See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/324299521

Fracture behaviour of cemented tailing backfill with pre-existing crack and

thermal treatment under three-point bending loading: Experimental studies
and particle flow code simulat...

Article  in  Engineering Fracture Mechanics · April 2018

DOI: 10.1016/j.engfracmech.2018.04.008

CITATIONS READS

9 228

2 authors, including:

Xu Wenbin
China University of Mining & Technology, Beijing
26 PUBLICATIONS   146 CITATIONS   

SEE PROFILE

All content following this page was uploaded by Xu Wenbin on 15 May 2018.

The user has requested enhancement of the downloaded file.

 Engineering Fracture Mechanics 195 (2018) 129–141

Contents lists available at ScienceDirect

Engineering Fracture Mechanics

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

Fracture behaviour of cemented tailing backfill with pre-existing

T
crack and thermal treatment under three-point bending loading:
Experimental studies and particle flow code simulation
⁎
Wenbin Xua,b,c, , Peiwang Caoa
a
College of Resources & Safety Engineering, China University of Mining &Technology (Beijing), Beijing 100083, PR China
b
Key Laboratory of Ministry of Education of China for Efficient Mining and safety of Metal Mines, University of Science and Technology Beijing,
Beijing 100083, PR China
c
Department of Civil Engineering, University of Ottawa, Ottawa, Ontario K1N 6N5, Canada

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

Keywords: The three-point bending test was performed on cemented tailing backfill prepared with different
Cemented tailing backfill offset notches by a loading system with high-speed camera. And the effect of heated temperature
Three-point bending on CTB fracture behavior has been investigated. These results indicate that the influence of pre-
Fracture characteristic existing notch and temperature on CTB fracture feature is obvious. The peak load increases
Crack propagation
linearly with the increasing of the offset ratio when the depth-height ratio of notch is fixed at the
same value, while the fracture peak loading decreases sharply with the increase of the notch
depth when the offset ratios are the same level. The fracture toughness decreases by 63.5% as the
heated temperature rises from 20 °C to 120 °C. The fracture toughness is directly proportional to
the porosity of CTB. The increasing of porosity leads to the decrease of the fracture toughness
mainly due to the decomposition of hydration products and the coarsening of the pore structure
of CTBs. The crack propagates along the interface among tailing particles, resulting in a rough
and jagged edge propagation path. The rack propagation pattern of CTB samples contains
transgranular fracture and intergranular fracture. The intergranular fracture is clearly pre-
dominant during crack propagation. The rack propagation and failure pattern of CTB specimens
from numerical modeling are conformed to the failure pattern of samples obtained by experi-
mental testing. PFC is an effective numerical analysis method for exploring the fracture me-
chanism of micro-cracking in CTB.

1. Introduction

The introduction of underhand cut and fill mining (UCFM) in 1980s made hard and fractured rock mining in hazardous ground
conditions safer. Over the past decades, UCFM method has been widely used all around world in view of the advantages of effective
ground control, environmental surface disposal of mine solid waste and high ore recovery [1–4]. Underhand mining uses cemented
tailings for backfilling the mined-out stope cut, and makes the mining operations in the following cuts safer, because the personnel
are always working beneath the backfill, which will not collapse during mining process [5]. As mining depths increases, the stress of
surrounding rock increases, and the ground conditions in stope cut become more challenging. In addition, based on an average
geothermal gradient (3 °C/100 m) of crustal rocks, the highest temperature may reach 60 °C at the depth of 2000 m without

⁎
Corresponding author at: College of Resources & Safety Engineering, China University of Mining &Technology (Beijing), Beijing 100083, PR China
E-mail address: xuwb08@163.com (W. Xu).

https://doi.org/10.1016/j.engfracmech.2018.04.008
Received 6 December 2017; Received in revised form 29 March 2018; Accepted 6 April 2018
Available online 07 April 2018
0013-7944/ © 2018 Elsevier Ltd. All rights reserved.
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

Fig. 1. Schematic diagram of the underhand cut and fill mining.

consideration of heat from exothermic ore deposit. Therefore, the temperature becomes another important factor affecting the sta-
bility of the backfilling body as the increasing of mining depth [6,7]. The schematic diagram of underhand cut and fill mining is
shown in Fig. 1.
A key issue in underhand cut and fill mining stope cut is the overall stability of the overlying backfill, and the quality of backfill
roof is required to develop enough strength to protect miners working under it [5]. As one of man-made structure materials, ce-
mented tailing backfill (CTB) in UCFM stoping composed of cement, tailings matrix and voids, is a complex heterogeneous material
and exhibits complex mechanical behaviors [8,9]. Bayram E. et al. [10]’s experimental results have indicated that the influence of
binder type and dosage on the mechanical properties and microstructure of cemented tailing backfill are significant. M. Fall has
presented a design method for mix proportioning of CTB to minimize the number of trial mixes and provided an appropriate mix
proportion [11]. However, most of the researches are focusing on the compositions of the material itself and relatively little studies
are conducted on mechanical properties or fracture characteristics of CTB. As fundamental problems in mining engineering, micro-
crack or damage produced by frequently mining disturbance and blasting vibration are primary source of degradation of cemented
backfill especially in underhand cut and fill mining stope and can significantly reduce the backfill service capacity, so that the crack
and fracture is a main type of distresses. Consequently, the cracking propagation mechanism and fracture features of cemented
backfill have draw more and more attention in recent years.
The mechanical behaviors of overlying CTB in UCFM stoping are significantly affected by stope geometry design, consistent fill
quality, loading conditions, seismic effects (blasting vibration and mining disturbance), and support placement as well as other
factors [12]. Han et al. [13] have reported that the riskiest position is located in the middle of two braced forces in the lower surface
of the bearing stratum. CTB is an artificial engineered mixture, and gains the mechanical through hydration reaction, has the similar
structure to rock or concrete materials [14]. Therefore some results and methods of fracture from the rock and concrete studies can be
resorted to analyze the fracture characteristic of CTB. Zuo et al. [15–17] have obtained that the fracture characteristics, the mi-
crostructure change and the crack propagation model of the granite after the heat treatment, and the crack propagation char-
acteristics of the different bias cracked basalt have also been revealed. Most studies show that the whole process of crack propagation
is important for the fracture characteristics of rock [18–20]. Besides, the particle flow code, as a kind of discrete element numerical
simulation software, has been widely used to analyze the micro mechanism of rock. These studies show that there is a potential link
between the PFC meso-scale strength parameter and the fracture toughness of the rock [21]. The fracture toughness of CTB depends
on solid content, binder-to-tailing ratio and curing age [22]. Its fracture characteristics are fundamentally different from those of rock
and concrete materials. Therefore, it is necessary to study the fracture mechanics behavior of CTB. However, as of the present, little
studies have conducted the fully investigation of fracture characteristic of CTB in UCFM stope. Consequently, a solid knowledge of the
crack propagation and fracture characteristic is vital for the rational design, and a safe and efficient analysis of the stability of CTB in
UCFM stope. Therefore, to better understand the crack propagation that occurs in the CTB. The above authors will present and discuss
some tests to understand the characteristic of CTB samples prepared with different offset notch (different offset ratio and seam-height
ratio) and thermal treatments during the whole process of the three-point bending testing. In addition, we will simulate the crack

130
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

Table 1
Chemical composition and content of used tailing.
Composition MgO Al2O3 SiO2 CaO Na2O MnO Fe2O3 Total

Content (%) 2.41 3.85 82.05 2.46 0.18 0.02 8.01 98.98

propagation, fracture pattern and fracture behavior of CTB by particle flow code simulation software to prove the reliability of
experimental results. The research results can provide reference for the stability and safety assessment of the overlying CTB in UCFM
stope.

2. Experimental program

2.1. Materials

The tailings in this present study were obtained from an iron mine, located in Hefei city, Anhui Province, southeast of China.
Ordinary Portland cement is used as a binder. The composition of the used tailing was shown in Table 1, it can be noted that the
tailing belong to acidic tailing because the silica content is much more than 82.05 wt% [23]. The size distribution of the tailing was
determined by the LS-C (Ⅱ) laser particle size analyser, as shown in Fig. 2. The coefficients of uniformity (Cu) and curvature (Cc) are
3.636 and 1.657, respectively.

2.2. Sample preparation

The tailings, binder and water were mixed and homogenized to produce the designed CTB mixtures. In this process, the ratio of
cement to tailing is 1:4 and the solid content is 75% by weight. The samples were removed from the mold after 24 h, and then sealed
and cured in a chamber maintained at approximately 90% humidity and 20 ± 1 °C for periods of 7 days. The size of rectangle sample
with an offset-notch are shown in Fig. 3(a), the different notch locations (b) of a notch deviated from the bending center line (b = 0)
are 0 mm, 20 mm, 40 mm and 60 mm respectively. Every artificial notch radius is about 0.4 mm width, the depth of notch is set to
4 cm, 10 mm and 20 mm, respectively. So the offset ratio (β = b /100) are 0, 0.25, 0.5, 0.75 and the depth-height ratio of notch
(α = a/40) were 0.1, 0.25, 0.5, respectively. The samples are divided into three groups according to the different notch depths as
shown in Table 2. To investigate the effect of temperature on the CTB, the samples with a 10 mm deep notch and no offset were
subjected to 40 °C, 60 °C, 80 °C, 100 °C and 120 °C thermal treatment respectively before conducting three-point bending testing.

2.3. Experimental procedure

CTB samples were heated in a high temperature furnace, and the temperature inside the furnace was raised at a rate of 3–4 °C/
min and kept for 3 h until it reached the assigned value, then naturally cooled down to the room temperature (20 °C). The samples
with different notch offset ratios were tested under the room temperature.
All three-point bending tests were performed at a rate of 5 × 10−4 mm/s which was the minimum loading speed of the system.
Three fulcrums were coated with lubricant to reduce friction before loading. The high speed camera selected 3900 frames per second
with a resolution of 640 × 512. The process of crack initiation and propagation during loading was captured by high speed camera,
and the load-time information was recorded by the high speed camera-load system, as shown in Fig. 3(b).

Fig. 2. Particle size distribution of the used tailings.

131
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

(a) Sample size (b) Sketch of three-point bending test

Fig. 3. Schematic drawing of samples and testing system (Unit: mm).

Table 2
The sample numbers and notch geometrical parameter values for all specimens.
Sample ID Offset/(mm) Offset ratio Notch depth /(mm) Depth-to-height ratio Temperature/(°C)

C0-4 0 0 4 0.1 20
C20-4 20 0.25
C40-4 40 0.50
C60-4 60 0.75

C0-10 0 0 10 0.25 20
C20-10 20 0.25
C40-10 40 0.50
C60-10 60 0.75

C0-20 0 0 20 0.5 20
C20-20 20 0.25
C40-20 40 0.50
C60-20 60 0.75

C0-10-40 0 0 10 0.25 40
C0-10-60 0 0 10 0.25 60
C0-10-80 0 0 10 0.25 80
C0-10-100 0 0 10 0.25 100
C0-10-120 0 0 10 0.25 120

3. Results and discussion

3.1. Offset-induced fracture characteristic of CTB

The fracture toughness KIC can be calculated by formula as following in accordance with the ASTM [24].
Pmax S a
KIC = f ( )
WH1.5 I H (1)

a a 0.5 a 1.5 a 2.5 a 3.5 a 4.5

fI ⎛ ⎞ = 2.9 ⎛ ⎞ −4.6 ⎛ ⎞ + 21.8 ⎛ ⎞ −37.6 ⎛ ⎞ + 38.7 ⎛ ⎞
⎝H⎠ ⎝H⎠ ⎝H⎠ ⎝H⎠ ⎝H⎠ ⎝H⎠ (2)

b
β=
L/2 (3)
a
α=
H (4)

Where Pmax is the peak loading, S is the effective span, W, L and H are respectively the sample width, Length and height, a, b is the
artificial notch depth and offset, α , β is the notch depth to height ratio and offset ratio.
The fracture parameters of CTB specimens with different notches are obtained as shown in Fig. 4, it can be seen in Fig. 4 that the
peak loading curves of CTB samples follow a similar development pattern with the increasing of notch offset ratio. The fracture peak
loadings of CTB samples are directly proportional to the offset ratios. After fitting the fracture of CTB with notch offset ratio, it is
apparent that the relationship between fracture peak loading and offset ratio follows a linear law with different slopes of 159.7, 165.7
and 193.4, irrespectively. The peak loading curve of CTBs with a deeper notch is always above those of smaller notch depth. It means
that the depth of notch affects the fracture strength. The deeper the notch is, the smaller the fracture peak loading is. To analyze the
effect of the notch depth on the fracture strength, the notch offset is fixed at the same level, the fracture peak loading exponent
decreases sharply with increase of the notch depth, as illustrated in Fig. 5. It implies that the fracture behavior of CTB is not only

132
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

650 C20-4
C40-4
C60-4 y = 159.7x + 399.1
550 C20-4 fitted curve R² = 0.962
C40-4 fitted curve
C60-4 fitted curve
450
y = 165.7x + 282.6
R² = 0.962

Pmax /KN
350

250
y = 193.4x + 185.2
R² = 0.956
150

50
0 0.25 0.5 0.75
Offset ratio
Fig. 4. Relationship between fracture peak loading and offset ratio of notch.

600 C0-a C20-a

C40-a C60-a
C0-a fitted curve C20-a fitted curve
500
C40-a fitted curve C60-a fitted curve

400
Pmax/KN

y = 766.1x-0.26
300 R² = 0.999

200 y = 746.2x -0.38

R² = 0.985
y = 778.5x-0.35
R² = 0.995 y = 789.5x -0.46
100 R² = 0.985

0
0 4 8 12 16 20 24
Notch depth /mm
Fig. 5. Relationship between fracture peak loading and notch depth.

dependent on the offset ratio but also dependent on the notch depth when the applied stress is in a bending state.
To quantify the effect of notch offset on the fracture loading of CTB accurately, the general relationship between the fracture peak
loading and the offset ratio, depth of the notch can be fitted as following, irrespectively:
Pmax = mβ + n (5)

Pmax = ke−a (6)

where Pmax is the peak loading, m, n, k are the fitted coefficient, and β , a are the offset ratio and depth of the notch.

3.2. Temperature-induced fracture characteristic of CTB

As we all known, the temperature is an important factor affecting the fracture strength of CTB in deep mine, to investigate the
effect of temperature on pore structure and fracture strength of CTB, the CTB samples with a 10 mm deep notch were subjected to
40 °C, 60 °C, 80 °C, 100 °C and 120 °C thermal treatment, irrespectively, before conducting three-point bending testing. The main
results of the effect of thermal treatment on the pore structure of CTBs are presented in Figs. 6 and 7. Fig. 6 shows the effect of
temperature on microstructure morphology and pore distribution of the cemented tailing backfill. From SEM micrographs in the
Fig. 6, it indicates that thermal treatment has a significant effect on the pore structure distribution within the CTB and a great number
of micocracks and pores appeared on the surface and dispersed in the matrix of the samples after different thermal treatments. A
higher thermal treatment means that the pore structure of CTBs becomes coarser. The porosity curve for the CTB samples, not heated
and heated at 40 °C, 60 °C, 80 °C, 100 °C and 120 °C, presented in Fig. 7, demonstrates that it leads to an obvious coarsening of the
pore structure of the CTBs when the heated temperature rises up to 120 °C. It can be illustrated that there is a small decrease in the
porosity with increased temperature (up to 40 °C). This means, up to this temperature, there is no important variation of the capillary
pore size within the heated CTB, and the temperature maybe further accelerate the cement hydration. When the heated temperature
exceeds the 40 °C, the surfaces of heated CTBs appear obviously rougher and the porosity of CTB samples increases sharply. Some
pores and cracks clearly connect with each other, facilitating the formation of the pores and fracture networks and decreasing in
fracture strength. The fracture toughness decreased by 30.1% when the heat treatment temperature rose from 80 °C to 120 °C. It is
attributed to the non-uniform expansion between the tailing grains and cement hydration products. When the heated temperature
was elevated from 40 °C to 120 °C, the loss in the content of water bound in hydration products (e.g. C-S-H gel) resulted in voids
formation. In addition, the cracks were formed within the tailing matrix due to the thermal mismatch of the hydration products,

133
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

ºC ºC

ºC ºC
Fig. 6. The SEM pictures of backfill samples after different thermal treatment: (a) 20 °C; (b) 40 °C (c) 80 °C; (d) 120 °C.

Fig.7. The curves of porosity and fracture toughness of backfill samples with different thermal treatment.

because the hydration products consists of various types of hydrated substances (e.g. C-S-H gel, ettringite, CaO·SO2·H2O) that have
different dehydration temperatures and rates, and also different levels of thermal expansion/shrinkage with increasing temperatures
[7,25].
The analysis of the aforementioned results of the influence of the thermal treatment on the fracture strength (Fig. 7) and pore
structure of CTB (Figs. 6 and 7) shows that there is a relationship between the porosity and fracture strength evolution of CTB after
heated by thermal treatment. The fracture toughness is directly proportional to the porosity rise. The decrease of the porosity and the
rise of fracture strength below 40 °C may be mainly attributed to water driven out and additional binder hydration. While the
increasing of porosity leads to the decrease of the fracture toughness due mainly to the decomposition of hydration products and the
coarsening of the pore structure of CTBs. So it can infer that the increasing temperature have two opposite influence on the fracture
strength, namely, strength-improving factor and strength-deteriorating factor.

134
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

400
Fig.9(e)
Fig.9(d)
350

300 Fig.9(c)

Loading/ kN
250

200

150

100
Fig.9(b)
50
Fig.9(f)
0
0 0.1 0.2 0.3 0.4 0.5 0.6 0.7 0.8 0.9 1 1.1
Time /s
Fig. 8. Loading-time curve of C40-10 CTB sample.

3.3. Real-time observation on crack propagation and fracture mechanism of CTB

A typical loading-time curve of CTB with a pre-existing notch under three-point bending testing is shown in Fig. 8. Some rack
initiation, propagation and failure process images of CTB sample are captured at different loading level by high speed camera, as
shown in Fig. 9. Fig. 9(a) is the original image of CTB at unloading state. The CTB sample is subjected to an external loading F on the
top. The other two support points are only schematic illumination since the actual positions cannot be focused within the range of
camera. On the basis of fracture theory, the pre-existing notch will propagate when the stress concentration around the gap exceed
the fracture toughness [26]. When the loading stress reaches 50 kN (Loading time is equivalently 250 ms), the pre-existing notch
starts to propagate, as shown in Fig. 9(c). It clearly demonstrates that the micro-cracks initiate from hydration products among the
coarse tailing particles. The main reason is that the tailing particles are almost wrapped by hydration productions and the bonding
strength and hardness produced by hydration reaction are far from the tailing grains. The loading stress cannot break the tailing grain
because of the adhesive among these tailing particles. Therefore, the crack propagates along the interface among tailing particles,
resulting in a rough and jagged edge propagation path. As the external loading increases, the micro-cracks gradually propagate and
coalescence. Fig. 9(c) and (d) illustrates that the crack propagate kinking in the matrix and the bypass the tailing particles. Cracks
occur along or between these tailing particles. And the deviated direction becomes more and more closed to the up loading point with
the increasing of loading time. When the loading reaches the peak value, the sample still has loading capacity, even the gap width
gets bigger and bigger, as shown in Fig. 9(e). As the loading time increases to 1020 ms, the CTB sample fully fractures. A macro-crack
open obviously appears in the sample, as shown in Fig. 9(f). The fracture failure process of CTB indicates that the failure pattern of
the intergranular fracture is the main fracture mechanism of CTB sample under three-point bending state.
For C0-10 sample with a non-offset notch, when the loading increases to 36 kN, a crack appears along the vertical direction, and
multi-deviated phenomena at the crack tips are shown in Fig. 10(a) and (b). It is because that the pre-existing notch located at the
middle of sample, namely, the offset of pre-crack is zero. That means that the offset of a notch degrade the fracture strength and
accelerate the crack propagation of CTB, which is also proved in the above results. With an increasing in loading value to peak value
approximate about 372 kN, the sample fractures with an open macro-crack.
To better show the effect of different offset on the behavior of crack propagation at the three-point bending stress state, deviated
angle θ was employed in Figs. 9 and 10. It indicates that the crack propagation path with different offset-notch has significantly
changed. With an increasing of offset ratio, the deviated angle increases. However, when the offset ratio is 0.75, the crack angle
becomes smaller. It is because that the crack initiates from the middle of sample. It also means that the crack offset can affect the
crack propagation and fracture path. The maximum deviated angle is about 25° for the offset ratio is set to 0.5.
In this section, the rack propagation process and pattern were respectively described. It was confirmed that the rack propagation
pattern of CTB samples contains transgranular fracture and intergranular fracture. And the intergranular fracture is clearly pre-
dominant way during crack propagation, showing the crack propagation pattern of the CTB is tensile fracture under three-point
bending, which is opposite to the rock failure behavior [27].

3.4. Effects of pre-existing notch on fracture behavior based on PFC analyses

In this section, the discrete element method (particle flow code: PFC 2D) is used to simulate the crack propagation and mode of
CTB samples prepared with different pre-existing notch locations under three points bending test. The PFC is suitable for presenting
grain-scale microstructural characters by taking tailing grain into CTB consideration as a discrete particle [28]. As illustrated in
Fig. 11, to be consistent with the lab tests, the CTB model in PFC simulation software, which is 200 mm in length and 40 mm in
height, is applied to simulate fracture features by the three point bending. The CTB model is built by the assembly of tailing particles
bonded with each other. A certain number of particles are randomly generated in the wall according to the radius and properties of
the particles, and then the internal stress is balanced by the radius adjustment method [29,30]. Sample size of model generated in this
simulation is consistent with the experimental size. Radius of particles follows Gaussian distribution that minimum radius is 0.2 mm.
The micro-scale properties of the numerical sample model are calibrated and shown in Table 4. Fig. 12 shows the numerical crack

135
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

(a) F=0 kN, s=0 ms (b) F=0 kN, s=250 ms

(c) F=252 kN, s=700 ms (d) F=345 kN, s=950 ms

(e) F=372kN, s=1000 ms (f) F=15 kN, s=1020 ms

Fig. 9. Whole failure process images of C40-10 sample at different loading state: (a) s = 0 ms; (b) s = 250 ms; (c) s = 770 ms; (d) s = 950 ms; (e)
s = 1000 ms; (f) s = 1020 ms.

136
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

(a) F=36 kN,s=230 ms (b) F= 156kN,s=640 ms

(c) F=287 kN, s=972ms (d) F= 12kN,s=1011ms

Fig. 10. Whole failure process images of C0-10 sample at different loading state: (a) s = 230 ms; (b) s = 640 ms; (c) s = 972 ms; (d) s = 1011 ms.

Fig. 11. Numerical simulation model of the C40-10 sample.

Table 3
The deviated angle θ values of CTB samples.
Sample ID Offset ratio θ

C0-4 0 2°
C20-4 0.25 11°
C40-4 0.50 15°
C60-4 0.75 5°
C0-10 0 70
C20-10 0.25 20°
C40-10 0.50 25°
C60-10 0.75 3°
C0-20 0 5°
C20-20 0.25 15°
C40-20 0.50 23°
C60-20 0.75 3°

137
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

Table 4
Input micro-scale parameters for modeling of CTB.
Lower bound of particle radius (mm) Rmin 0.2
Ratio of particle radius Rmin/Rmax 3
Particle density (kg/m3) ρ 2700
Ratio of normal to shear stiffness of the particle kn/ks 2.5
Yong’s modulus of the particle (GPa) Ec 1.1
Friction coefficient of particle μ 0.75
Ratio of normal to shear stiffness of the parallel bond kn / ks 2.5
Shear strength of the parallel bond (MPa) τc 15.0 ± 2.5
Tensile strength of the parallel bond (MPa) σt 10.0 ± 1.5

propagation of C40-10 sample under the same three-point bending as experiment used. It can be seen that the crack propagation
during the simulation are highly in accordance with the experimental results as described above. The crack initiates at the tip of pre-
existing notch, and the propagation direction extend toward up loading point. Then the crack propagates gradually with the loading
before failure. The crack propagation velocity profiles are visualized directly during simulation at pre-peak, peak and post-peak
stages, as shown in Fig. 13. The rack propagation and failure pattern of CTB specimens from numerical modeling are conformed to
the failure pattern of samples with different offsets (Fig. 14). For high offset specimen (β = 0.75) under three-point bending, the rack
propagates at the middle of sample, and it is unlikely to occur that cracks initiates at the tip of the pre-existing notch and develops
along a deviated angle to the up loading point position. The crack propagation angles in the Table 3 can verify the results. On the

(a) s=0 ms

(b) s=250 ms

(c) s=700 ms

(d) s=950 ms

(e) s=1000 ms

(f) s=1020
Fig. 12. The crack propagation process of C40-10 sample. (a) s = 0 ms; (b) s = 250 ms ; (c) s = 700 ms; (d) s = 950 ms; (e) s = 1000 ms; (f)
s = 1020 ms.

138
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

(a) s=250 ms

(b) s=950 ms

(c) s=1025 ms
Fig. 13. Variations of velocity profile of C40-10 sample during crack propagation process. (a) s = 250 ms; (b) s = 950 ms; (c) s = 1025 ms

(a) C0-4 sample

(b) C20-4 sample

(c) C40-4 sample

(d) C60-4 sample

Fig. 14. Numerical and experimental results of the rack paths for different offset notch samples. (a) C0-4 sample; (b) C20-4 sample; (c) C40-4
sample; (d) C60-4 sample

139
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

600

550

500

Pmax/KN
450

400 experiment
simulation
350

300
0 0.25 0.5 0.75
Offset ratio
Fig. 15. Peak loading contrast of CTB samples with different offsets.

whole, simulation results show that the peak loading is consistent with the experimental results, and the average error is less than
3.8% as shown in Fig. 15, which indicates that PFC is an effective numerical analysis method for exploring the fracture mechanism of
micro-cracking in CTB.

4. Conclusions

In this study, three-point bending testing were performed on the CTB specimens with different offsets notch (different offset ratio
and depth-height ratio) and thermal treatments (at 40 °C, 60 °C, 80 °C, 100° and 120 °C) under a high-speed camera-loading system.
The effect of notch on crack propagation and pattern were respectively discussed. Main conclusions obtained in the study are as
following:

(1) It is obvious to see the effect of offset and thermal on the crack propagation and fracture characteristic of CTB under three-point
bending. The peak load increases linearly with the increasing of the offset ratio when the depth-height ratio of the samples is fixed
at the same value, while the fracture peak loading exponent decreases sharply with increase of the notch depth, when the offset
ratios are the same level. The relationship between fracture peak loading and offset ratio follows a linear law.
(2) Increasing the heated temperature up to 120 °C leads to an obvious coarsening of the pore structure of the CTBs. The fracture
toughness decreases by 63.5% when the heated temperature rises from 20 °C to 120 °C. There is a relationship between the
porosity and fracture strength evolution of CTB after heated by thermal treatment. The fracture toughness is directly proportional
to the porosity rise. The increasing of porosity leads to the decrease of the fracture toughness due mainly to the decomposition of
hydration products and the coarsening of the pore structure of CTBs.
(3) The crack offset can affect the crack propagation and fracture path, with an increasing of offset ratio, the deviated angle increases.
The crack propagates along the interface among tailing particles, resulting in a rough and jagged edge propagation path. The rack
propagation pattern of CTB samples contains transgranular fracture and intergranular fracture. And the intergranular fracture is
clearly predominant way during crack propagation,
(4) The crack propagation velocity profiles are visualized directly during simulation at pre-peak, peak and post-peak stages. The rack
propagation and failure pattern of CTB specimens from numerical modeling are conformed to the failure pattern of samples with
different offsets by experimental testing. PFC is an effective numerical analysis method for exploring the fracture mechanism of
micro-cracking in CTB.

Acknowledgement

The first author would like to acknowledge China Scholarship Council as financial support for my visiting work at University of
Ottawa. This work was supported by the Natural Science Foundation of China (grant number: 51504256); the Open fund of the Key
Laboratory of Ministry of Education for Efficient Mining and Safety of Metal Mines (grant number: USTBMSLAB201702; and
Fundamental Research Funds for the Central Universities (grant number: 2014QZ02).

References

[1] Jiang H, Cao Y, Huang P, Fang K, Li B. Characterisation of coal-mine waste in solid backfill mining in China. Mining Technol 2015;124(1):56–63.
[2] Huang YL, Zhang JX, Zhang Q. Backfilling technology of substituting waste and fly ash for coal underground in China coal mining area. Environ Eng Manage J
2011;10(6):769–75.
[3] Zhang JX, Miao XX, Guo GL. Development status of backfilling technology using raw waste in coal mining. J. Mining Safety Eng. 2009;26(4):395–401.
[4] Song ZJ, Guo DW, Chen Q, Gu ZK. Research on the solid wastes utilization of shouyun iron mine. Int. J. Earth Sci. Eng. 2014;7(2):658–62.
[5] Ayhan K, Erol Y, Bayram E, Ibrahim A, Haci D. Effect of properties of tailings and binder on the short-and long-term strength and stability of cemented paste
backfill. Mater Lett 2005;59:3703–9.
[6] Fall M, Célestin JC, Pokharel M, Touré M. A contribution to understanding the effects of curing temperature on the mechanical properties of mine cemented
tailings backfill. Eng Geol 2010;114:397–413.
[7] Fall M, Samb SS. Effect of high temperature on strength and microstructural properties of cemented paste backfill. Fire Saf J 2009;44:642–51.

140
W. Xu, P. Cao Engineering Fracture Mechanics 195 (2018) 129–141

[8] Fall M, Pokharel M. Coupled effects of sulphate and temperature on the strength development of cemented tailings backfills. Cem Concr Compos
2010;32:819–28.
[9] Nasir O, Fall M. Coupling binder hydration, temperature and compressive strength development of underground cemented paste backfill at early ages. Tunn
Undergr Space Technol 2010;25:9–20.
[10] Bayram E, Ayhan K, Ferdi C, Haci D, Ibrahim A. Cemented paste backfill of sulphide-rich tailings: importance of binder type and dosage. Cem Concr Compos
2009;31:268–74.
[11] Fall M, Benzaazoua M, Saa EG. Mix proportioning of underground cemented tailings backfill. Tunn Undergr Space Technol 2008;23:80–90.
[12] Pakalnis R, Caceres C, Clapp K, Morin M, Brady T, Williams T et al. Design spans-underhand cut and fill mining. In: Proceedings, 2005 CIM-AGM Annual
Meeting, Toronto, Ontario, Canada, 2005, 9.
[13] Han B, Zhang SX, Deng J, Wang XL. Determining method of backfill strength for underhand drift cut-and-fill Stopping based on reliability theory. J. China Univ.
Min. Technol. 2006;35(3):372–6.
[14] Benzaazoua M, Fall M, Belem T. A contribution to understanding the hardening process of cemented paste fill. Min Eng 2004;17(2):141–52.
[15] Zuo JP, Wang JT, Sun YJ, Chen Y, Jiang GH, Li YH. Effects of thermal treatment on fracture characteristics of granite from Beishan, a possible high-level
radioactive waste disposal site in China. Eng Fract Mech 2017;182:425–37.
[16] Zuo JP, Wang XS, Mao DQ. SEM in-situ study on the effect of offset-notch on basalt cracking behavior under three-point bending load. Eng Fract Mech
2014;131:504–13.
[17] Zuo JP, Wang XS, Mao DQ, Wang CL, Jiang GH. T-M coupled effects on cracking behaviors and reliability analysis of double-notched crustal rocks. Eng Fract
Mech 2016;158:106–15.
[18] Ali T, Derrick JB, Ali F, Joseph FL. Fracture initiation in cavity expansion of rock. Int J Rock Mech Min Sci 2016;85:84–91.
[19] Dong W, Wu ZM, Zhou XM, Wang N, Kastiukas G. An experimental study on crack propagation at rock-concrete interface using digital image correlation
technique. Eng Fract Mech 2017;171:50–63.
[20] Nicolas M, Ted B. Extended finite element method for cohesive crack growth. Eng Fract Mech 2002;69:813–33.
[21] Jeoungseok Y. Application of experimental design and optimization to PFC model calibration in uniaxial compression simulation. Int J Rock Mech Min Sci
2007;44(6):871–89.
[22] Xu WB, Tian XC, Cao PW. Assessment of hydration process and mechanical properties of cemented paste backfill by electrical resistivity measurement.
Nondestruct Test Evaluat 2018;33(2):198–212.
[23] Héctor MC, Brett HR, Rainer S, Bernd N. Metal extractability in acidic and neutral mine tailings from the Cartagena-La Unión Mining District (SE Spain). Appl
Geochem 2008;23:1232–40.
[24] Ding SD. Fracture mechanics. Beijing: China Machine Press 1997:30p. (in Chinese).
[25] Gallé C., Sercombe J., Permeability and pore structure evolution of silico-calcareous and hematite high-strength concretes submitted to high temperatures.
Mater. Struct. 2001; 34: 619–28.
[26] Broek D. Elementary Engineering Fracture Mechanics. Dordrecht: Martinus Nijhoff; 1982. p. 90p.
[27] Zuo JP, Xie HP, Dai F, Ju Y. Three-point bending test investigation of the fracture behavior of siltstone after thermal treatment. Int J Rock Mech Min Sci
2014;70:133–43.
[28] Jia L, Chen M, Zhang W, Xu T, Zhou Y, Hou B, et al. Experimental study and numerical modeling of brittle fracture of carbonate rock under uniaxial compression
2013;50:58–62.
[29] Zhou J, Zhang LQ, Pan ZJ. Numerical investigation of fluid-driven near-borehole fracture propagation in laminated reservoir rock using PFC2D. J Nat Gas Sci Eng
2016;36:719–33.
[30] Chen J, Wang LB, Huang XM. Micromechanical modeling of asphalt concrete fracture using a user-defined three-dimensional discrete element method. J Cent
South Univ 2012;19:3595–602.

141

View publication stats