Engineering Fracture Mechanics 194 (2018) 136–144

Contents lists available at ScienceDirect

Engineering Fracture Mechanics

journal homepage: www.elsevier.com/locate/engfracmech

The influences of mode II loading on fracture process in rock using

T
acoustic emission energy
⁎
Qing Lina,b, , Dengtian Maoc, Shan Wangb, Shiyuan Lia,b
a
State Key Laboratory of Petroleum Resources and Prospecting, China University of Petroleum, Beijing 102249, China
b
Department of Engineering Mechanics, College of Petroleum Engineering, China University of Petroleum, Beijing 102249, China
c
CNOOC Safety & Technology Services Co., Ltd, Tianjin 300457, China

A R T IC LE I N F O ABS TRA CT

Keywords: The influences of mode II loading on fracture process in Berea sandstone were investigated by
Fracture process zone comparing mode I and mixed-mode fractures that were created by center and eccentric notch
Acoustic emission (AE) beams under three-point bending. Fracture process was characterized by Acoustic Emission (AE)
AE energy and related energy. Experimental results show mode II loading does not affect the length of fully-
Mixed-mode fracture
developed fracture process zone and maximum value of AE energy. However, it affects the zone
Mode II loading
from reaching fully-developed stage, i.e., mixed-mode fracture before or at peak load, and mode I
after the peak. It also breaks the similarity and symmetry of AE events pattern.

1. Introduction

An understanding of fracture processes in rocks is of fundamental importance to solve various engineering problems since cracks
or discontinuities are common structural features for rocks. It involves a detailed characterization of fracture process zone at the
crack front [1,2]. Fracture process zone is generally associated with mode I fracture, but the mixed-mode fracture actually dominates
in-situ conditions.
When a crack starts to propagate under the mixed-mode loading, the fracture path generally follows the criterion of local
symmetry, i.e., the local mode I conditions [3]. Although a mixed-mode fracture advances towards the tensile side, the mode II (KII)
loading still has an essential influence on fracture propagation. The previous research demonstrates that a fracture will propagate
under a mode I (KI) loading that is smaller than pure mode I toughness (KIc) for rocks and other brittle materials, when KII loading is
applied [4–6]. However, those studies are based on linear elastic fracture mechanics assuming a small fracture process zone, and the
influences of KII loading on the fracture process zone in rocks are mostly unknown. Recently, experimental studies reported a mixed-
mode fracture consisting of a fracture process zone two times longer than the mode I fracture [7,8]. The detailed image analyses also
demonstrated that the fracture process zone is an opening zone for both types of fractures [8]. This presents a challenge to interpret
how mode II loading, orthogonal to the orientation of the opening, fundamentally affects the development of the fracture process
zone. In this research, the mixed-mode refers to the combination of mode I and mode II.
It is well known that the fracture process zone in rocks consists of microcracks that produce acoustic emission (AE), and thus AE
has been an effective experimental technique in the study of rock fracture and fracture process zone [9]. The systematic evaluation of
the fracture process zone in rocks using AE began in the late 1970s [1]. AE is a transient elastic wave generated from the sudden
release of strain energy caused by microcracking or other irreversible changes in the material. The cluster of detected AE provides
information to characterize the fracture process zone, especially its size [3]. AE acquired waveform reflects the energy released from

⁎
Corresponding author at: Department of Engineering Mechanics, College of Petroleum Engineering, China University of Petroleum, Beijing 102249, China.
E-mail address: qinglincupb@cup.edu.cn (Q. Lin).

https://doi.org/10.1016/j.engfracmech.2018.03.001
Received 1 February 2018; Accepted 1 March 2018
Available online 02 March 2018
0013-7944/ © 2018 Elsevier Ltd. All rights reserved.
Q. Lin et al. Engineering Fracture Mechanics 194 (2018) 136–144

Nomenclature V recorded AE signal voltage

r distance between AE location and AE sensor
E AE energy t time
G dissipation energy cp p-wave velocity
GF fracture energy σt tensile strength
KI mode I stress intensity factor w crack opening displacement
KIc mode I fracture toughness a notch length
KII mode II stress intensity factor S specimen span
β ratio of KII/KI H specimen height
θ rotation angle between two coordinates

the source, and in practice, it can be used to obtain an estimate of the energy released in the fracture process zone [10–12]. Indeed,
AE energy is not direct information, since the determination of actual microseismic energy involves a comprehensive sensor cali-
bration and source characterization [13]. However, AE energy can be regarded as a useful energy index to study the fractures in rock
and other quasi-brittle materials [11,12,14,15]. Thus, in this study AE events and AE energy were used to investigate the influences of
mode II loading on the fracture process of rock by the comparison of mode I and mixed-mode fractures.

2. Testing material and acoustic emission

2.1. Testing material and experiment setup

The tested material, Berea sandstone, is a flat-bedded, light gray, medium- to fine-grained protoquartzite cemented with silica and
clay [3]. It is early Mississippian and quarried in northern Ohio. The grains range from 0.1 to 0.8 mm. It is homogeneous but slightly
anisotropic in elastic response. The sandstone was used in this study due to its relatively long fracture process zone [3]. The elastic
properties of Berea sandstone are: porosity = 20%, Young’s modulus under compression = 10–14 GPa, Poisson’s ratio = 0.31–0.34,
tensile strength, σt = 3.4–3.6 MPa, uniaxial compressive strength (UCS) = 28–32 MPa. The p-wave velocity measured perpendicular
to bedding was 2160 m/s and parallel to bedding it was 2290 m/s.
A simple testing approach to achieve opening or mixed-mode fracture is a beam under three-point bending with a center for
specimen CN-1 (β = 0%), or eccentric notch for specimen EN-1 (β = 20%), EN-2 (β = 30%), EN-3 (β = 40%). The eccentricity factor
β is the normalized distance, with respect to the half-span, from the centerline (Fig. 1). The nominal size of the beams was height
H = 60 mm, span S = 150 mm, thickness B = 25 mm and the notch length a = 12 mm. A variety of apparent KII/KI = 0, 7.4, 10.0,
11.4% was produced for the specimen with β = 0, 20, 30, 40%, respectively, as determined by a boundary element code assuming
linear elastic fracture mechanics applies [16]. Note that those values of KII/KI only provide a knowledge about mode II loading,
especially when the specimen sizes are similar, but it does not represent the actual loading condition after fracture initiation.
The fracture test was performed in a closed-loop servo-hydraulic load frame, which allowed control of crack propagation through
a constant rate (i.e., 0.15 μm/s) of crack mouth opening displacement (CMOD) at the notch. As the feedback signal, CMOD was
measured by a clip gage extensometer, which measured displacement between two clips bonded to the bottom surface of the beam. A
21 kN load cell was used to measure the applied load and an LVDT with ± 1 mm linear range recorded the deflection of the beam.
Fig. 2 shows a typical loading curve including load vs. CMOD and load-point displacement (deflection). Because Berea sandstone is a
relatively soft rock, all the specimens displayed a similar pattern of the loading curve regardless of the loading conditions.

Fig. 1. Experimental setup and specimen dimensions.

137
Q. Lin et al. Engineering Fracture Mechanics 194 (2018) 136–144

800 800
(a) (b)
700 700

600 600

500 500
Load (N)

Load (N)
400 400

300 300

200 200

100 100

0 0
0 0.1 0.2 0.3 0.4 0.5 0 0.1 0.2 0.3
CMOD (mm) Deflection (mm)

Fig. 2. The typical loading curves: (a) load vs. crack mouth opening displacement (CMOD); (b) load vs. deflection.

2.2. Acoustic emission

The AE system consisted of eight AE sensors (Physical Acoustics model S9225), eight preamplifiers (Physical Acoustics model
1220C), and a data acquisition system with four DAQ cards (National Instruments model PCI-5112). As shown in Fig. 1, four sensors
on each side of the specimen were spaced about 45 mm radially from the notch tip (Fig. 1). Preamplifiers were set to 40 dB gain. One
sensor was selected as the trigger, and trigger levels of either 11 or 15 mV were set for each of the tests performed. Once recording
was triggered, signals were band-pass filtered (0.1–1.2 MHz) and sampled at 20 MHz over a 200 μs window, with a 100 μs pretrigger.
Generally, AE provides information about microcrackings during the experiments: AE source location and AE energy. AE source
location or simply AE location represents a microcrack location, it can be determined through the arrival time of the p-wave in each
waveform [17,18]. By minimizing the error between the definition of wave speed (Eq. (1)) and the distance equation (Eq. (2)), AE
location is:

ri = cP (ti−to) + ε (1)

ri = (x i−x )2 + (yi −y )2 + (z i−z )2 (2)

where ri is the distance between AE location and the ith AE sensor, cP is P-wave speed of the material, to is the event time, ti is arrival
time recorded from the ith sensor, ε = error (residual), and xi, yi, and zi = coordinates of the ith AE sensor. The square of the residual
is
n
ε2 = ∑ ([ (x i−x )2 + (yi −y )2 + (z i−z )2 + cP (to−ti )]2 )
i=1 (3)

The residual was minimized in a least-squares sense using the Levenberg-Marquardt algorithm due to the non-linear nature of Eq. (3).
AE energy is related to the area of an amplitude-time history and it is determined by the root-mean-square of the voltage
transients [12,14,15]. With consideration of attenuation by including distance ri, the relative AE energy recorded from an individual
event is [12]:
te
Ei = ri2 ∫ts Vi (t )2 dt
(4)

where V(t) is the recorded voltage of the AE signal from the ith sensor; ts is the starting time and te is the ending time of the signal.
The units of relative AE energy are mV∗μt1/2∗mm2. AE energy is taken as the average value from all (N = 8) sensors:
N
1
E=
N
∑ Ei
1 (5)

Based on AE energy [19], the events can be categorized by large magnitude, Level 1 Ek/Em ≥ 100, medium magnitude, Level 2
100 ≥ El/Em ≥ 10, and small magnitude, Level 3 El+1/Em ≥ 10, where E1 = maximum energy and Em = minimum, and

(E1 > E2 > ⋯ > Ek ) > (Ek + 1 > ⋯ > El ) > (El + 1 > ⋯ > Em) (6)

All AE locations have a position error of less than 5 mm. The locations with large AE energy generally involve a smaller error,
typically < 3 mm. For example, the accuracy of AE locations within Level 1 is 1.5 mm.

138
Q. Lin et al. Engineering Fracture Mechanics 194 (2018) 136–144

3. Experimental results

3.1. The patterns of AE events during the loading from 90% pre-peak to peak

The sandstone did not produce significant AE events until the load reached 90% of the peak load (i.e., 90% pre-peak), and then
the clustering of AE events was observed above the notch. Fig. 3 shows a series of AE locations from specimen CN-1, EN-1, EN-2, EN-
3, with respect to 90–100% pre-peak. The AE analyses show that there are two zones of AE locations: a small region that includes AE
events in energy Level 1 as large black square dots; a large region that includes all AE events. The small zone contains high energy and
is spatially more concentrated, whereas the large zone covers a much larger area and AE events scatter inside it. For simplicity, the AE
energy is projected along the specimen center (i.e., the fracture path) to obtain a histogram. Distribution of AE energy shows that
more than 90% of AE energy is concentrated in the small zone. The rest of AE events produce little energy. Fracture process zone is a
region with nonlinear material deformation. It is unlikely that a region containing less than 10% of the total AE energy will produce

Fig. 3. AE locations, and AE energy distributions along fracture path, 90–100% pre-peak: (a) CN-1; (b) EN-1; (c) EN-2; (d) EN-3. AE energy unit: mV*μt1/2*mm2.

139
Q. Lin et al. Engineering Fracture Mechanics 194 (2018) 136–144

significant material deformation. Thus, the small zone is regarded as the fracture process zone, where the large zone simply re-
presents a material damage region. Indeed, AE locations always have a position error, but the accuracy of AE locations within energy
Level 1 is 1.5 mm, and thus it is reliable to report the size of fracture process zone.
Fig. 3a shows the results of the mode I specimen CN-1. The AE energy histogram shows that the fracture process zone has a
dimension of 4 × 2 mm with its tip at position A (x = 0 mm, y = 15 mm). The material damage zone occupies 13 × 10 mm that is
more than 10 times larger than fracture process zone. Both zones display a symmetric pattern along the fracture path. It is important
to note that there are some AE events observed under the notch, occupying a length of 2–3 mm approximately. Those AE events are
due to experimental error and the fact that when the notch tip is the position where fracture initiates, a significant amount of AE
activities are generated from the notch. As shown in AE energy histogram, most AE energy is detected surrounding the notch. In
addition, the diffraction and reflection of waves by the notch can result in experimental errors for the waveform from the AE
transducer, and consequently, position errors for the AE location determination. Thus, it is not surprising to observe AE events below
the notch, but they are not considered as part of the fracture process zone.

Fig. 4. AE locations, and AE energy distributions along fracture path, 95–90% post-peak: (a) CN-1; (b) EN-1; (c) EN-2; (d) EN-3. AE energy unit: mV*μt1/2*mm2.

140
Q. Lin et al. Engineering Fracture Mechanics 194 (2018) 136–144

For mixed-mode specimens, the fracture path does not follow the notch orientation and a fracture kink is observed (Fig. 3b–d).
Thus, a new coordinate (x1, y1) is established with the notch tip as the origin. The y1 – axis follows the fracture path. All AE locations
can be transformed into the new coordinate based on Eq. (7):
x1 = xcosθ−(y−11)sinθ, y1 = xsinθ + (y−11)cosθ (7)
The unit for both coordinates is millimeter. The angle θ is the rotation angle, and they are marked for mixed-mode specimens in
Fig. 3b–d.
Similarly, the AE energy histogram along y1 – axis can be obtained by projection of the transformed AE locations. Then the
dimensions of the fracture process zone are determined to be 9 × 2, 9 × 2 and 9 × 3 mm for specimen EN-1, EN-2 and EN-3 re-
spectively with position A as the tip (Fig. 3b–d). The dimensions of material damage region for them are also larger.
Experimental results indicate the consistency with results obtained from speckle interferometry and digital image correlation,
such that the length of fracture process zone in mixed-mode fracture is almost twice than in mode I [7,8]. However, detailed analyses
of AE energy along the fracture path reveal differences between mixed-mode and mode I specimens. For mode I specimen, the value
of AE energy reaches the maximum at the notch tip; for mixed-mode specimens, the value of AE energy is relatively small at the notch
tip. Note that it is a fact that an ideal traction-free crack produces no AE events. Hence, it can be concluded that the fracture process
zone reaches its fully-developed stage when a relatively small value of AE energy is observed. Its position is regarded as where the
fracture process zone ends and the traction-free crack forms.
Based on the AE energy histogram, a traction-free crack occurs at the notch tip for mixed-mode specimens when load reaches the
peak. However, the fracture is still under process at the notch tip for mode I specimen. Fig. 3c and d indicate the end of the fracture
process zone (position B), which is approximately 2 mm above the notch. Mixed-mode results suggest that the traction-free crack
actually forms at the notch tip before the peak load. Note that all the specimen sizes and loading curves are similar, and the material
is homogeneous and only slightly anisotropic. Thus, the differences are considered to be the consequence of mode II loading. In
summary, it is the influence of mode II loading such that the fully-developed fracture process zone forms “earlier” in mixed-mode
fracture than in mode I, although it is debatable whether it occurs “earlier” when mode II loading increases, i.e., a larger KII/KI.
Experimental results show a wider material damage region and an asymmetric pattern of AE events distribution along the fracture
path for mixed-mode fractures. For instance, specimen EN-3 (Fig. 3c) demonstrates the size of material damage region is 16 × 20 mm
approximately, whose width is significantly larger than specimen CN-1. The material damage region does not present a symmetric
pattern along fracture path because more AE events are detected on the right of the fracture path. However, the fracture process zone
maintains its symmetry for mixed-mode fractures. Hence, similarly it can be concluded that the mode II loading results in a wider
damage region and an asymmetric pattern of AE locations, but it does not affect the core region, i.e., fracture process zone. Indeed, it
is debatable whether the width of damage region and degree of asymmetric pattern increases when mode II loading increases.

3.2. The patterns of AE events during the loading from peak to 90% post-peak

After the load passes the peak, specimen CN-1 is observed to have a relatively small AE energy at the notch tip at 90% post-peak,
(Fig. 4a), i.e., fracture process zone reaches its fully-developed stage. The phenomenon is also reported by Bažant and Yu [20], such
that the mode I fracture process zone completely forms at the post-peak for a quasi-brittle material. Thus, the fracture process zone is
10 × 2 mm and the material damage region is 20 × 10 mm. For mixed-mode specimens, fracture process zone changes insignif-
icantly. The detailed observation of AE energy histogram shows that the traction-free crack already moves to y1 = 6 mm for specimen
EN-1 (Fig. 4b), and the fracture process zone is 9 × 3 mm. Specimen EN-2 and EN-3 displays a similar zone, 12 × 4 and 12 × 3 mm
respectively. Similarly, the material damage region is also observed to be wider, and more AE events are detected in the right of
fracture path (Fig. 4d).

3.3. The total AE energy along fracture path for different loadings

Since the measurements from AE events are incremental values with respect to various loadings, they can be summed to obtain a
relatively complete image of the fracture processes. Fig. 5 is the total AE energy released along fracture path from 70% of peak (the
fracture initiation) to different loading of post-peak. The tip (position A) and end (position B) of fracture process zone are marked on
individual AE energy histogram for the specimens with respect to various loadings. The notch tip is also marked, because the different
coordinates were established for different fractures. For mode I fracture, notch tip is at y = 11 mm; for mixed-mode fracture, notch
tip is at y1 = 0 mm. Fig. 5 shows a similar length of fully-developed fracture process zone, i.e., 9–12 mm for all specimens. With
consideration of maximum grain size of 0.8–1 mm for Berea sandstone, fracture process zone is about 11–14 times larger than
maximum grain size. Experimental results suggest that fully-developed fracture process zone is not affected by mode II loading.
The other fundamental concern of this study is the AE energy. AE energy generally can be regarded as a portion of the energy
dissipated during the fracture process, although it is debatable for the relation between AE and fracture energy [14,21]. However, it is
reasonable to assume that once the traction-free crack occurs, the AE energy reaches its maximum value. Then, the AE energy will
remain constant in an ideal manner since AE activities are insignificant for a traction-free crack.
After detailed survey of AE energy histogram, there is no ideal match between the position of maximum value of AE energy and
the end of fracture process zone (position B). For instance, when the load reaches 80% of post-peak for specimen EN-2 (Fig. 5c), the
AE energy is 6 mV∗μt1/2∗mm2 at position B, but the maximum value is more than 8 mV∗μt1/2∗mm2. Although the AE locations with
high AE energy have smaller position errors (less than 1.5 mm), the summation of AEs from different loadings is also a process of

141
Q. Lin et al. Engineering Fracture Mechanics 194 (2018) 136–144

[1] pre70%-peak [2] pre70%-post90% [3] pre70%-post80%

44 44 44
40 40 40
36 36 36
32 32 32 A

y (mm)
y (mm)

y (mm)
28 28 A 28
(a) CN-1 24 24 24
20 A 20 20 B
B
16 16 16
12 12 12
8 8 8
Notch tip 4 4 4
y = 11 mm
0 2 4 6 8 0 2 4 6 8 0 2 4 6 8
AE energy (x1010) AE energy (x1010) AE energy (x1010)

40 40 40
36 36 36
32 32 32
28 28 28
A
24 24 24
A
y1 (mm)
y1 (mm)

y1 (mm)
20 20 20
(b) EN-1
16 16 16
A B
12 12 B 12
8 B 8 8
4 4 4
0 0 0
-4 -4 -4
Notch tip 0 2 4 6
0 2 4 6 0 2 4 6
y1 = 0 mm
AE energy (x1010) AE energy (x1010) AE energy (x1010)

40 40 40
36 36 36
32 32 32
28 28 28 A

24 24 A 24
y1 (mm)
y1 (mm)

y1 (mm)

20 20 20
(c) EN-2 A
16 16 16
B B
12 12 12
8 B 8 8
4 4 4
0 0 0
-4 -4 -4
Notch tip
0 2 4 6 0 2 4 6 8 10 0 2 4 6 8 10
y1 = 0 mm
AE energy (x1010) AE energy (x1010) AE energy (x1010)
40 40 40
36 36 36
32 32 32
28 28 28 A
24 24 A 24
y1 (mm)
y1 (mm)

y1 (mm)

20 20 20
A
(d) EN-3 16 16 16
12 B 12
B
12
B
8 8 8
4 4 4
0 0 0
-4 -4 -4
Notch tip
y1 = 0 mm 0 2 4 6 0 2 4 6 8 0 2 4 6 8 10

AE energy (x1010) AE energy (x1010) AE energy (x1010)

(caption on next page)

142
Q. Lin et al. Engineering Fracture Mechanics 194 (2018) 136–144

Fig. 5. Total AE energy distributions along fracture path, (a) CN-1; (b) EN-1; (c) EN-2; (d) EN-3. The loadings are: pre70%-peak; pre70%-post90%; pre70%-post80%
respectively. AE energy unit: mV*μt1/2*mm2.

accumulation of experimental errors. Thus, it has the difficulty to obtain an ideal observation of maximum value of AE energy at the
end of fracture process zone. However, considering the AE energy distribution, it is reasonable to treat 6 mV∗μt1/2∗mm2 as its
maximum value for specimen EN-2.
Based on the AE energy histograms shown in Fig. 5, the maximum value of AE energy is observed to be more or less a constant
value, i.e., 4–6 mV∗μt1/2∗mm2 for all specimens. Thus, experimental results suggest that the maximum value of AE energy is also not
affected by mode II loading.

4. Discussion

Comparisons between mode I and mixed fractures reveal the influences of mode II loading. AE energy, as an energy index,
provides a valuable parameter to characterize the fracture processes. Based on AE energy and its magnitude, the fracture process zone
is “filtered” from the material damage region; the length of fully-developed fracture process zone and the maximum AE energy are
observed to maintain more or less a constant value for all specimens, with little influences from mode II loading.
These observations help interpret the reported experimental results based on speckle interferometry and digital image correlation
[7,8]. Their results indicated that the fracture process zone in Berea sandstone under mixed-mode loading is longer than mode I
fracture. The reported length is 10–12 mm for mixed mode fracture, and 5–7 mm for mode I. However, their critical assumption is
that the fracture process zone is completely formed at the peak load. In addition, the detailed studies of displacements concluded that
the fracture process zone is an opening zone for two types of fractures [8]. Thus, it poses a challenge to explain how an opening zone
can be fundamentally influenced by mode II loading, in particularly when their orientations are orthogonal. In other words, a
question needs to be answered how the shear stresses create a situation such that the length of an opening zone is doubled. Especially
the specimen sizes are similar and the material is homogeneous and only slightly anisotropic.
The AE energy analyses provide a reasonable interpretation: mode II loading actually does not change the fully-developed fracture
process zone length, but it does affect its formation stage. Mode II loading has a consequence such that the fracture process zone
completely forms “earlier” in mixed-mode fracture than in mode I. For mixed-mode fracture, the zone occurs before or at peak load;
for mode I fracture, it occurs after the peak. Thus, experimental results from digital image correlation [8] can be reinterpreted: the
fracture process zone, as an opening zone, has similar length, i.e., 9–12 mm for the opening and mixed-mode fractures in tested
sandstone.
The formation of fracture process zone is a process to create, propagate and coalesce microcracks in rock and other quasi-brittle
materials. All those phenomena release the energy that also can be regarded as the energy dissipated into the fracture process zone,
i.e., dissipation energy G. Generally the cohesive crack model is popularly used to interpret fracture process zone, and the dissipation
energy can be described by the softening law [20], as shown in Fig. 6. The dissipation energy increases with the increasing opening
displacement. Once the critical displacement is reached, the maximum dissipation energy is also reached, which is the region where
the softening curve covers, i.e., fracture energy GF (Fig. 6). The dissipation energy during the fracture process can be detected by AE
through the generation of AE signals. Thus, it is reasonable to assume that AE energy is a portion of the dissipation energy, although it
is a challenge task to know their relation [14,15,21]. However, if the relation exists no matter if it is linear or nonlinear, AE energy is
a component of the dissipation energy, and the maximum value of AE energy is also a component of fracture energy. Because the
maximum value of AE energy is observed to be similar for two types of fractures, fracture energy can also be considered as a constant
value for two types of fractures. Thus, experimental results in this study also indicate that the fracture energy is not affected by mode
II loading. Because fracture energy is a fundamental material property for the fracture process zone, it can be concluded that mode II
loading does not influence the fracture process zone fundamentally. However, it is important to note that mode II loading is relatively
small for the specimens in this study, and thus a larger KII/KI is recommended for future testing and research.

Fig. 6. Cohesive crack model and dissipation energy.

143
Q. Lin et al. Engineering Fracture Mechanics 194 (2018) 136–144

AE and its related energy are local responses, which provide information for AE events distribution pattern during fracture
processes. The pattern of mode I fracture display two characteristics: similarity and symmetry. Both fracture process zone and
damage region demonstrate a similarity and symmetry along the fracture path for mode I fracture. The mode II loading breaks the
similarity indeed, since the fracture is kinked from the notch. Also, the symmetry is broken because more AE events are observed on
one side of the fracture path. However, the asymmetric pattern of AE events is observed only for the material damage region, and the
fracture process zone as a core region still maintains its symmetry for mixed-mode fractures. Furthermore, the material damage
region is wider in mixed-mode fracture than mode I, probably because of the influences of mode II loading.

5. Summary

For rock and other quasi-brittle materials, shear loading always has a fundamental influence on material failure and fracture
mechanism. For instance, Mohr-Coulomb failure criterion indicates that the material will fail under a tensile stress that is smaller than
its tensile strength, when a shear loading is introduced. Similarly, if a fracture is under a mixed-mode loading, it will initiate and
propagate under a KI loading that is smaller than KIc when a KII loading is present [4–6]. However, those studies [4–6] are based on
linear elastic fracture mechanics, without considering the influences of fracture process zone in the material. Fracture in rock and
quasi-brittle materials always involves the development of a fracture process zone. In this study, experimental results are reported to
show the shear/mode II loading influences on the fracture process zone, although more experimental and theoretical studies are
recommended. Based on the results from fracture and AE testing on Berea sandstone, the following conclusions are drawn:

(1) Mode II loading does not influence the length of the fully-developed fracture process zone, which was estimated to be 9–12 mm
for all specimens. With consideration of maximum grain size of 0.8–1 mm for Berea sandstone, it about 11–14 times larger.
(2) Mode II loading does not influence the maximum value of AE energy, which is estimated to be more or less a constant value, i.e.,
4–6 mV∗μt1/2∗mm2 for all specimens. If it is assumed that a certain portion of the fracture energy is detected by AE, the fracture
energy is also not affected by mode II loading.
(3) Mode II loading causes the fracture process zone to reach fully-developed in the mixed-mode fracture “earlier” than the mode I. In
other words, the fully-developed fracture process zone occurs in mixed-mode fracture before or at peak load, while it occurs in
mode I after the peak.
(4) Mode II loading not only breaks the similarity of AE pattern during fracture processes, but also its symmetry. With more AE
events on one side of fracture path, it results in an asymmetric pattern of AE events for material damage region. However, the
fracture process zone as a core region still maintains its symmetry for two types of fractures. Also, mode II loading results in a
wider material damage region.

Acknowledgement

The authors sincerely thank the National Natural Science Foundation of China (Grant No. 51304225, 51704307), the Science
Foundation of China University of Petroleum, Beijing (No. 2462013YJRC036), and the Foundation of State Key Laboratory of
Petroleum Resources and Prospecting, China University of Petroleum, Beijing (No. PRP/open-1402) for the financial support.

References

[1] Hoagland RG, Hahn GT, Rosenfield AR. Influence of microstructure on fracture propagation in rock. Rock Mech 1973;5:77–106.
[2] Zietlow WK, Labuz JF. Measurement of the intrinsic process zone in rock using acoustic emission. Int J Rock Mech Min Sci Geomech Abstr 1998;35:291–9.
[3] Cotterell B, Rice JR. Slightly curved or kinked cracks. Int J Fract 1980;16(2):155–69.
[4] Ingraffea AR. Theory of crack initiation and propagation in rock. In: Atkinson BK, editor. Fracture mechanics of rock. New York: Academic Press; 1989. p. 71–110.
[5] Chang SH, Lee CI, Jeon S. Measurement of rock fracture toughness under mode I and II and mixed-mode conditions by using disc-type specimens. Eng Geol 2002;66:79–97.
[6] Ayatollahi MR, Aliha MRM, Saghafi H. An improved semi-circular bend specimen for investigating mixed mode brittle fracture. Eng Fract Mech 2011;78:110–23.
[7] Lin Q, Biolzi L, Labuz JF. Opening and mixed-mode fracture initiation in a quasi-brittle material. J Eng Mech-ASCE 2013;139:177–87.
[8] Lin Q, Yuan H, Biolzi L, Labuz JF. Opening and mixed-mode fracture processes in a quasi-brittle material via digital imaging. Eng Fract Mech 2014;131:176–93.
[9] Lockner D. The role of acoustic emission in the study of rock fracture. Int J Rock Mech Min Sci Geomech Abstr 1993;30(7):883–99.
[10] Zang A, Wagner CF, Dresen G. Acoustic emission, microstructure, and damage model of dry and wet sandstone stressed to failure. J Geophy Res 1996;101(B8):17507–21.
[11] Baud P, Meredith PG. Damage accumulation during triaxial creep of darley dale sandstone from pore volumetry and acoustic emission. Int J Rock Mech Min Sci Geomech
Abstr 1997;34:3–4.
[12] Muralidhara S, Prasad BKR, Eskandari H, Karihaloo BL. Fracture process zone size and true energy of concrete using acoustic emission. Constr Build Mater 2010;24:479–86.
[13] Shah KR, Labuz JF. Damage mechanisms in stressed rock from acoustic emission. J Geophys Res 1995;100(B8):15527–39.
[14] Landis EN, Baillon L. Experiments to relate acoustic emission energy to fracture energy of concrete. J Eng Mech-ASCE 2002;128(6):698–702.
[15] Prasad BKR, Sagar RV. Relationship between AE energy and fracture energy of plain concrete beams: experimental study. J Mater Civ Eng-ASCE 2008;20(3):212–20.
[16] Ostanin IA, Mogilevskaya SG, Labuz JF, Napier J. Complex variables BEM for elasticity problems with constant body force. Eng Anal Bound Elem 2011;35:623–30.
[17] Lin Q, Fakhimi A, Haggerty M, Labuz JF. Initiation of tensile and mixed-mode fracture in sandstone. Int J Rock Mech Min Sci 2009;46:489–97.
[18] Hardy Jr HR. Acoustic emission/microseismic activity: volume 1: Principles, techniques and geotechnical applications. 1st ed. The Netherlands: A.A. Balkema Publishers;
2003.
[19] Otsuka K, Date H. Fracture process zone in concrete tension specimen. Eng Fract Mech 2000;65:111–31.
[20] Bažant ZP, Yu Q. Size-effect testing of cohesive fracture parameters and nonuniqueness of work-of-fracture method. J Eng Mech-ASCE 2011;137:580–8.
[21] Carpinteri A, Lacidogna G, Corrado M, Di Battista E. Cracking and crackling in concrete-like materials: a dynamic energy balance. Eng Fract Mech 2016;155:130–44.

144