China University of Mining and Technology,
221008 Xuzhou, Peoples Republic of China
e-mail: yangsqi@hotmail.com
technique were adopted for uniaxial compression test.
The real-time crack coalescence process of sandstone
containing a single fissure was recorded during the
whole loading. In the end, the influence of the crack
coalescence on the strength and deformation failure
behavior of brittle sandstone sample containing a sin-
gle fissure is analyzed under uniaxial compression. The
present research is helpful to understand the failure
behavior and fracture mechanism of engineering rock
mass (such as slope instability and underground rock
burst).
Keywords Brittle sandstone Fissure length
Fissure angle Mechanical parameter Failure mode
Crack coalescence Tensile crack Surface spalling
Acoustic Emission (AE)
1 Introduction
In past several decades, a number of experimental
and numerical investigations (Bobet 2000; Bobet and
Einstein 1998; Wang and Shrive 1995; Wong et al.
2002; Tang and Kou 1998; Tang et al. 2001; Huang and
Wong 2007; Yang et al. 2008) have been conducted for
fractured rock under compression since Nemat-Nasser
and Horii (1982) investigated the fracturing mecha-
nisms of flaw plates (model material) and obtained
that flaw length is one of the parameters controlling
the failure mode of sample. Bobet (2000) and Bobet
and Einstein (1998) have investigated the 2-D cracks
123
228
propagation and coalescence on rock-like materials
containing two inclined open or closed flaws. Wang
and Shrive (1995) reviewed and summarized the exist-
ing typical models and fracture criteria for brittle frac-
ture in compression, and established a sound fracture
model from knowledge of observed fracture behavior
and based on both a stress and an energy criterion.
Wong et al. (2002), Tang and Kou (1998), Tang et al.
(2001) performed uniaxial and biaxial compression on
numerical model samples containing a number of large,
pre-existing flaws and a row of suitably oriented smaller
flaws by RFPA2D software, which shows that the crack
growth is stable and stops at some finite crack length
with the increase of lateral pressure; while a lateral
tensile stress with even a small value will result in an
unstable crack growth after a critical crack length is
attained. Huang and Wong (2007) performed a series of
uniaxial compressive tests on the frozen PMMA (Poly-
methyl Methacrylate) with pre-existing 3-D cracks.
Based on the test results, it was found that for 3-D sur-
face cracks, the interactions of cracks affected crack
growth and extension in two aspects, i.e. the inter-
action either prompting crack extension or restrain-
ing it. The mechanism of crack interaction depends
mainly on the location of cracks in spatial domain.
Yang et al. (2008) carried out conventional triaxial
compression experiment investigation for cylindrical
marble samples with two pre-existing closed cracks
in non-overlapping geometry by fixing the flaw length
and ligament length, which investigated the strength
and deformation property of marble with four various
flaw geometries under different confining pressures.
The above experimental and numerical results for pre-
cracked rock have distinctly shown that the geometries
of fissures in the rocks have an important influence on
the strength and failure behaviors. However, previous
investigations focused mainly on physical test of model
materials (rock-like materials) due to easy fabrication
of fissures or numerical simulation for brittle rock mate-
rials. Less experimental studies were performed for real
rock material to investigate the initiation, propagation
and coalescence of pre-cracked rock materials. Some
represented experimental results for rock material with
pre-existing fissures can be summarized as follows.
Wong and Chau (1998) observed three main modes
of crack coalescence (Tensile crack, shear crack and
mixed cracks) in flawed samples with two fissures in
uniaxial compression. Wong et al. (2001) investigated
experimentally crack coalescence and peak strength
123
S.-Q. Yang, H.-W. Jing
of model materials containing three parallel fractional
flaws, which showed that the mechanisms of cracks
coalescence depended on the flaw arrangement and the
frictional coefficient on the flaw surface. Prudencio and
Van Sint Jan (2007) presented the results of biaxial tests
on physical models of rock with non-persistent joints.
Tests showed three basic failure modes: failure through
a planar surface, stepped failure, and failure by rotation
of new blocks. Planar failure and stepped failure were
associated with high strength, and small failure strains,
whereas rotational failure was associated with a very
low strength, ductile behavior, and large deformation.
Besides the above physical tests on rock-like materials,
some real rock samples have also been tested to investi-
gate the mechanical behavior of pre-cracked rock mate-
rials. The real rock samples were usually machined into
thin slices with some slots. For example, Feng et al.
(2009) carried out a number of experimental studies to
investigate the mechanism of multi-crack interaction in
limestone samples (the size is about 30 15 3 mm)
with two or three flaws of different arrangements under
the coupled uniaxial compressive stress and chemi-
cal solutions with different ionic concentrations and
pH values. The experimental results indicated that the
influence of chemical corrosion was quite complicated,
depending on chemical ions and their concentrations
and pH values, mineral components of rock, geometry
and the number of flaws. Li et al. (2005) carried out
the uniaxial compression experiment on pre-cracked
marble (the size is about 110 62 25 mm), to inves-
tigate the propagation, crack coalescence of pre-exist-
ing fissures in marble samples. Two types of newborn
cracks are observed: wing (tensile) cracks and second-
ary (shear) cracks. Both types of cracks initiate from
the tips of the flaws and propagate in a stable manner.
Wong and Einstein (2009) investigated the cracking
behaviors in modeled gypsum and Carrara Marble sam-
ples (152 76 32 mm) containing single open fissure
(12.5 mm long and 1.3 mm width) under uniaxial com-
pression. Seven different crack types were identified
based on their geometry and propagation mechanism
(tensile/shear) by a high-speed video observation.
Sandstone is a kind of typical sedimentary rock,
which is formed by the accumulation of thousands
of years after sand grains deposited in the riverbed
through water erosion. The microscopic behavior of
sandstone material is distinctly different from that of
marble material, which leads to the strength failure and
crack coalescence in sandstone samples are distinctly
Strength failure and crack coalescence behavior
different from that in marble sample. Of course, sand-
stone can be divided into less porous sandstone and
porous sandstone, which is usually a function of poros-
ity. Therefore, porosity has a key effect on strength fail-
ure and crack coalescence behavior. Fujii and Ishijima
(2004) performed the uniaxial compression tests on
sandstone samples with an inclined slit from the spec-
imen surface or an inclined initial fracture from the
specimen surface. The size of sandstone sample is
120 60 60 mm, and an inclined slit (open flaw)
of 1 mm in thickness and 20 mm in length was intro-
duced. The experimental results showed the crack from
the inclined slit from the sample surface grew at a small
angle to the initial direction and curved slightly towards
the free surface in all cases for sandstone samples.
However no clear relation was observed between the
load at which the main fracture initiated and the incli-
nation of the slit. Lu et al. (2008) carried out the exper-
imental study on cylindrical red sandstone (the size is
50 mm in diameter and 100 mm in length) containing
one or two pre-existing fissures under chemical action,
which investigated the influence of water chemistry on
the mechanical behavior of pre-cracked sandstone. The
research result provided understanding of the funda-
mental nature of rock failure and damage mechanics
models for simulating nonlinear evolution of tested red
sandstone under coupled mechanical-hydraulic-chem-
ical processes.
As it is well known, Acoustic Emission (AE) tech-
nique is an efficient measuring tool to reflect the evo-
lution and propagation of defects in materials, which
has been widely applied to investigate the damage and
fracture behavior internal the rock material in many
past studies (Li et al. 2005; Labuz et al. 2001; Chang
and Lee 2004; Moradian et al. 2010; Pestman and Van
Munster 1996; Debecker and Vervoort 2009; Ganne
and Vervoort 2007; Eberhardt et al. 1999; Tham et al.
2005; Jouniaux et al. 2001). Debecker and Vervoort
(2009) carried out a series of Brazilian tensile tests and
uniaxial compression tests for layered slate by visual
recordings and AE technique, which investigated in
detail the fracture sequence and growth direction in
anisotropic slate samples. The obtained experimen-
tal results provided a deep insight in the different
parameters that influenced the strength and deforma-
tion behavior of slate. Ganne and Vervoort (2007) stud-
ied the effect of stress path on pre-peak damage in rock
material induced by macro-compressive andtensile
stress fields. The results show that previously induced
229
(by compressive stresses) intragranular cracks influ-
ence the occurrence and growth of the intergranu-
lar cracks caused by tensile stresses, while previously
induced (by tensile stresses) intergranular cracks influ-
ence the occurrence and growth of the intragranular
cracks caused by compressive stresses. Eberhardt et al.
(1999) analyzed the influence of grain size on the ini-
tiation and propagation thresholds of stress-induced
brittle fractures by AE monitoring. Tham et al. (2005)
characterized the micro- and macro-failure behaviors
of granite and marble plate samples under uniaxial
tension by using an AE technique. The results reveal
clearly the failure processes of rock including initi-
ation, nucleation and propagation of micro-fractures
when the axial stress is close to the peak strength of
rock. But up to now, AE technique was less used to
explore the initiation, propagation and coalescence of
brittle rock containing fissures (Jouniaux et al. 2001;
Wassermann et al. 2009; Van De Steen et al. 2005; Lei
et al. 2004). Van De Steen et al. (2005) investigated the
effect of a tensile stress gradient on fracture initiation
and growth in Brazilian disc rock sample containing a
hole by AE technique, which showed that the maximum
local tensile stresses the material was able to withstand
increase as the stress gradient increases. The Brazil-
ian disc with a hole split along the loaded diameter or
macro-fracturing starts at the hole (Fig. 1). Lei et al.
(2004) obtained the detailed time-space distribution of
AE events during the catastrophic fracture of rock sam-
ples containing a pre-existing joint or potential fracture
plane under triaxial compression using a high-speed
32-channel waveform recording system (Fig. 2).
Therefore in this research, in order to better under-
stand the strength and fracture coalescence phenomena
in brittle rock material, uniaxial compression exper-
iments were conducted for brittle sandstone samples
containing a single fissure by a rock mechanics servo-
controlled testing system. Moreover, this investigation
relied on photographic monitoring and AE technique
to obtain the real-time crack coalescence process in the
process of the whole deformation failure, which was not
preformed for brittle sandstone in previous studies. The
emphasis of this research is focused on the investigation
the influence of the geometry of single fissures (fissure
length and fissure angle) on the strength and defor-
mation failure behavior of brittle sandstone material,
to analyze real-time crack coalescence process in real
sandstone material under uniaxial compression on basis
of photographic monitoring and AE technique, and to
123
230 S.-Q. Yang, H.-W. Jing

1200 Q Q Quartz; F Feldspar; K Kaolinite;

F
Ch Chlorite; I Illite; C Calcite;
1000
S Smectite; L Siderite;

Intensity (cps)
I/S Illite/Smectite interstratified minerals;
800
O Others.
600

400
K Q
Ch F FCh Q
F Ch
Q
Q
200 S F Ch F C Q KQ
I/S I FO Q Q
I C Q K F L Ch K Ch
0
3 8 13 18 23 28 33 38 43 48 53 58 63

2 ()

Fig. 3 XRD analysis of sandstone material in this research

sandstone has a crystalline and blocky structure, which

Fig. 1 Two typical failure modes in Brazilian disc rock samples
containing a hole Van De Steen et al. (2005)
is a fine-grained heterogeneous material with average
unit weight about 2620 kg/m3 . Moreover tested sand-
stone is porous. According to the result of X-ray dif-
fraction (XRD) (Fig. 3), the minerals in the sandstone
samples are mainly feldspar, quartz, kaolinite and chlo-
rite, which more detailed composition of this rock are
described as follows: the content (weight %) of feldspar
is 51.7%, quartz 17.4%, kaolinite 11.2%, chlorite 9.4%,
illite 2.3%, smectite 2.1%, illite/smectite interstratified
minerals 2.4%, calcite 1.1%, siderite 0.5%, and others
1.9%.

2.2 Preparation for sample with single fissure

In the present research, sandstone samples containing a

Fig. 2 Typical space AE distribution of cylindrical rock sample
containing pre-existing joints Lei et al. (2004)
study the effect of the crack coalescence on the strength
and deformation failure behavior of brittle sandstone
sample.
2 Experimental studies
2.1 Sandstone material
To investigate the strength failure and crack coales-
cence behavior of rock containing a single fissure under
uniaxial compression, the sandstone material located
in Linyi City, Shandong province of China was cho-
sen for the experimental object in this research. The
single fissure were prepared to carry out uniaxial com-
pression tests. The samples were cut from the same
rectangular block. When cutting, the sandstone sam-
ples were machined along the same direction in order
to avoid the effect of anisotropy on the experimental
results of sandstone. In order to obtain the exact results
as well as the best comparison, all the tests were per-
formed in natural and dry conditions.
Following the method suggested by the ISRM
(Fairhurst and Hudson 1999), the height to width ratio
of tested sample should be in the range of 2.03.0 in
order to minimize the effect of the end friction effects on
the testing results. Therefore, the size of all tested sand-
stone samples is rectangular with 120 mm in height,
60 mm in width and 30 mm in thickness. As a result,
all tested samples with the height to width ratio of 2.0
ensure a uniform stress state within the central part of
the samples.

123
Strength failure and crack coalescence behavior 231

The geometry of fissured sandstone is described in
Fig. 4. Notice, the term fissure is used to describe
an artificially created flaw or crack; however the term
crack is adopted to describe the new fracture or fail-
ure in the process of loading. The geometry of single fis-
sure is defined by two geometrical parametersfissure
length 2a and fissure angle (the angle of fissure with
the direction of the horizontal direction) as shown in
Fig. 4.
Although for rock-like materials, it is easier to make
fissures by inserting steel or copper slices during mold-
ing process, it is more difficult to cut fissures in the thick
samples of real rock (Li et al. 2005). High pressure
water-jet cutting machine was used to cut open single
fissure in the intact sandstone samples. Machined open
fissure width was about 2.5 mm. In order to investigate
the effect of pre-existing single fissure geometry on
the strength, deformation and failure behavior of sand-
stone under uniaxial compression, different geometries
of single fissure by varying one of the two parameters
(2a and ) while keeping another constant were chosen
in this study. Detailed description for sandstone sam-
ples with different single fissure geometries is listed in
Table 1.

Fig. 4 Geometry of single

1
fissure in the sandstone 1
samples, in which is
fissure angle, 2a is fissure
length. In this research,
open fissure in sandstone Crack
Upper fissure tip
sample was machined by
high pressure water-jet
cutting
2a
120mm

Fissure Fissure
Under fissure tip

Pre-existing open Open fissure in sandstone

60mm 30mm 1
fissure width was sample was machined by high
about 2.5mm pressure water-jet cutting

Table 1 Pre-existing single fissure geometries of sandstone samples under uniaxial compression
Sample W/ mm H/ mm T/ mm M/g /(kg/m3 ) / 2a/mm Note

GS-03# 60.4 120.3 30.0 572.41 2625.9 N/A 0 Intact sample

GS-04# 60.1 120.4 29.8 559.54 2599.4 45 5 Fissure length
GS-07# 60.2 120.8 29.0 551.09 2622.1 45 10
GS-10# 60.3 120.0 29.3 549.53 2605.4 45 15
GS-14# 60.3 120.5 30.1 567.32 2611.9 45 20
GS-17# 60.1 120.3 29.4 555.72 2637.2 45 25
GS-19# 60.5 120.0 29.9 571.30 2645.5 15 15 Fissure angle
GS-11# 60.4 120.0 28.3 538.22 2637.6 45 15
GS-25# 60.5 120.2 29.9 561.64 2596.4 60 15
GS-28# 60.3 120.2 29.3 552.48 2615.0 75 15
W Width, H height, T thickness, M mass

123
232 S.-Q. Yang, H.-W. Jing

Fig. 5 MTS815.02 rock

MTS815.02
mechanics servo-controlled
testing system and
AE21C-06 Acoustic
Emission System, in which
the camera and lamp were
used to record clearly the
crack coalescence process
of sandstone
Camera Lamp
AE21C-06 Acoustic Sample with AE sensor
Emission System

2.3 Experimental equipment and procedure
Uniaxial compression experiments for intact and fis-
sured sandstone samples were carried out on a rock
mechanics servo-controlled testing system (MTS815.
02) with the maximum loading capacity of 2700 kN
and the maximum displacement capacity of 25 mm, as
shown in Fig. 5. This servo-controlled system can test
samples in load or displacement control while the data
are recorded and analyzed in real-time. The axial stress
was imposed on the surface of rock sample until the
failure took place. All the tests were conducted under
displacement- controlled conditions with a strain rate
of (1.3 2.0) 105 .
In this research, we measured the axial force with
the loading capacity of 600 kN and the axial deforma-
tion with displacement capacity of 5 mm. During the
whole uniaxial compression experiments, loads and
deformations of the tested sandstone samples were
recorded simultaneously. Moreover two rigid steel
blocks (63 mm in width, 32 mm in thickness and 15 mm
in height) were placed between the loading frame and
rock sample. At the same time, two antifriction gas-
kets were placed between two rigid steel blocks and the
end face of rock sample, which decreased distinctly the
effect of the end friction effects on the testing results
of sandstone samples with the height to width ratio
2.0.
During the uniaxial compression test, in order to
investigate the crack coalescence mechanism in the
sandstone samples containing single fissure, the AE
technique were adopted. The AE counts and AE energy
rate were recorded by AE21C-06 Acoustic Emission
System (Fig. 5). The frequency of AE system ranges
from 5 KHz to 1 MHz and the recording event rate is
0.1 s. At the same time, the camera was used to trig-
ger manually to capture images when the coalescence
cracks developed or sample failure occurred. When
measuring AE signals, it is a pity that only one AE
sensor were stuck on the sandstone samples (Fig. 5)
using vaseline as a coupling agent and fixed slightly
by cellulose tape, which is approximately same with
Li et al. (2005). Notice the AE sensor is nearby pre-
existing fissure in the sample.
3 Strength and deformation behavior
Axial stress-axial strain curves for sandstone samples
containing single fissure under uniaxial compression
are shown in Fig. 6, in which the axial stress-axial strain
curve of intact sandstone sample is also plotted. From
Fig. 6, it is very clear that the fissure length and fissure
angle have a key effect on the strength and deformation
behavior of sandstone samples under uniaxial compres-
sion, which will be analyzed in detail as follows.
3.1 Uniaxial stress-strain curves of sandstone with
different single fissure geometries
In accordance with the axial stress-axial strain curve
shown in Fig. 6 and the failure mode plotted in Fig. 7,
we can obtain that the sandstone in present study is a
kind of brittle rock and takes on typically axial splitting
failure, i.e. the axial stress drops abruptly to zero in two
seconds after peak stress and a huge failure sonic can
be heard during the dropping. Notice the intact sample
was fractured into many small blocks under uniaxial
compression. The axial stress-axial strain behavior of
sandstone containing a single fissure can be approx-
imately divided into four typical stages, i.e. fissure
closure, elastic deformation, crack growth and prop-
agation, and strain-softening.
At the stage of fissure closure, the uniaxial com-
pression stressstrain curve of sandstone sample con-
taining different single fissure geometries shows the

123
Strength failure and crack coalescence behavior 233

a 240
45 2a=0mm
(Intact sample)
200
Axial stress 1 / MPa

2a=5 mm
160

2a=20 mm
120 2a=10 mm

80 2a=15 mm

40
2a=25 mm

0
0 2 4 6 8 10 Fig. 7 Typically axial splitting failure of intact sandstone sample
-3 in the present study
Axial strain 1 / 10

b 240
2a 15 mm
Intact sample
But it needs to be noted that even though at this stage
200
of elastic deformation, some initiated cracks from the
Axial stress 1 / MPa

75 tips of fissure in flawed sample have still been possible

160
120
80
40
45
0
0 2 4 6
Axial strain 1 / 10
Fig. 6 Axial stress-axial strain curves of sandstone samples con-
taining a single fissure under uniaxial compression, in which Fig
(a) and (b), respectively, show the effect of fissure length and
fissure angle on axial stress-axial strain curves
downward concave and the initial nonlinear deforma-
tion at low stress levels, which results from the clo-
sure of some primary pores and voids in the sample
with the increasing compression stress. Moreover the
rigidity of the sandstone sample at the initial stage of
deformation (Fig. 6) has better consistency and is not
dependent to fissure length and fissure angle, which
results from better homogeneity of tested rock mate-
rial. At the stage of elastic deformation, the loading
begins with the increase of axial stress after the primary
fissure closure in intact and flawed sandstone sample,
and the elastic deformation dominates the stress-strain
curve of sample. The linear stress-strain behavior of
intact and flawed sandstone sample does not change
although there are some irrecoverable processes such
as the fissure closure or crack initiation at this stage.
to propagate along the direction of axial stress, which
60 is different from intact sample.
However at the stage of crack growth and propa-
15 gation, the sandstone sample continuously produces
the stress concentration nearby the tips of pre-exist-
ing single fissure, which results in that some new
cracks will initiate and propagate from the upper and
8 10 under tips along the direction of axial stress. There-
-3 fore the stress-strain curve departs from the elastic
behavior and shows distinctly nonlinear deformation.
At the stage of strain-softening, the macroscopic crack
in intact and flawed sandstone sample comes out rap-
idly and the post-peak behavior in the stressstrain
curves shows a rapid drop. At this time, the sample
can only support a lower axial stress, even though no
axial stress.
It needs to be noted that the stressstrain curves
of sandstone sample containing single fissure show an
abrupt change of slope before peak strength, which is
coincident with the propagation of some cracks at the
upper and under tips of pre-existing fissures. For the
abrupt change in the stress-strain curve, we will make
a detailed analysis and explanation in the fourth sec-
tion. Moreover, compared with intact sample, samples
containing single fissure take on the localization defor-
mation failure nearby peak strength, which is closely
related to the fissure length and fissure angle. For the
samples containing longer fissures (keeping = 45 )
or larger fissure angles (keeping 2a = 15 mm), the
localization deformation failure nearby peak strength
is more distinct, as shown in Fig. 6.

123
234 S.-Q. Yang, H.-W. Jing

Table 2 Mechanical parameters of sandstone samples containing a single fissure under uniaxial compression
Sample / 2a/mm c /MPa E S /GPa E 50 /GPa 1c /103 Note

GS-03# N/A 0 212.08 35.95 24.37 7.500 Intact sample

GS-04# 45 5 170.04 31.47 22.63 6.768 Fissure length
GS-07# 45 10 136.39 30.74 20.91 5.671
GS-10# 45 15 122.80 28.61 20.54 5.358
GS-14# 45 20 85.29 28.21 17.09 4.308
GS-17# 45 25 77.25 28.56 16.81 3.674
GS-19# 15 15 139.28 32.86 22.95 5.663 Fissure angle
GS-11# 45 15 115.17 28.11 21.13 4.827
GS-25# 60 15 149.96 32.25 21.68 5.881
GS-28# 75 15 181.71 33.97 24.93 6.348

3.2 Effect of single fissure geometry on strength
and deformation parameters of sandstone
In this section, the influence of single fissure geometry
is investigated on the strength and deformation parame-
ters of sandstone samples under uniaxial compression,
which are listed in detail in Table 2. Before analyz-
ing the effect of single fissure geometry on strength
and deformation parameters of sandstone, we firstly
investigate the influence of heterogeneity of sandstone
material on axial stress-axial strain curve of samples
containing a single fissure (2a = 15 mm and = 45 )
under uniaxial compression, which is presented in
Fig. 8. From Fig. 8, the heterogeneity has nearly no
effect on the strength and deformation parameters of
140
2a = 15 mm
120 45
Axial stress 1 / MPa
two samples containing a same single fissure, which
have a good consistency. Therefore, the dispersion of
sandstone sample containing a single fissure in this
research is very small, and the strength and deforma-
tion parameters listed in Table 2 can be used to discuss
approximately the effect of single fissure geometry on
the mechanical parameters of sandstone under uni-
axial compression. But due to limited tested samples
from the same block, we can not repeat the test for
other samples with different fissure lengths and angles.
Moreover, because sandstone is a kind of natural geo-
logical material; complete axial stress-axial strains of
two samples in Fig. 8 have some minor differences,
which results from the effect of the heterogeneity on
the crack coalescence process of rock.
The definition for every mechanical parameter in
Table 2 is described as follows. The c is defined as the
uniaxial compressive strength and the 1c is defined
as peak axial strain value. E S represents the elastic
modulus; but E 50 is defined as deformation modulus.

100 GS-10# More detailed descriptions on Youngs modulus can be

80 found in Yang and Jiang (2010). In accordance with
Table 2, the effect of fissure length and fissure angle
60
on the c , 1c , E S and E 50 of sandstone sample con-
GS-11#
40 taining a single fissure are analyzed and discussed in
detail (Figs. 9, 10). The uniaxial compressive strength,
20
Youngs modulus and peak axial strain of sandstone
0 samples with pre-existing single fissure are all lower
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 than that of intact sandstone sample, while the reduc-
Axial strain 1 / 10 -3
tion extent has a significant relation with the fissure
length and fissure angle.
Fig. 8 Effect of heterogeneity on axial stress-axial strain curve
of sandstone samples containing a single fissure (2a = 15 mm
The uniaxial compressive strength of intact sam-
and = 45 ) under uniaxial compression ple was 212.08 MPa, while it could be seen that the

123
Strength failure and crack coalescence behavior 235

Fig. 9 Effect of fissure a 250 b 10

length on the strength and Uniaxial compressive strength Peak axial strain
deformation parameters of 200 8
sandstone samples with the

c / MPa
6

1c / 10-3
same fissure angle 150
( = 45 ) under uniaxial
compression 100 4

50 2

0 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
2a /mm 2a /mm

c 40 d 30
35 25
30 20
ES / GPa

E50 / GPa
25 15
20 Elastic modulus 10 Deformation modulus
15 5
10 0
0 5 10 15 20 25 30 0 5 10 15 20 25 30
2a /mm 2a /mm

Fig. 10 Effect of fissure a 250 b 10

c =212.08MPa
angle on the strength and -3
1c =7.50010
deformation parameters of 200 8
sandstone samples with the
c / MPa

-3

same fissure length 150 6

1c / 10

(2a = 15 mm) under uniaxial

100 4
compression
Uniaxial compressive strength 2 Peak axial strain
50

0 0
0 15 30 45 60 75 90 0 15 30 45 60 75 90
/ /
c 40 d 30
ES 35.95GPa E50 24.37GPa
35 25

30 20
E50 / GPa
ES / GPa

25 15

20 10 Deformation modulus
Elastic modulus
15 5

10 0
0 15 30 45 60 75 90 0 15 30 45 60 75 90
/ /

UCS of samples containing single fissure ranged from
77.25 MPa ( = 45 and 2a = 25 mm) to 181.71 MPa
( = 75 and 2a = 15 mm), while the reduction
extent of uniaxial compressive strength was between
14.3% and 63.6%. In accordance with Fig. 9a, one can
conclude that the uniaxial compressive strength will
decreases with the increase of fissure length. How-
ever, for the relation between uniaxial compressive
strength and fissure angle, according to Fig. 10a, it can
be seen that the uniaxial compressive strength firstly
decreases and then increases with increasing fissure
angle.

123
236
The intact sandstone sample failed at a strain of
0.75% under uniaxial compression, while sandstone
samples containing a single fissure failed at a strain
of approximately 0.370.68%, which is less than that
of intact sample. Moreover, the peak axial strain 1c
of sandstone sample is dependent distinctly to fissure
length and fissure angle. In accordance with Fig. 9b,
one can conclude that the peak axial strain decreases
with the increase of fissure length. However, for the
relation between peak axial strain and fissure angle, in
accordance with Fig. 10b, it can be seen that the 1c
has a nonlinear relation with the fissure angle.
The elastic modulus and deformation modulus of
intact sample were 35.95 and 24.37 GPa, respectively.
However, the elastic modulus of sample containing a
single fissure ranged from 28.11 GPa ( = 45 and
2a = 15 mm) to 33.97 GPa ( = 75 and 2a = 15 mm),
while the reduction extent of elastic modulus was
between 5.5 and 21.8%. The deformation modulus
of sample containing a single fissure ranged from
16.81 GPa ( = 45 and 2a = 25 mm) to 24.93 GPa
( = 75 and 2a = 15 mm), while the reduction extent
of deformation modulus was between 2.3 and 31%.
On the whole, the Youngs modulus of sandstone has
a decreasing trend with the increase of fissure length,
but a nonlinear relation with the fissure angle. Notice,
for the fissured sample with the same angle ( = 45 )
the elastic modulus of sample tends to a stable value
28 GPa after the fissure length increases to 15 mm.
Among all the samples containing single fissure, the
deformation moduli of the samples GS-14# ( = 45
and 2a = 20 mm) and GS-17# ( = 45 and 2a = 25 mm)
are lowest and approximately 17 GPa, which results
from that the axial stress corresponded to an abrupt
change of slope before peak strength is lower than
half of peak strength. Except for the samples GS-
14# and GS-17# , the average deformation modulus
of other samples containing a single fissure is about
22.37 GPa, which reaches 91.8% of the deformation
modulus 24.37 GPa of intact sample, and the disper-
sion coefficient (the percentage of the differential value
between the maximum and minimum value to the aver-
age value) is approximately 18%.
From the above analysis, one can conclude that, for
the sample with longer fissure or the fissure angle of
45 the corresponding mechanical parameters are all
lower, which shows that the longer fissure and fissure
angle of 45 in the sample have significant effect on
the strength and deformation parameters of sandstone
123
S.-Q. Yang, H.-W. Jing
sample. Moreover the influence of single fissure geom-
etries on the uniaxial compressive strength and peak
axial strain is higher than that on the Youngs modulus
for this kind of brittle sandstone material.
4 Crack coalescence behavior
In the research, the intact sandstone sample leads to
typically axial splitting failure (Fig. 7), which is a
good agreement with many laboratory testing results
of brittle rocks. It needs to be explained that we did
not observe any crack coalescence for intact sample
before peak strength, moreover the intact sample failed
in two seconds after peak stress and a huge failure
sonic was heard during testing. At the same time, the
intact sample was fractured into many small blocks, see
Fig. 7. However, for samples containing a single fissure,
under uniaxial compression, different failure modes
were observed in comparison with that of intact sample.
The crack coalescence was all observed from upper and
under tips of pre-existing fissure in sandstone samples
containing single fissure, and ultimate failure mode is
distinctly dependent to the fissure length and fissure
angle. Therefore, in this section, a systematic evalua-
tion is made on crack coalescence in sandstone samples
by photographic monitoring and AE technique.
4.1 Crack coalescence type of sandstone samples
containing a single fissure
Nine different crack types (Fig. 11) were identified
based on their geometry and crack propagation mecha-
nism (tensile, shear, lateral crack, far-field crack and
surface spalling) by analyzing the ultimate failure
modes of sandstone samples containing a single fis-
sure (Figs. 12, 13). In accordance with Fig. 11, one can
see that five of them (crack types IVI) are tensile, one
of them (crack type VI) is lateral crack, one of them
(crack type VII) is shear, one of them (crack type VIII)
is far-field and one of them (crack type IX) is surface
spalling. All crack types can be categorized as follows:
(1) Crack type I: Wing crack initiates simultaneously
from upper and under tips of single fissure. The
tensile wing crack usually initiates along the verti-
cal direction to pre-existing fissure, and then grad-
ually departs towards the direction of axial stress,
and finally develops along the direction of axial
Strength failure and crack coalescence behavior 237

Fig. 11 Various crack types

from single fissure
T T
identified in the present T T T
study. T Tensile crack,
Lc Lateral crack, S Shear
crack, F Far-field crack,
Ss Surface spalling
T
T T T
T

I: Tensile crack II: Tensile crack IV: Tensile crack

(tensile wing crack) (tensile wing crack) III: Tensile crack (Anti-tensile crack) V: Tensile crack

F F
Ss
Lc S

F
Lc S F Ss
F

VI: Lateral crack VII: Shear crack VIII: Far-field crack IX: Surface spalling

stress. Moreover due to very fine grain sizes in
sandstone material, this coalescence path of ten-
sile wing crack I is very smooth.
(2) Crack type II: Wing crack often initiates simul-
taneously from upper and under tip of single fis-
sure. The wing cracks initiates firstly along the
direction to the fissure, and then gradually turns
towards the direction of axial stress, and finally
develops towards the end face of sample.
(3) Crack type III: Tensile crack initiates from upper
or under tip of single fissure. Usually, the tensile
crack III develops along the direction completely
parallel to the axial stress, but the crack coales-
cence path is not very smooth.
(4) Crack type IV: The crack coalescence is reverse
to that of crack type III, which is referred to as
anti-tensile crack. The anti-tensile crack also
initiates from upper or under tip of single fis-
sure, and develops along the direction of the axial
stress.
(5) Crack type V: The tensile crack usually initiates
from this position which is located at a distance
away from the tips of single fissure. The tensile
crack initiates along the vertical direction to the
fissure, and then gradually departs towards the
direction of axial stress, and finally develops along
the direction of axial stress.
(6) Crack type VI: The crack initiates from upper or
under tip of single fissure. The lateral crack initi-
ates firstly along the direction to the fissure, and
then gradually turns towards the horizontal direc-
tion, and finally develops towards the lateral face
of sample.
(7) Crack type VII: The shear crack initiates from
upper or under tip of single fissure, which the
crack coalescence path is parallel to the direction
of pre-existing fissure. The shear crack and pre-
existing fissure are coplanar.
(8) Crack type VIII: The far-field crack does not usu-
ally initiate from the tips of single fissure. More-
over, the far-field crack coalescence path is not
very smooth, which may be vertical or horizontal
in the process of the loading.
(9) Crack type IX: The surface spalling shown in
Fig. 11 is often observed for brittle rock material.
The surface spalling usually appears after some
tensile cracks (such as crack types I-III) or accom-
panies with the process of tensile crack (such as
crack type IV) propagation.
In accordance with the above nine crack types, one
can analyze the ultimate failure mode and cracking pro-
cess of sandstone samples containing a single fissure
under uniaxial compression (Figs. 12, 13 and Table 3).

123
238 S.-Q. Yang, H.-W. Jing

Fig. 12 Effect of fissure

4 2
length on ultimate failure
modes of brittle sandstone 3 Ss
4
samples containing single Ss
fissure under uniaxial
1
compression ( = 45 )
1

2 1
4 Ss
4
3
Ss
3 3

2a=5mm, 45 2a=10mm, 45

10
4
Ss
3
4 1 2
1
4
4

1 2 6
3 3

4 1
5
6
5

2a=15mm, 45 2a=20mm, 45

3
17 14
2
Ss 6
7 1

7
Ss
4

5 1
Local Local
Magnification Magnification

2a=25mm, 45 2a=25mm, 45 2a=20mm, 45

It is very clear that the macroscopic failure mode of
sandstone samples containing a single fissure (Figs. 12,
13) are all a mixture of several cracks among the above
nine various crack types. For example, the failure mode
of the sample GS-17# (2a = 25 mm, = 45 ) is a mix-
ture of cracks IIII and VIIIX. From Table 3, some
qualitative conclusions can be summarized as follows.
Table 3 summarizes initiated crack types of sand-
stone samples containing single fissure with different
lengths and angles in response to the applied axial
loads. As indicated in Table 3, tensile crack I and III
are often the first crack to initiate from the tips of pre-
existing fissure. But sometimes, the tensile crack V
located at a distance away from the tips of single fis-
sure is also the first crack, and is only observed in the
sample GS-19# (2a = 15 mm, = 15 ) which is pos-
sible to result from the heterogeneity of rock material
or lower fissure angle. For tensile crack IV, it is very

123
Strength failure and crack coalescence behavior 239

Fig. 13 Effect of fissure

angle on ultimate failure Ss
modes of brittle sandstone Ss
samples containing single
fissure under uniaxial
compression (2a = 15 mm)
Ss 5
Ss
Ss
5

1
2 3 2
2 3
1
1 2 2
6 1
6
1 3
3 4
2
5 1 4 3
5 1

15, 2a=15mm 45, 2a=15mm 60, 2a=15mm 75, 2a=15mm

Table 3 Initiated crack types of sandstone samples containing single fissure with different lengths and angles
Sample Crack types
I:Tensile II:Tensile III:Tensile IV:Tensile V:Tensile VI:Lateral VII:Shear VIII:Far-field IX:Surface spalling

GS-03# Intact sample: typically axial splitting failure (Fig. 7)

GS-04# 1 (more)

GS-07# 1 (more)

GS-10# 1 (more)

GS-14# 1

GS-17# 1 (small) (small)

GS-19# 1 (small)

GS-11# 1 (small)

GS-25# 1

GS-28# 1

1 Indicates that the crack is the first crack to initiate from the tips of pre-existing fissure

clear that anti-tensile crack IV often accompanies with
tensile crack III towards the reverse direction, which is
usually observed in the sample with the fissure length
of 15 mm.
It needs to be noted that tensile wing crack I is usu-
ally easier to initiate and nucleate from the tips of the
longer fissures, e.g. the fissure length is 20 or 25 mm.
At keeping the fissure length of the sample 15 mm as a
constant, only the sample with fissure angle of 45 ini-
tiates tensile wing crack I from the tips of single fissure.
Tensile wing crack II is observed only when the fissure
length is equal to 20 or 25 mm. But surface spalling
IX initiates in all other samples containing single fis-
sure except for the sample GS-14# . However the tensile

123
240
crack III and far-field crack VIII are all observed in all
the flawed samples. The above phenomenon shows that
cracks III, VIIIIX are easier to occur during the load-
ing for brittle rock material.
In accordance with Table 3, tensile crack V is
rare, which is only observed in the sample GS-19#
(2a = 15 mm, = 15 ) Tensile crack IV initiates from
the tips of fissure only when the fissure angle is lower
than 60 but lateral crack VI is only observed in
the samples with the fissure length of 15 mm except
for the sample GS-11# . Moreover it can be concluded
that the shear crack VII usually occurs in the sam-
ples containing the longer fissures, such as 2a = 20
or 25 mm. But sometimes, the shear crack VII is also
observed in the fissured sample (2a = 15 mm, = 60 )
which shows the fissure angle 60 makes the sample
easier to emanate the crack VII along the direction of
pre-existing fissure.
From Figs. 12 and 13, it can also be seen that
some wing cracks appear to be arrested a certain dis-
tance from the upper and lower boundary of the sam-
ple, which is probably related to the confining stress
induced by the friction between the sample and the
steel platen.
4.2 AE behaviors of intact and flawed sandstone
samples with single fissure geometries
AE monitoring is very useful for exploring the fail-
ure process and crack coalescence mechanism of brittle
rocks, especially for the flawed rocks with pre-existing
fissures. Therefore, during the uniaxial compression
test, the AE technique is adopted for the sandstone sam-
ples in this research.
Figure 14 shows the AE counts and AE energy of
intact sandstone sample during the whole deformation
failure, however, Figs. 15, 16, 17, 18, 19 show the AE
counts of some typical sandstone sample containing
single fissure in the process of uniaxial compression
test. In accordance with the AE results as shown in
Figs. 14, 15, 16, 17, 18, 19, a detailed investigation
for the AE characteristics of brittle sandstone samples
with different fissure geometries was carried out. The
AE characteristics of intact and flawed sandstone con-
taining single fissure can be approximately divided into
three typical periods, i.e. quiet period, active period and
remission period (Fig. 14a).
123
S.-Q. Yang, H.-W. Jing
At the quiet period, the uniaxial compression stress-
time curve of intact and fissured sandstone sample all
show the downward concave and the initial nonlin-
ear deformation at low stress levels. Notice, the quiet
period corresponds to the stage of fissure closure, there-
fore the AE events are not very active and the AE counts
are also very rare compared with the active period,
which results from only occurrence of fissure closure at
this stage. Moreover we can see that for hard and brittle
sandstone material, the AE behavior at quiet period is
not dependent to the fissure geometries including fis-
sure length and fissure angle.
At the active period, the stressstrain curves of sand-
stone samples undergo two different stages, i.e. the
stage of elastic deformation, and the stage of crack
growth and propagation. With the increase of time,
the axial stress increases linearly at the stage of elas-
tic deformation, but when increasing to yielding stress,
the axial stress-time curve occurs to depart from the lin-
ear curve and shows distinctly nonlinear behavior. For
intact sample (Fig. 14), the AE events are quite active,
and the AE counts are very dense and stable. More-
over with the increase of axial stress, the AE counts
and energy of intact sample all increase continuously.
But after yielding stress, although the AE counts of
intact sample do not increase with the increase of time,
the AE energy increase rapidly as shown in elliptical
zone (Fig. 14), which further validates abrupt splitting
failure of brittle sandstone (Fig. 7).
However for flawed sample with single fissure
(Figs. 15, 16, 17, 18, 19), the AE behaviors during the
deformation failure are different to that of intact sample
(Fig. 14) at the active period, which is distinctly related
to fissure length and fissure angle. The AE counts of
flawed sample with single fissure is more decentralized
than that of intact sample, moreover the AE counts of all
flawed samples with single fissure have several larger
peak values before peak strength, which correspond
to the crack initiation and propagation in the samples.
Notice, the quiescence of the AE observed before the
failure is probably due to the increasing attenuation as
cracks propagate. This is often observed particularly
when spalling occurs near the surface Wassermann
et al. (2009).
In accordance with the AE counts shown in Figs. 15,
16, 17, one can conclude that the fissure length has
a great influence on the AE behavior at the active
period. When the fissure length is shorter (2a = 5 mm)
in Fig. 15, the dense extent of AE counts of flawed
Strength failure and crack coalescence behavior 241

Fig. 14 AE counts and AE a 250 400

energy of intact sandstone
sample during the whole AE Counts (100)
Remission period
deformation failure. The Axial stress
number in double bracket in 200
300

AE Counts (100) /s
vertical axis labels is the

Axial stress 1 / MPa

multiplier of coordinates
150
Quiet period Active period

200

100

100
50

0 0
0 100 200 300 400 500 600
Time / seconds

b 250
900
AE energy (100)
200 Axial stress
720

AE energy (100) rate

Axial stress 1 / MPa

150
540

100
360

50 180

0 0
0 100 200 300 400 500 600
Time / seconds

sample is approximately similar to that of intact sam-
ple (Fig. 14a), but the flawed sample has several larger
peak values. For the samples containing medium fis-
sure (2a = 15 mm) in Fig. 16, several larger peak val-
ues of AE signals are observed at the stage of elastic
deformation, while the AE counts are very rare at the
stage of crack growth and propagation, which shows
that there are no distinct crack coalescence at this stage.
However for the samples containing longer fissures
(2a = 20, 25 mm) as shown in Fig. 17, the AE behav-
ior is reverse to that of flawed samples (2a = 15 mm),
whereby the AE events are not active and AE counts are
very rare at the stage of elastic deformation, but very
dense at the stage of crack growth and propagation.
From Figs. 18, 19 one can also see the effect of fis-
sure angle on the AE behaviors at the active period.
For flawed samples with the lower angle ( = 15 ) the
AE events are not very active and AE counts are very
rare expect for several larger AE peak values. However
for flawed samples with the larger angle ( = 60 )
the AE behaviors are approximately the same to that
with the lower angle ( = 15 ) at the stage of elastic

123
242 S.-Q. Yang, H.-W. Jing

180 400 a 100 250

3
AE Counts (100)
2 4 AE Counts (100)
150 Axial stress 4

AE Counts (100) /s
Axial stress 1 / MPa

Axial stress 3

Axial stress 1 / MPa

80 200

AE Counts (100) /s
1 300 5
2
120 6
60 150
90 200 1

40 100
60
100
30 20 50

0 0 0 0
0 100 200 300 400 500 600 0 30 60 90 120 150 180 210
Time / seconds Time / seconds
80 250
Fig. 15 Relation between axial stress, AE counts and time of b B
2, 3
5
sandstone sample containing short fissure (2a = 5 mm, = 45 )
AE Counts (100) 5
Axial stress 4

AE Counts (100) /s
6

Axial stress 1 / MPa

200
60
7
140 300 150
AE Counts (100)
120 Axial stress 40 1
3 250
Axial stress 1 / MPa

AE
AE Counts (100) /s
A events
100
4
100 5 were not
2 200
20 observed
80 1 50
1 150
60 1
0 0
100 0 40 80 120 160 200 240
40
Time / seconds
20 50

0 0 Fig. 17 Relation between axial stress, AE counts and

0 50 100 150 200 250 time of sandstone sample containing the longer fissure a
Time / seconds 2a = 20 mm, = 45 ; b 2a = 25 mm, = 45 . Note, the AE
counts in the square were not observed
Fig. 16 Relation between axial stress, AE counts and time of
sandstone sample containing medium fissure (2a = 15 mm, = 160 250
45 ) AE Counts (100)
4 5
140 Axial stress 6
Axial stress 1 / MPa

200

AE Counts (100) /s
120 3

2
100 150
deformation, while the AE counts at the stage of crack
80
growth and propagation are more and denser than that 1 100
with the lower angle ( = 15 ) 60

40
At the remission period (i.e. the stage of strain-soft- 50
20
ening), the AE counts and energy of intact sample are
0 0
very rare. The AE behaviors of flawed sample with 0 50 100 150 200 250 300 350 400 450
shorter fissure (2a = 5 mm) or higher angle ( = 60 ) Time / seconds
are approximately the same to that of intact sample,
which results from little occurrence of macroscopic Fig. 18 Relation between axial stress, AE counts and time of
sandstone sample containing the lower angles (2a = 15 mm, =
cracks after peak strength. But, the AE behaviors of 15 )
flawed sample with medium fissure (2a = 15 mm) are
not dependent to the fissure angle, which shows that the
AE counts undergo firstly a relative steady phase and AE events in the square in Fig. 17b were not observed,
then increase abruptly with increasing time, as shown which is due to the departure between AE sensor and
in Figs. 16 and 18. However, for flawed sample with sample surface.
longer fissure (2a = 20 and 25 mm), it can be seen from In sum, the above analysis shows clearly that the AE
Fig. 17 that the AE counts are more and relative denser, behaviors of sample are closely dependent to the crack
which results from gradual crack coalescence in the coalescence in the sample, which will be analyzed and
samples after peak strength. It needs to be noted that the discussed detailed in next section.

123
Strength failure and crack coalescence behavior 243

160 400
AE Counts (100) 1
2
Figure 20 shows typical real-time crack coales-
140
Axial stress 1 / MPa

Axial stress 3 cence process of sandstone samples containing a

120
100
80
60
40
20
0 0
0 50 100 150 200 250 300 350 400 450
Time / seconds
Fig. 19 Relation between axial stress, AE counts and time of
sandstone sample containing the higher angles (2a = 15 mm,
= 60 )
4.3 Real-time crack coalescence process of sandstone
samples containing a single fissure
Besides previous AE technique, the photographic mon-
itoring was also adopted during uniaxial compression
test to clarify further the precedence relationship of
crack coalescence in sandstone sample. Based on the
combined results on AE and photographic monitoring,
the real-time crack coalescence process of sandstone
containing a single fissure can result in a more detailed
analysis as follows.
AE Counts (100) /s
1 300 single fissure (2a = 25 mm, = 45 ) by using photo-
graphic monitoring. From Fig. 20, it is very clear that
200 photographic monitoring can be used to identify and
characterize the crack coalescence process of flawed
100 sandstone. Notice, the denoted number and letter shown
in Fig. 20 are corresponding to Fig. 17b. Figure 21
shows the influence of real-time crack coalescence on
the strength and deformation failure behavior of sand-
stone samples containing a single fissure (2a = 25 mm,
= 45 ) under uniaxial compression.
In accordance with Figs. 17b and 20, 21 we can
analyze real-time crack coalescence process of sand-
stone sample containing a single fissure (2a = 25 mm,
45 ). Before point A (1 = 27.55MPa = 35.7%c )
as shown in Fig. 17, the stress concentration at the tips
of fissure does not reach the material strength nearby
the tips, which does not result in a crack initiation. But
when the axial stress is loaded to point A, the sam-
ple begins to initiate the wing crack I from the upper
and under tips of fissure, but the crack is too fine to be
observed with the unaided eyes or in video recordings.
Notice the initiated stress of wing crack I is identi-
fied by AE counts, as shown in Fig. 17b. After point
A, with the increase of axial deformation, the wing

Fig. 20 Typical real-time

crack coalescence process A B
of sandstone sample
3
containing a single fissure 1
2
(2a = 25 mm, = 45 ). The 1

denoted number and letter

shown in the figure are
corresponding to Fig. 17b 1
1

1 = 27.55MPa 1 = 34.1230.38MPa 1 =71.44MPa c = 77.2570.97MPa c = 77.2570.97MPa

7
6

4 7

5 5

1 = 70.9768.9MPa 1 = 75.365.95MPa 1 = 72.562.3MPa 1 = 67.455.8MPa 1 =55.816.9MPa

123
244
Fig. 21 Effect of real-time
crack coalescence on the
80
strength and deformation
behavior of sandstone
samples containing a single
fissure (2a = 25 mm, 60
Axial stress 1 / MPa
= 45 ) under uniaxial
compression
40
20
0
0.0
crack I further lengthens and propagates gradually to
the edge of the sample. When the sample is loaded
to 34.12 MPa (44.2%c ), the wing crack I undergoes
a large and rapid evolution to the edge of the sam-
ple, which leads to the axial supporting capacity drops
to 30.38 MPa in a smaller range of axial strain. After-
wards, with the increase of axial deformation, the wing
crack I is very difficult to lengthen due to the limita-
tion of sample boundary, but widens a lot. The con-
tinuous increase of deformation leads to the sample
being loaded to 77.25 MPa (100%c ). At this moment
of about 171 s, the cracks 23 (crack type III) develop
rapidly to the edge of sample in one second, and two
subsequent tensile cracks III can only be snapped with
photographic monitoring.
However after the sample is loaded to peak strength,
the axial supporting capacity does not drop to zero. For
the moment, the axial stress of this sample is about
70.97 MPa and corresponding axial strain is 3.67
103 . Notice, the sample has still a supporting struc-
ture (Fig. 21). Afterwards, with the increase of axial
deformation, the shear crack VII emanates from the
upper tip of fissure and the corresponding axial stress
drops rapidly from 70.97 to 68.9 MPa. At this time, the
axial supporting capacity begins to increase slowly with
increasing deformation. While due to obvious damage
of supporting structure, the reloading average modulus
E a (about 20 GPa) is distinctly lower than the elastic
modulus (28.56 GPa) before peak strength. When the
sample reaches the axial stress 75.3 MPa, the crack
5 (crack type III) begins to initiate and propagate from
the under tip of fissure along the direction of axial load-
123
S.-Q. Yang, H.-W. Jing
I+III Ea=20 GPa
ES=28.56GPa
Crack type I Ea=20 GPa
Ea=15 GPa
I+II+III+VII
I+III+VII I+III+VII
2a=25mm, 45 I+II+III+VII
+VIII+IX
1.0 2.0 3.0 4.0 5.0 6.0
Axial strain 1 / 10-3
ing, which leads to the axial supporting capacity drops
to 65.95 MPa in a small range of axial strain. After-
wards, with increasing axial deformation, the axial
stress begins to increase slowly and the reloading aver-
age modulus is approximately equal to 20 GPa, which
indicates that the initiation and coalescence of the crack
5 does not affect the average modulus, but reduce
the axial supporting capacity of fractured rock. With
the increase of axial deformation, the axial supporting
capacity begins to increase slowly. But due to more
obvious damage of supporting structure, the reloading
average modulus (15 GPa) is distinctly lower than pre-
vious average modulus (20 GPa) and the axial stress-
axial strain begins to depart gradually towards the axial
strain. When the sample was loaded to 67.4 MPa (cor-
responding axial strain is 5.06 103 ), crack 6 (crack
type II) occurs very fast. The axial stress of the sam-
ple drops to 55.8 MPa and corresponding axial strain is
5.09 103 . Finally, with the increase of axial defor-
mation, some far-field crack VIII and surface spalling
IX are also observed, which results in rapid reduction
of axial supporting capacity until 16.9 MPa in an almost
constant axial strain.
By using photographic monitoring, Fig. 22 also
presents the influence of real-time crack coalescence on
the strength and deformation failure behavior of sand-
stone samples containing a single fissure (2a = 15 mm,
= 15 ) under uniaxial compression. Notice, the
denoted number shown in Fig. 22 is corresponding to
Fig. 18.
Compared Figs. 22 with 21, it can be seen that more
main crack coalescences shown in Fig. 21 occur after
Strength failure and crack coalescence behavior 245

Fig. 22 Effect of real-time III+V+IV

III+V+IV+IX
crack coalescence on the
strength and deformation V+IV
behavior of sandstone 3
samples containing a single Ea=27.0GPa
150
fissure 2 Ea=30.7GPa m
g 5
(2a = 15 mm, = 15 ) o
Crack type V h p
under uniaxial compression 120
e
n

Axial stress 1 / MPa

c f III+V+IV+VI
III+V+IV+IX +VIII+IX
1 d
90 ES=32.86GPa 6
b
Ea=33.3GPa 6
1
60
a 6
4

30
2a=15mm, 15 q

0
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0
-3
Axial strain 1 / 10

peak strength, while more main crack coalescences in
Fig. 22 take place before peak strength, which results
from the difference of fissure length and angle in the
sample.
In accordance with Fig. 22, we can also analyze
real-time crack coalescence process of sandstone sam-
ple containing a single fissure (2a = 15 mm, = 15 ).
Before point a(1 = 58.40 MPa = 41.9%c ) as shown
in Fig. 22, the stress concentration at the tips of fissure
does not reach the material strength nearby the tips,
which does not result in a crack initiation. But when
the axial stress is loaded to point a, the sample begins
to produce rapidly the wing crack 1 (crack type V)
from the upper and under tips of fissure, which can be
observed with the naked eyes. With the crack 1 propa-
gates to the edge of the sample, the axial stress under-
goes a minor drop to 57.4 MPa from 58.4 MPa. After
point a, the crack 1 is very difficult to lengthen due
to the limitation of sample boundary. The axial stress
increases linearly with the axial deformation as the elas-
tic modulus of 32.86 GPa (Fig. 23a). When the sample
is loaded to point c(1 = 99.9 MPa = 71.7%c ), the
crack 2 (crack type IV) produces from the under tip of
fissure and propagates towards the upper edge of the
sample rapidly, which leads to the corresponding axial
stress drops rapidly from 99.9 to 98.4 MPa. At this time,
the sample has still a good supporting structure. There-
fore with the increase of axial deformation, the axial
supporting capacity begins to increase with approxi-
mately same average modulus 33.3 GPa (Fig. 23b)
as elastic modulus 32.86 GPa (Fig. 23a), which indi-
cates that the initiation and propagation of cracks 12
have no influence on the elastic deformation behav-
ior of the sample. When the sample is loaded to point
e (1 = 114.2 MPa = 82%c ), the crack 3 (crack
type III) emanates from the upper tip of fissure and
the corresponding axial stress drops rapidly from 114.2
to 112.4 MPa. Afterwards, the axial supporting capac-
ity begins to increase slowly with increasing deforma-
tion. While due to some minor damage of supporting
structure, the reloading average modulus 30.7 GPa
(Fig. 23c) is a little lower than the elastic modulus
(32.86 GPa).
Afterwards, when the sample is loaded to point
g(1 = 135.1 MPa = 97%c ), the crack 4 (crack type
IV) emanates from the upper tip of fissure and the
corresponding axial stress drops rapidly from 135.1
to 129.4 MPa. For the moment, the axial support-
ing capacity begins to increase slowly with increas-
ing deformation. While due to obvious damage of
supporting structure, the reloading average modulus
27.0 GPa (Fig. 23d) is distinctly lower than the elas-
tic modulus (32.86 GPa). The continuous increase
of deformation leads to the sample being loaded
to 139.28 MPa (100%c ), at this moment, the crack
5 (crack type III) develops rapidly in one second,
which leads to the axial supporting capacity drops to
127.7 MPa in a smaller range of axial strain. From
Fig. 23, we can see clearly that the axial stress of the
sample has a better linear relation with the axial strain

123
246 S.-Q. Yang, H.-W. Jing

Fig. 23 Relation between 116 e

Axial stress 1 / MPa

axial stress and axial strain 73

Axial stress 1 / MPa

ES=32.86GPa b Ea=33.3GPa
112
after several crack 70
coalescences in brittle 67 108
sandstone sample
64 y = 33.27 x - 36.55
containing a single fissure y = 32.86 x - 30.63 104
(2a = 15 mm, = 15 ) 61 R2 = 1.00 R2 = 1.00
a 100
under uniaxial compression 58 d
(All the symbols represent 55 96
experimental results and the 2.6 2.7 2.8 2.9 3 3.1 3.2 4 4.1 4.2 4.3 4.4 4.5 4.6
lines represent the Axial strain 1 / 10 -3
Axial strain 1 / 10 -3

regression results)
g Peak strength
135
Axial stress 1 / MPa

Axial stress 1 / MPa

Ea=30.7GPa 140 m

130 Ea=27.0GPa

125 136

y = 30.71 x - 26.72 y = 26.96 x - 13.41

120
R2 = 1.00 132 R2 = 0.99
115
f
h
110 128
4.5 4.6 4.7 4.8 4.9 5 5.1 5.2 5.3 5.3 5.4 5.5 5.6 5.7
Axial strain 1 / 10-3 Axial strain 1 / 10-3

132
after each larger crack coalescence in the sample, which y = 22.26 x - 1.12 Ultimate failure
Axial stress 1 / MPa

R2 = 0.97
results probably from a kind of brittle and hard sand- 130
stone material in this research. However after peak o
128
strength, the relation between axial stress and axial Ea=22.3GPa

strain shown in Fig. 24 is distinctly different from that 126 n

shown in Fig. 23. After the axial stress of the sam- 124 p
ple drops to point n(1 = 126.22 MPa = 90.6%c ) 123.917.3MPa
122
from 127.7 MPa, the axial supporting capacity begins to
5.5 5.6 5.7 5.8 5.9 6 6.1 6.2
increase slowly with the increase of axial deformation.
Axial strain 1 / 10 -3

But due to more obvious damage of supporting struc-

ture, the reloading average modulus (22.26 GPa) is dis- Fig. 24 Relation between axial stress and axial strain in brittle
tinctly lower than previous average modulus (27 GPa) sandstone sample containing a single fissure (2a = 15 mm, =
and the axial stress-axial strain begins to depart grad- 15 ) after peak strength under uniaxial compression
ually towards the axial strain. When the sample was
loaded to point o, the crack 6 occurs very fast. The Table 4, the numbers 17 represents the succession of
axial stress of the sample drops to 17.3 MPa and the crack coalescence, which corresponds to that as shown
corresponding axial strain is 6.10 103 . Moreover, in Figs. 12,13. It can be seen that each larger stress
some far-field crack VIII and surface spalling IX are drop in the axial stress-axial strain curve all indicates a
also observed, which results in rapid reduction of axial larger crack propagation, which also corresponds to a
supporting capacity until 17.3 MPa in an almost con- larger amplifying of AE counts. Notice, the initiation,
stant axial strain. propagation and coalescence of some cracks occur after
By adopting the same method as sample GS-17# peak strength, which shows the re-fracture mechanical
(2a = 25 mm, = 45 ) and GS-19# (2a = 15 mm, behavior of flawed sample after failing. Moreover, the
= 15 ), the real-time crack coalescence process corresponding axial stress of initiating the first crack is
of all other samples with a single fissure can also be closely dependent to fissure length and fissure angle.
identified and characterized on basis of photographic With the increase of fissure length, the corresponding
monitoring and AE technique. Table 4 summarizes the axial stress of initiating the first crack tends to decrease
successive axial stress values of crack coalescence in except for the sample GS-10# , which shows that the ini-
the sandstone samples containing a single fissure. In tiation of the first crack in the sample with longer fissure

123
 Table 4 Succession of crack coalescence in sandstone sample containing single fissure
Sample Subsequence/Corresponding axial stress formed the crack in the sample (Unit: MPa)

1 2 3 4 5 6 7

GS-04# III VIII VIII, IX VIII, IX

(128.2) (148 144.3) (170.04 = c ) (160.9 68.3)
GS-07# III IX VIII, IX
(95.93 88.45) (103.9, 136.39 = (112.17 8.49)
Strength failure and crack coalescence behavior

c , 135.2
133)
GS-10# I IV III, IV VI, VIII, IX
(101.6i ) (120.8 118) (122.8=c 90.8) (89.8 12.9)
GS-14# I II III VIII VIII, III VII, III
(43.77i 40.35) (67.73 65.5) (79.28 78.4) (85.29 = c 81.7) (76.1 73.6) (69.5 59.8)
GS-17# I III III VII III II VIII, IX
(27.55i , 34.12 (77.25 = c 70.97) (77.25 = c 70.97) (70.97 68.9) (75.3 65.95, (67.4 55.8) (55.8 16.9)
30.38) 72.5 62.3)
GS-19# V IV III IV, IX III, IX VI, VIII, IX
(58.4i 57.4) (99.9 98.4) (114.2 112.4) (135.1 129.4) (139.28 = c 127.7) (123.9 17.3)
GS-11# I VIII III IV, IX VIII, IX
(49.2i , 64.3) (83.980.7) (115.17=c 108.9) (109.9104.6) (88.8 4.4)
GS-25# III III, IV III, VIIX
(117i , 146.57 144.3) (149.96 = c 143.6) (143.6 122.1 24)
GS-28# III (103.54i VI, VII, IX (179.9
101.1, 181.71 = 22.8)
c 179.9)
The letter i indicates the initiated axial stress of the first crack

123
247
248
Fig. 25 Typical real-time
crack coalescence process
of sandstone sample
containing a single fissure
(2a = 10 mm, = 45 )
Crack type III Crack type III+IX
1 = 95.9388.45MPa 1 = 103.9MPa
is easier than that with shorter fissure. However the cor-
responding axial stress of initiating the first crack has
a nonlinear relation with the fissure angle, but for the
sample GS-11# with = 45 , the corresponding axial
stress is lowest in all the samples with fissure length of
15 mm to initiate the first crack. Moreover the hetero-
geneity of sandstone material has a significant effect
on the axial stress to initiating the first crack.
In accordance with Table 4, one can further con-
clude that more crack coalescences occur in the samples
with longer fissure or smaller fissure angle, which indi-
cates that the localization deformation failure (Fig. 6)
is easier to initiate in the samples with longer fissure
or smaller fissure angle. For the sample with shorter
fissure (e.g. GS-07# ) as shown in Fig. 25, the surface
spalling IX often occurs in brittle rock material but does
not complete at once. From Fig. 25, it can be seen that
the surface spalling IX in the sample GS-07# undergoes
four processes with an increase of axial deformation.
5 Conclusions
The aim of this experimental study is to investigate
the strength, deformation failure, and crack coales-
cence behavior in brittle sandstone material containing
a single fissure under uniaxial compression by photo-
graphic monitoring and AE technique. On the basis of
the experimental results of sandstone samples contain-
ing a single fissure, the following conclusions can be
drawn.
The fissure length and fissure angle have a key
effect on the strength and deformation behavior of
sandstone samples under uniaxial compression. In the
present research, tested intact sandstone sample is a
kind of typical brittle rock and takes on typically axial
123
S.-Q. Yang, H.-W. Jing
Ss Ss Ss
Ss
Ss
Crack type III+IX Crack type III+IX Crack type III+VIII+IX
1 = 136.39MPa =c 1 =135.2133MPa 1 = 112.178.49MPa
splitting failure. Compared with intact sample, the
flawed sample containing a single fissure fails with
lower strengths, smaller Youngs modulus and smaller
peak axial strain, which is closely related to the fis-
sure length and fissure angle. Moreover, the flawed
sample takes on the localization deformation failure
behavior. The mechanical parameters of flawed sand-
stone decrease with the increase of fissure length, but
firstly decrease and then increase with increasing fis-
sure angle. Moreover the influence of single fissure
geometries on the uniaxial compressive strength and
peak axial strain is higher than that on the Youngs
modulus for this kind of brittle sandstone material.
Nine different crack types are identified based on
their geometry and crack propagation mechanism (ten-
sile, shear, lateral crack, far-field crack and surface
spalling) for single fissure in response to the applied
axial loads. Tensile crack I and III are often the first
crack to initiate from the tips of pre-existing fissure.
But sometimes, the tensile crack V is also the first
crack, and is only observed in the sample GS-19#
(2a = 15 mm, = 15 ), which shows that this fis-
sure angle makes the sample easier to emanate this
tensile crack V. Tensile crack IV initiates from the
tips of fissure only when the fissure angle is lower
than 60 , but lateral crack VI is only observed in the
samples with the fissure length of 15 mm except for
one sample GS-11# . However, anti-tensile crack IV
often accompanies with tensile crack III towards the
reverse direction. The shear crack VII usually occurs
in the samples containing the longer fissures, while the
cracks III, VIIIIX are easier to occur during the load-
ing for brittle rock material. In accordance with nine
crack types, the failure mode of flawed sandstone sam-
ple containing single fissure under uniaxial compres-
sion is analyzed. The failure modes of flawed samples
Strength failure and crack coalescence behavior
are all a mixture of several cracks among nine crack
types.
According to AE monitoring results in the process
of deformation failure, the AE characteristics of intact
and flawed sandstone containing single fissure can be
approximately divided into three typical periods, i.e.
quiet period, active period and remission period. At
quiet period, for hard and brittle sandstone material,
the AE events are not very active and the AE counts are
very small, which results from the occurrence of fissure
closure. Moreover the AE behavior at quiet period is
not dependent on the fissure length and fissure angle.
At the active period, the AE counts of intact sample are
very dense and stable, but the AE behaviors of flawed
samples are different with that of intact sample, which
is distinctly related to fissure length and fissure angle.
The AE counts of flawed sample with single fissure is
more decentralized than that of intact sample, moreover
the AE counts of all flawed samples with single fissure
have several larger peak values before peak strength,
which correspond to the crack initiation and propaga-
tion in the samples. However, at the remission period,
the AE counts and energy of intact sample are very
rare, but AE behaviors of flawed sample with shorter
fissure (2a = 5 mm,) or higher angle ( = 60 ) approx-
imate the one of intact sample, which results from little
occurrence of macroscopic cracks after peak strength.
While, the AE behaviors of flawed sample with medium
fissure (2a = 15 mm,) are not dependent on the fissure
angle, which shows that the AE counts undergo firstly
a relative steady phase and then increase abruptly with
increasing time. For flawed sample with longer fissure
(2a = 20 and 25 mm), the AE counts are more and rel-
ative denser, which results from gradual crack coales-
cence in the samples after peak strength.
In order to confirm the subsequence of crack coa-
lescence in sandstone sample, the photographic mon-
itoring was adopted during uniaxial compression test.
On basis of the photographic monitoring results, the
real-time crack coalescence process of sandstone con-
taining a single fissure are identified and characterized.
Moreover, the corresponding axial stress of crack coa-
lescence in the sandstone samples containing a single
fissure is also obtained. Each larger stress drop in the
axial stress-axial strain curve all indicates a larger crack
propagation, which also corresponds to a larger amount
of AE counts. The corresponding axial stress of initiat-
ing the first crack is closely dependent to fissure length
and fissure angle. With the increase of fissure length,
249
the corresponding axial stress of initiating the first crack
tends to decrease except for the sample GS-10# , which
shows that the initiation of the first crack in the sam-
ple with longer fissure is easier than that with shorter
fissure. However, the corresponding axial stress of ini-
tiating the first crack has a nonlinear relation with the
fissure angle, but for the sample GS-11# with = 45 ,
the corresponding axial stress is lowest in all the sam-
ples with fissure length of 15 mm to initiate the first
crack. At the end, two typical effects of the real-time
crack coalescence on the strength and deformation fail-
ure behavior of brittle sandstone sample are analyzed
under uniaxial compression.
This paper investigates only the strength failure and
crack coalescence behavior of brittle sandstone con-
taining a single fissure with different lengths and angles
under uniaxial compression by photographic monitor-
ing and AE technique. In order to understand deeply the
failure behavior and fracture mechanism of engineer-
ing rock mass, more experimental investigations for
flawed brittle sandstone samples with two or three fis-
sures (varying ligament angle, ligament length, fissure
angle and fissure length, etc.) are being conducted by
advanced techniques (such as AE technique and pho-
tographic monitoring, etc.).
Acknowledgments The authors would like to acknowledge the
financial support by the National Natural Science Foundation
of China (no. 50709008), China Postdoctoral Science Founda-
tion funded project (no. 20100470068) and Blue Project of Col-
leges and Universities of Jiangsu Province in China (2010). We
also would like to express our sincere gratitude to Prof. Andr
Vervoort and another anonymous reviewer for their valuable
comments, which have greatly improved this paper.
References
Bobet A (2000) The initiation of secondary cracks in compres-
sion. Eng Fract Mech 66:187219
Bobet A, Einstein HH (1998) Fracture coalescence in rock-type
materials under uniaxial and biaxial compression. Int J Rock
Mech Min Sci 35:863889
Chang SH, Lee CI (2004) Estimation of cracking and damage
mechanisms in rock under triaxial compression by moment
tensor analysis of acoustic emission. Int J Rock Mech Min
Sci 41:10691086
Debecker B, Vervoort A (2009) Experimental observation of
fracture patterns in layered slate. Int J Fract 159:5162
Eberhardt E, Stimpson B, Stead D (1999) Effects of grain size on
the initiation and propagation thresholds of stress-induced
brittle fractures. Rock Mech Rock Eng 32(2):8199
Fairhurst CE, Hudson JA (1999) Draft ISRM suggested method
for the complete stress-strain curve for the intact rock in
123
250
uniaxial compression. Int J Rock Mech Min Sci 36:279
289
Feng XT, Ding WX, Zhang DX (2009) Multi-crack interaction
in limestone subject to stress and flow of chemical solutions.
Int J Rock Mech Min Sci 46(1):159171
Fujii Y, Ishijima Y (2004) Consideration of fracture growth from
an inclined slit and inclined initial fracture at the surface of
rock and mortar in compression. Int J Rock Mech Min Sci
41:10351041
Ganne P, Vervoort A (2007) Effect of stress path on pre-peak
damage in rock induced by macro-compressive andten-
sile stress fields. Int J Fract 144(2):7789
Huang ML, Wong RHC (2007) Experimental study on propa-
gation and coalescence mechanisms of 3D surface cracks.
Chin J Rock Mech Eng 26(9):17941799 (in Chinese)
Jouniaux L, Masuda K, Lei X et al (2001) Comparison of the mi-
crofracture localization in granite between fracturation and
slip of a preexisting macroscopic healed joint by acoustic
emission measurements. J Geophys Res 106(B5):8687
8698
Labuz JF, Cattaneo S, Chen L-H (2001) Acoustic emission
at failure in quasi-brittle materials. Constr Build Mater
15:225233
Lei XL, Masuda K, Nishizawa O et al (2004) Detailed analysis
of acoustic emission activity during catastrophic fracture of
faults in rock. J Struct Geol 26:247258
Li YP, Chen LZ, Wang YH (2005) Experimental research on
pre-cracked marble under compression. Int J Solids Struct
42:25052516
Lu ZD, Ding WX, Feng XT, Zhang YL (2008) Experimental
study on mechanical-hydraulic-chemical coupling process
in cracked rocks. Chinese Journal of Rock Mechanics and
Engineering 27(4):796804 (in Chinese)
Moradian ZA, Ballivy G, Rivard P, Gravel C, Rousseau B (2010)
Evaluating damage during shear tests of rock joints using
acoustic emissions. doi:10.1016/j.ijrmms.2010.01.004
Nemat-Nasser S, Horii H (1982) Compression-induced nonlin-
ear crack extension with application to splitting, exfoliation,
and rockburst. J Geophys Res 87(B8):68056821
Pestman BJ, Van Munster JG (1996) An acoustic emission study
of damage development and stress-memory effects in sand-
stone. Int J Rock Mech Min Sci Geomech Abstr 33(6):585
593
123
S.-Q. Yang, H.-W. Jing
Prudencio M, Van Sint Jan M (2007) Strength and failure modes
of rock mass models with non-persistent joints. Int J Rock
Mech Min Sci 44(6):890902
Tang CA, Kou SQ (1998) Crack propagation and coalescence
in brittle materials under compression. Eng Fract Mech
61:311324
Tang CA, Lin P, Wong RHC, Chau KT (2001) Analysis of crack
coalescence in rock-like materials containing three flaws-
part II: numerical approach. Int J Rock Mech Min Sci
38:925939
Tham LG, Liu H, Tang CA etal (2005) On tension failure of 2-
D rock specimens and associated acoustic emission. Rock
Mech Rock Eng 38(1):119
Van De Steen B, Vervoort A, Napier JAL (2005) Observed and
simulated fracture pattern in diametrically loaded discs of
rock material. Int J Fract 131:3552
Wang EZ, Shrive NG (1995) Brittle fracture in compression:
mechanisms, models and criteria. Eng Fract Mech 52:1107
1126
Wassermann J, Senfaute G, Amitrano D et al (2009) Evidence
of dilatant and non-dilatant damage processes in oolitic
iron ore: P-wave velocity and Acoustic Emission analyses.
Geophys J Int 177(3):13431356
Wong RHC, Chau KT (1998) Crack coalescence in a rock-like
material containing two cracks. Int J Rock Mech Min Sci
35:147164
Wong RHC, Chau KT, Tang CA, Lin P (2001) Analysis of crack
coalescence in rock-like materials containing three flaws-
part I: experimental approach. Int J Rock Mech Min Sci
38:909924
Wong RHC, Tang CA, Chau KT, Lin P (2002) Splitting failure
in brittle rocks containing pre-existing flaws under uniaxial
compression. Eng Fract Mech 69:18531871
Wong LNY, Einstein HH (2009) Systematic evaluation of crack-
ing behavior in specimens containing single flaws under uni-
axial compression. Int J Rock Mech Min Sci 46(2):239249
Yang SQ, Jiang YZ (2010) Triaxial mechanical creep behavior
of sandstone. Min Sci Technol 20(3):339349
Yang SQ, Jiang YZ, Xu WY, Chen XQ (2008) Experimental
investigation on strength and failure behavior of pre-cracked
marble under conventional triaxial compression. Int J Solids
Struct 45(17):47964819