Rock Mechanics and Rock Engineering

https://doi.org/10.1007/s00603-022-02772-4

ORIGINAL PAPER

Effect of Confining Pressure and Strain Rate on Mechanical Behaviors

and Failure Characteristics of Sandstone Containing a Pre‑existing
Flaw
Wei You1 · Feng Dai1 · Yi Liu1 · Zelin Yan1

Received: 22 July 2021 / Accepted: 31 December 2021

© The Author(s), under exclusive licence to Springer-Verlag GmbH Austria, part of Springer Nature 2022

Abstract
In underground engineering constructions, fractured rock masses are likely to be subjected to three-dimensional static stress
and dynamic loading simultaneously. In this study, via a modified split Hopkinson pressure bar (SHPB) system, a series
of triaxial dynamic tests are carried on sandstone containing a pre-existing flaw under different radial confining pressures
ranging from 5.7 to 22.7 MPa and different strain rates varying by 85–252 ­s−1. The results indicate that both the dynamic
strength and energy dissipation density of single-flawed sandstone feature a positive correlation with the rising radial con-
fining pressure and strain rate; while the dynamic elastic modulus, failure strain and energy utilization efficiency all exhibit
an evident radial confining pressure enhancement effect, these are insensitive to the strain rate. Moreover, the fragment
distributions of single-flawed sandstone are adequately characterized by combining sieving tests and generalized extreme
value function. Under a higher strain rate, the single-flawed specimen is featured by smaller average fragment size and wider
fragment distribution range, while the higher confining pressure has the opposite effect. By post-mortem examination, three
main failure types are classified, and the fracturing mechanism of new cracks is revealed by microscopic observation using
a scanning electron microscope (SEM). The macro oblique shear crack and through shear crack are mainly induced by the
shear slip mechanism, while the quasi-coplanar crack is tensile dominated.

Highlights

• Conducted a series of triaxial SHPB tests on the three- • Revealed the fracturing mechanism of new cracks from a
dimensional confined sandstone containing a pre-existing microscopic perspective by the scanning electron micro-
flaw. scope.
• Investigated the radial confining pressure and strain rate
effect on the dynamic mechanical response and failure
characteristics of flawed rocks.

Keywords Confining pressure · Dynamic mechanical response · Single-flawed rock · Fracturing characteristics ·
Fragmentation distribution

1 Introduction

With the rapid growth of the demand for space and

* Feng Dai resources, quantities of engineering structures are con-
fengdai@scu.edu.cn structed at different depths underground, including air-raid
1
State Key Laboratory of Hydraulics and Mountain River shelter, hydropower house, and geothermal mining, as sche-
Engineering, College of Water Resource and Hydropower, matically illustrated in Fig. 1a. Rock masses are initially
Sichuan University, Chengdu, Sichuan 610065, subjected to three-dimensional static stress which increases
People’s Republic of China

13
Vol.:(0123456789)
 W. You et al.

Fig. 1  Fractured rock masses

confronted with three-dimen-
sional static stress and dynamic
loading in various underground
engineering scenario

with the earth depth in underground engineering (Hoek engineering constructions are likely to be confronted with
and Brown 1980). However, except the three-dimensional static stress and dynamic disturbance simultaneously (Liu
static stress, rock masses in engineering construction are et al. 2021), the coupled static–dynamic loading on flawed
usually confronted with dynamic disturbance induced by rocks has also been performed (Feng et al. 2018; Xiao et al.
blasting, drilling, seismic events or earthquakes (Zhang and 2020; Yan et al. 2020). Feng et al. (2018) revealed that the
Zhao 2014; Xia and Yao 2015; Liu et al. 2018; Jiang et al. coupled static–dynamic stress evidently affects the strength
2021; Liu and Dai 2021), as outlined in Fig. 1b. It is thus and failure mode of multi-flawed specimens. The strength,
crucial to investigate the dynamic mechanical responses of elastic modulus, and released elastic energy of specimens
rocks under three-dimensional static stress, especially for feature an increase first, then decrease as the static pre-stress
the flawed rocks that contain many discontinuous fissures increases. Yan et al. (2020) reported that the failure modes
and widely exist in various engineering structures (Hoek and of single-flawed specimens changed from a shear-domi-
Bieniawski 1984; Wong and Chau 1998; Wong and Einstein nated “X” shape failure under individual dynamic loading
2009; Wei et al. 2021). to mixed shear–tensile failure under coupled static–dynamic
Up to now, investigations of flawed rock are mainly con- loading. It should be noted that these studies are limited to
centrated on the pure static or dynamic loading conditions. one-dimensional static stress conditions.
Numerous static or dynamic uniaxial compression tests have For flawed rock under three-dimensional stress condi-
been conducted on rocks containing single flaw (Bobet and tions, some static experiments have been conducted (Huang
Einstein 1998; Li et al. 2005, 2017; Yang and Jing 2011; Zou et al. 2016a and b; Yang et al. 2021). Huang et al. (2016b)
and Wong 2014; Zou et al. 2016), double flaws (Wong and showed that the crack evolution and failure mode are mainly
Chau 1998; Lee and Jeon 2011; Li et al. 2019; Zhou et al. affected by the fissure angle under lower confining pressure,
2019 and 2020; Feng et al. 2021) and even multiple flaws while the effect of confining pressure becomes more signifi-
(Sagong and Bobet 2002; Cao et al. 2016a and b; Huang cant under higher confining pressure. Yang et al. (2021) con-
et al. 2019; Yang and Zhang 2020). These results show that ducted conventional triaxial compression tests on sandstone
the mechanical properties and cracking behaviors of flawed specimens containing two non-coplanar flaws with different
rocks under static loading are different from that under ligament angles. The crack is easier to coalesce when the
dynamic loading, but both of them are significantly affected ligament angle is parallel to the maximum principal stress,
by the flaw geometric configuration, including the length and and five typical coalescence modes of the two non-copla-
angle of the flaw, the space between two flaws, and the length nar flaws were classified. The effects of high strain rate on
and angle of the rock bridge. Since rocks in underground the mechanical and failure behaviors of three-dimensional

13
Effect of Confining Pressure and Strain Rate on Mechanical Behaviors and Failure Characteristics…

confined flawed rocks, however, have not been considered in introduced into triaxial dynamic tests to simulate engineer-
the previous literature. To reveal the mechanical properties ing flawed rocks simultaneously confronted with three-
of rocks confronted with a three-dimensional stress state dimensional static stress and dynamic loading. The influ-
under high strain rates, some dynamic triaxial tests were ence of radial confining pressure and strain rate on dynamic
carried out through modified SHPB apparatus with either a strength, deformation properties and energy dissipation of
pressure (Christensen et al. 1972; Hokka et al. 2016; Si et al. the single-flawed rocks are first systematically studied. By
2019; Gong et al. 2019; Du et al. 2020a and b) or displace- combining sieving tests and three parameters generalized
ment (Yuan et al. 2011; Liu et al. 2019) boundary condi- extreme value function, the fragmentation distributions of
tion. However, such triaxial dynamic tests have only been the confined single-flawed rocks are then adequately char-
conducted on intact rocks, and investigation on the dynamic acterized under different loading conditions. Meanwhile,
mechanical response and failure characteristics of flawed the final failure patterns of specimens are classified through
rocks under three-dimensional confinement has never been post-mortem examination, and the fracturing mechanism
reported. of new cracks is further revealed by scanning electron
In the present study, via a modified SHPB apparatus, microscope.
a sandstone specimen containing a prefabricated flaw is

Fig. 2  Thin section observation of Neijiang sandstone material with a single and b orthogonally polarized light; c energy-dispersive X-ray spec-
troscopy analysis and mineral composition

13
 W. You et al.

Table 1  Basic physical and mechanical properties of tested Neijiang that the mineral particles are evenly distributed with an aver-
sandstone age size of about 0.2 to 0.3 mm. By the energy-dispersive
Density (kg/ Uniaxial Indirect ten- Elastic Poisson’s X-ray spectroscopy (EDS) method, Fig. 2c shows the elec-
m3) compres- sile strength modulus ratio tronic spectrum and mineral composition of the sandstone,
sion strength (MPa) (GPa) in which O, Si, Al, and Mg are the main elements, and the
(MPa)
Neijiang sandstone is mainly composed of quartz (43%),
2354 61 7 6.7 0.23 plagioclase (35%), orthoclase (5%), rock debris (14%), and
some clay minerals (3%). Using a hydraulic servo-con-
trolled MTS-793 rock testing system, preliminary tests are
2 Experiment Setup conducted to obtain some essential physical and mechani-
cal properties of the tested Neijiang sandstone, as listed in
2.1 Physical and Mechanical Properties of Neijiang Table 1. With the loading rate of 0.005 mm/s, the uniaxial
Sandstone compression strength of 61 MPa is calculated by using the
cylinder specimen with a diameter of 50 mm and an aspect
Fine-grained homogenous Neijiang sandstone (Sichuan ratio of 2:1; the indirect tensile strength of 7 MPa is calcu-
province, China) is used in this study. Under an optical lated by using the flattened Brazilian disc specimen with a
microscope with a single and orthogonally polarized light, geometry of 50 × 25 ­mm2 (diameter × thickness).
the thin section observation of sandstone can be obtained, as
shown in Fig. 2a, b. It can be observed that the sandstone is a
crystalline and blocky structure without distinct defects, and

Fig. 3  Basic information of the single-flawed specimen: a a photograph of the prepared specimens, b the specimen geometry, c a typical static
strain–stress curve and d the schematic loading conditions

13
Effect of Confining Pressure and Strain Rate on Mechanical Behaviors and Failure Characteristics…

2.2 Specimen Preparation and Test scheme bar first impacts the incident bar producing a comprehensive
incident wave (εi), as depicted in Fig. 4b. As the incident
All test specimens used in this study are cored from one wave reaching the bar–specimen interface, part of the wave
rock block along the same direction to reduce the influence is reflected back as the reflected wave (εr), and the remainder
of anisotropy, and some specimens are shown in Fig. 3a. goes through the specimen and travels into the transmitted
The sandstone is first manufactured into cuboid specimens bar as the transmitted wave (εt). All these three strain sig-
with the geometry of 45 × 35 × 35 m ­ m3, and then a prefabri- nals can be collected by the data acquisition system with a
cated flaw is cut at the middle region of the cuboid specimen sampling frequency of 1­ 07 Hz in the tests. A red copper disk
using the drilling bit and diamond wire saw, as presented in with the geometry of 10 × 2 m ­ m2 (diameter × thickness) is
Fig. 3b. The length and width of the prefabricated flaw is glued on the left end of the incident bar as a pulse shaper to
about 7 and 1 mm, respectively. The angle between the flaw facilitates the dynamic stress balance of specimens (Frew
and the dynamic loading direction is 45°. After carefully et al. 2001; Dai et al. 2010).
polishing using a grinding machine, all the specimens have The confining pressure system includes an axial confin-
a surface roughness less than 0.02 mm and the deviation of ing pressure system and a radial confining pressure system.
perpendicularity between two adjacent surfaces is smaller The axial pre-stress on the specimen is applied by driving
than 8′. the transmitted bar moving toward the incident end via the
The typical static strain–stress curve of the single-flawed axial confining pressure system. Radial confining pressure
specimen under the uniaxial compression test is shown in is applied by the hydraulic oil in the chamber of the radial
Fig. 3c. The determined static uniaxial compressive strength confining system. To ensure the validity of triaxial confined
of single-flawed specimen (σc) is 56.8 MPa. For investigat- dynamic tests, the specimen should be well sealed as spe-
ing the radial confinement effect on dynamic mechanics cifically shown in Fig. 4c. Both ends of the specimen are
and failure characteristics of fractured rock under a three- properly lubricated by high vacuum grease and then sand-
dimensional stress state, single-flawed specimens would be wiched between the incident and transmitted bars. A heat
subjected to different radial confining pressures firstly, and shrink tube first wraps the specimen and the steel hoops fix
then the dynamic pulse act on the axial direction of rock the tube on the bars. The hot-melt adhesive is used to further
specimen, as depicted in Fig. 3d. Four radial confining pres- prevent the high-pressure oil from seeping through the heat
sures are set as 5.7, 11.4, 17.0 and 22.7 MPa, and this stress shrink tube. In addition, two high-strength alloy sheets are
range approximately covers the vertical ground stress under attached on the side surface of the single-flawed specimen
the depth ranging from 200 to 900 m (Wasantha and Ranjith to prevent the high-pressure oil from penetrating the heat
2014). Such engineering depth represents typical excava- shrink tube through the prefabricated flaw. Figure 4d dem-
tion depth of underground structure. Numerous underground onstrates the radial confining pressure histories under four
engineering projects are built in this depth range, e.g., the different initial stress states in the whole loading process.
underground powerhouse of Baihetan hydropower station, The variation magnitude of radial confining pressure is less
Beishan nuclear waste repository, and Dongguashan copper than 1 MPa, which indicates that the radial chamber is large
mine. A high enough axial pre-stress of 22.7 MPa is adapted enough to ensure the stability of confining pressure in the
in all tests to prevent the specimen from separating the bars. dynamic impact process.
The specimens are divided into four groups according to the
different radial confining pressures, and six specimens are
tested in each group for different strain rates.
3 Methodology
2.3 Testing Apparatus and Procedure
3.1 Data Acquisition and Processing
As shown in Fig. 4a, the triaxial dynamic tests on the single-
flawed specimen are conducted using the modified SHPB According to the studies by Du et al.(2020a), the one-
device, which mainly consisted of a bar system, data acquisi- dimensional stress wave theory by Kolsky (1949) is also
tion system, and confining pressure system. The bar system applicable in the triaxial SHPB tests. However, it should
consists of three high-strength steel bars, including a 0.3 m be noted that a detachment wave shall be produced due to
striker bar, a 3.0 m incident bar and a 2.0 m transmitted bar. the separation between the reaction plate and flange in the
The elastic modulus, density and diameter of these bars are triaxial SHPB tests (Chen et al. 2018). The data processing
211 GPa, 7800 kg/m3 and 50 mm, respectively. The data methods highly depend on whether the detachment wave
acquisition system mainly contains strain gauges mounted arriving at the strain gauge after the reflected wave. With
on the bars and a high-frequency dynamic digital acquisition careful check, the detachment wave passes through the strain
instrument. During the dynamic loading process, the striker gauge after the reflected wave for all tests in this study. In

13
 W. You et al.

Fig. 4  Modified split Hopkinson pressure bar (SHPB) with radial and axial confining system: a a photograph, b the schematic of dynamic wave
transmission, c the sealing process and d the radial confining pressure histories

13
Effect of Confining Pressure and Strain Rate on Mechanical Behaviors and Failure Characteristics…

this case, the dynamic forces on the incident end ( Pi ) and ⎧

the transmitted end ( Pt ) of the specimen can be calculated ⎪ UI = Eb Ab Cb ∫ (𝜀i − 𝜀i 𝜀pre )dt
2

by the following equation: ⎪

⎪
⎨ UR = Eb Ab Cb ∫ 𝜀r dt (3)
2
,
Pi = Ab Eb (𝜀i + 𝜀r ), Pt = Ab Eb 𝜀t , (1) ⎪
⎪
where Ab and Eb are the cross-sectional areas and Young’s ⎪ UT = Eb Ab Cb ∫ (𝜀t − 𝜀t 𝜀pre )dt
2

modulus of the steel bars. ⎩

Once the dynamic stress equilibrium (i.e., Pi ≈ Pt) is
where UI, UR, and UT are the incident, reflected, and trans-
achieved in the dynamic loading process, the dynamic strain
mitted energy, respectively. According to the first law of
rate 𝜀̇ , dynamic nominal strain εdyn, and dynamic nominal
thermodynamics, the dissipated energy (ΔU) for the con-
stress σdyn of the specimen can be expressed as follows:
fined single-flawed specimen in the dynamic loading process
⎧ 2Cb could be expressed as Eq. (4).
⎪ 𝜀̇ = − l 𝜀r
⎪ s ΔU=UI − UR − UT . (4)
⎨ 𝜀dyn = − l ∫0 𝜀r dt ,
⎪ 2Cb t
(2) In the triaxial SHPB tests, most of the dissipated
⎪ s
energy (ΔU) is consumed by the rock crushing energy
⎪ Ab Eb
⎪ 𝜎dyn = (𝜀t − 𝜀pre ) (U c) causing new crack surfaces, and a negligible part
⎩ As dissipates in the form of thermal radiation and acoustic
emission. Therefore, the dissipated energy can be approx-
where Cb is the longitudinal wave velocity of the steel bars,
imately regarded as the crushing energy of the specimen
5201 m/s; εpre is the pre-strain caused by the axial pre-stress;
(i.e., Uc ≈ ΔU). Generally, energy utilization efficiency
ls and As are the length and the cross-sectional area of speci-
(ed) is defined as the ratio of useful work done to the total
men, respectively.
input energy, and the energy dissipation density (e ρ) is
For the dynamic triaxial SHPB tests, the traditional
defined as the dissipated energy per unit volume of rock
energy calculation formula is no longer applicable. Con-
specimen by Eq. (5).
sidering the elastic strain energy restored in the bars, a new
{
modified formula to determine the energy partitions is given ed =Uc ∕UI ≈ ΔU∕UI
as follows (Yan et al. 2021): , (5)
e𝜌 =Uc ∕Vs ≈ ΔU∕Vs

where eρ is the energy dissipation density for rock fragmen-

tation; Vs is the volume of the specimen.

Fig. 5  Sieving test: a a standard sieve and b sieved fragments

13
 W. You et al.

3.2 Sieving Tests and GEV Function curve and reflects the concentrated trend of fragment distri-
bution, which characterizes the size of average fragments,
Dynamic fragmentation of rocks covers a wide range of and the higher location parameter indicates a larger average
applications in many fields, including fault activity research, fragment size. It should be noted that the cumulative dis-
rock mass landslide, earthquake, oil or gas reservoir drilling, tribution function of GEV should meet the requirement of
mining and blasting, and hydraulic fracturing (Hou et al. [1 + ξ((x-η)/ω)] > 0.
2015; Li et al. 2018). In this study, the manual sieving tests
are carried out on all destroyed single-flawed sandstone
specimens after dynamic loads. With the mesh size of 40, 4 Results and Discussion
20, 10, 5, 2, 1, 0.5, 0.25, and 0.075 mm, a standard sieve is
adopted in the sieving tests. The standard sieve and the typi- 4.1 Dynamic Stress Balance
cal specimen fragments with different sizes after sieving are
shown in Fig. 5a, b, respectively. To quantitatively describe The dynamic stress balance on both ends of the specimen
the dynamic fragmentation characteristics, a generalized is a prerequisite for the quasi-static data analysis method in
extreme value (GEV) distribution function is introduced to the SHPB tests (Dai et al. 2010; Zhou et al. 2012). In our
characterize the statistical characteristics of rock fragments tests, the pulse shaping technique is employed to produce a
of single-flawed specimens under different radial confining ramp impact wave with a duration of about 350 μs, which
pressures after dynamic loading. Many studies have shown provides enough time for wave reverberation in the speci-
that the GEV distribution fits the experimental data well men to achieve a dynamic stress equilibrium state. Figure 6a
for analyzing rock fragmentation characteristics (Du et al. demonstrates typical time-varying stresses on the two ends
2020b; Hou et al. 2015). The GEV statistical distribution is of the confined single-flawed specimen. The dynamic stress
a continuous probability distribution based on the extreme on the incident end is the sum of the incident and trans-
value theory and the cumulative distribution function is mitted waves, marked as In + Re; the dynamic stress on the
expressed as follows: transmitted end of the specimen can be expressed by the
1
transmitted wave, noted as Tr. At the beginning of dynamic
F(x;𝜂, 𝜔, 𝜉) = e−(1+𝜉(
x−𝜂 − 𝜉
𝜔
))
, (6) loading, the incident stress (In + Re) is generally greater than
the transmitted stress (Tr). The dynamic stresses on both
where η, ω and ξ are the location parameter, the scale ends of the specimen are almost identical after several wave
parameter and the shape parameter, respectively. The shape reverberations, which means that the dynamic stress equilib-
parameter ξ reflects the kurtosis and skewness of distribu- rium of the three-dimensional confined single-flawed speci-
tion curve; the scale parameter ω measures the distribution men is well achieved in the triaxial dynamic tests. Due to the
range of rock fragment size, and a large value of the scale axial constraint, the starting point of the incident wave and
parameter represents the wider distribution range; the loca- the transmitted wave are equal to the static axial pre-stress.
tion parameter η determines the position of distribution

Fig. 6  Dynamic stress balance check and the determination of strain rate: a the typical time-varying stresses on the two ends of specimen and b
the evolution curves of the strain rate and axial nominal stress

13
Effect of Confining Pressure and Strain Rate on Mechanical Behaviors and Failure Characteristics…

Fig. 7  Dynamic nominal stress–strain curves of single-flawed specimen at various strain rates under the radial confining pressure of a 5.7, b
11.4, c 17.0 and d 22.7 MPa

After the dynamic stress equilibrium is achieved, the respectively. Note that this nominal stress–strain curve only
time-varying strain rate and axial nominal stress are obtained exhibits the dynamic loading part while excluding the static
by Eq. (2). Figure 6b depicts the typical evolution curves of pre-loading part. Compared with the static uniaxial compres-
the strain rate 𝜀(t)
̇ and axial nominal stress. In the vicinity sion test (Fig. 3c), the dynamic nominal stress–strain curve
of the peak stress, it is clear that there exists a flat plateau directly enters the linear deformation stage without a compac-
in the time–strain–rate curve during 168 and 214 μs, which tion stage for each radial confining pressure group. This may
indicates that the specimen is axially deformed at a con- be because the specimen has been compacted by the radial and
stant speed. In this study, the average value of the plateau is axial confining pressures before the dynamic loading. After
defined as the strain rate. For all tested specimens under dif- nonlinear growth, the dynamic nominal stress–strain curve
ferent radial confining pressure, the strain rates range from reaches the peak value and then begins to decline. Note that the
85 to 252 ­s−1. In this study, the dynamic stress equilibrium dynamic nominal strain of the single-flawed specimen under
of all the tested specimens has been critically checked and the radial confining pressure of 17.0 MPa at the strain rate of
the following result analysis is based on valid SHPB tests. 96 ­s−1 recovers back quickly after the dynamic nominal stress
reaches the peak value. This specimen is unbroken but slightly
4.2 Dynamic Strengths and Deformation Properties damaged at the tip of the prefabricated flaw (see Fig. 7c). The
peak value of the dynamic nominal stress history is defined
Figure 7 depicts the dynamic nominal stress–strain curve as the dynamic strength (σdyn) of the confined single-flawed
of the confined single-flawed specimen under four different specimen, and the corresponding dynamic nominal strain is
radial confining pressures of 5.7, 11.4, 17.0 and 22.7 MPa, set as the dynamic failure strain (εf). The slope of the dynamic

13
 W. You et al.

Fig. 8  Dynamic strength vari-

ation of confined single-flawed
specimens in different loading
conditions

Fig. 9  Dynamic deformation

properties of confined single-
flawed specimens in different
loading conditions: a dynamic
elastic modulus and b dynamic
failure strain

nominal stress–strain curve in the liner increasing stage is confining pressures (p) and various strain rates ( 𝜀̇ ). At a
defined as the dynamic elastic modulus (Ed). given radial confining pressure, the dynamic strength shows
Figure 8 shows the relationship between dynamic strength high strain rate dependence and increases with a rising strain
(σ dyn) of single-flawed specimens and different radial rate. Meanwhile, the dynamic strength also increases with

13
Effect of Confining Pressure and Strain Rate on Mechanical Behaviors and Failure Characteristics…

the increase of radial confining pressure, indicating that

the radial confining pressure enhances the dynamic bear-
ing capacity of rocks. These variation trends of dynamic
strength could be expressed by the fitting surfaces, and the
fitting formula is as follows:

𝜎dyn = 0.445𝜀+5.126p
̇ − 0.00315𝜀p+51.532
̇ R2 = 0.9528.
(7)

This formula well reflects the mutual relationship

between confining pressure and strain rate on dynamic
strength. As the radial confining pressure (p) increases,
the coefficient of strain rate (0.445–0.00315p) decreases,
indicating the increment of dynamic strength will decrease
in the same strain rate range, which is similar to the results
reported by Gong et al.(2019) and Du et al.(2020a). With
the increase of strain rate ( 𝜀̇ ), the coefficient of radial con-
Fig. 11  Failure patterns of the single-flawed specimens under differ-
fining pressure (5.126–0.00315 𝜀̇ ) decreases, suggesting ent radial confining pressures and strain rates
the confining pressure enhancement effect on the dynamic
strength will also reduce, which agrees with the conclu-
sion drawn by Fu et al.(2021). The initiation, propaga- different strain rates in each radial confining pressure
tion and interaction of micro-cracks mainly cause the final group. The strain rate is represented by the rainbow color
failure of rock. On the one hand, under a high strain rate, and increases successively from left to right in each radial
micro-cracks are hard to propagate in a short time and confining pressure group. Compared to the dynamic
more micro-structures in rock specimens are involved in strength, the strain rate and confining pressure depend-
the resistance to damage, leading to the macro strength of ency of the dynamic elastic modulus and dynamic failure
rocks increase. On the other hand, the confining pressure strain are not apparent.
makes micro-cracks close and restrain the lateral deforma-
tion of rocks, which also enhances the rock strength. Fig-
ure 9 presents the dynamic elastic modulus and dynamic
failure strain of confined single-flawed specimens under

Fig. 10  Three main failure types of confined single-flawed specimen in different loading conditions

13
 W. You et al.

Fig. 12  The effect of strain rate and radial confining pressure on radial confining pressure of 5.7 MPa; c the mass cumulative fre-
fragmentation characteristics: a the mass cumulative frequency and quency variation and d fragment size distribution at similar strain rate
b fragment size distribution at different strain rates under the same but different radial confining pressures from 5.7 to 22.7 MPa

4.3 Failure Behavior and Fragmentation tip of the prefabricated flaw to the root of the specimen,
Characteristics which makes the single-flawed specimen showing a mac-
roscopic single diagonal failure pattern and breaking into
With the high radial confining pressure, the failed specimen two large pieces.
still gathered together even after the dynamic impact. There- Failure type II (Fig. 10b): In addition to two OSc, two
fore, the failure patterns of confined single-flawed speci- macro quasi-coplanar cracks (QCc) intersect the prefabri-
mens could be examined by post-mortem examination. After cated flaw at a small angle. These two QCc also extend from
careful examination, three main failure types are classified the tip of the prefabricated flaw in another diagonal direc-
according to the position relationship between new cracks tion, which makes the single-flawed specimen display an
and prefabricated flaw, as depicted in Fig. 10. OSc-QCc mixed “X” shape failure pattern. Generally, four
Failure type I (Fig. 10a): There are mainly two macro major fragments are formed after destroy.
oblique shear cracks (OSc) intersecting with the prefabri- Failure type III (Fig. 10c): Accompanied by two QCc,
cated flaw at a large angle. These two OSc initiate from the a macro through shear crack (TSc) directly passes through

13
Effect of Confining Pressure and Strain Rate on Mechanical Behaviors and Failure Characteristics…

the middle of the prefabricated flaw and extends along the a relatively higher strain rate, the middle and small size
diagonal direction of specimen. The single-flawed specimen fragments apparently increase after the specimen failed
finally exhibits a TSc-QCc mixed “X” shape failure pattern. (b3 and b4 in Fig. 12b). To research the radial confining
Except for the new macroscopic cracks, some small rock pressure effect on the fragment distribution of the con-
debris probably spalls from the surface of the failure speci- fined single-flawed specimen, Fig. 12c, d demonstrates
men. Figure 11 exhibits the final failure patterns of all bro- the mass cumulative frequency variation and fragment
ken specimens. The solid symbol, half filled symbol and size distribution at a similar strain rate about 140 ­s−1 but
hollow symbol represent the type I, type II, and type III increasing radial confining pressure from 5.7 to 22.7 MPa.
failure pattern, respectively. In our tests, the type II failure It can be observed that the position of the mass cumulative
pattern is the most common, and the percentages of each frequency curves increases first and then decreases with
failure pattern are about 13, 52 and 35%, respectively. The increasing confining pressure. Under a lower radial con-
type I failure pattern is only observed in the 5.7 MPa radial fining pressure of 5.7 MPa, the specimen exhibits a type I
confining pressure group and rather lower strain rates. With failure pattern with two large fragments (d1 in Fig. 12d).
the increase of radial confining pressure, the type III failure The corresponding mass cumulative frequency curve is
pattern becomes more obvious, which means that the effect thus at the bottom position. When the single-flawed speci-
of radial confining pressure on the failure pattern occupies mens exhibit type II or III failure (d2–d4 in Fig. 12d), the
an important role. mass cumulative frequency curve gradually goes to a lower
Figure 12 illustrates the results of manual sieving position as the radial confining pressure increases.
tests for revealing the influence of dynamic strain rate All the location parameters and scale parameters of the
and radial confining pressure on the fragment distribu- fitted GEV line of each failed specimen are illustrated in
tion of tested rocks. Using a l­og10 transformation of the Fig. 13. It is easily observed that the location parameter
mesh size, the mass cumulative frequency variations and decreases, but the scale parameter increases with the ris-
corresponding GEV fit line at different strain rates under ing strain rate, which indicates the higher the strain rate,
the 5.7 MPa radial confining pressure group are plotted in the smaller average fragment size and wider fragment dis-
Fig. 12a. After the examination of the least squares fit, all tribution range. Meanwhile, the mean value of the loca-
the coefficients of determination R2 are greater than 0.99, tion parameters in each radial confining pressure group
indicating that the fragment size distribution could be ade- decreases initially and increases afterward with the ris-
quately described by the three parameters GEV function. ing confining pressure, while the mean value of the scale
As the strain rate increases, the percentage of small size parameter exhibits an opposite trend. Since half of the
fragments increases and the mass cumulative frequency specimens in the 5.7 MPa radial confining pressure group
curves is in a higher position. Under a lower strain rate, the are type I failure, the average fragment size and the frag-
single-flawed specimen breaks into two large fragments ment distribution range are rather larger and narrower,
with some small debris (b1 and b2 in Fig. 12b). Under respectively. When the single-flawed specimen shows

Fig. 13  The variation of two GEV parameters for single-flawed specimens under different radial confining pressures and strain rates: a location
parameter and b scale parameter

13
 W. You et al.

Fig. 14  Incident, reflected, transmitted, and dissipated energy variation of the single-flawed specimens under different radial confining pressures
and strain rates

type II or III failure pattern, the higher the radial confin- are determined. Figure 14 presents the incident energy of
ing pressure, the larger is the constraint for the specimen confined single-flawed specimens under different strain
and the more is the fracturing difficulty, which leads to rates in each radial confining pressure group. The abscissa
the increase of average fragment size and the decrease of at the bar graph bottom represents different radial confining
fragment distribution range. pressure groups, and the strain rate increases successively
from left to right in each radial confining pressure group. A
4.4 Energy Parameters Analysis higher strain rate is generally obtained by raising the impact
velocity of the striker bar. Under the same radial confining
Using Eqs. (3)–(5), the energy partitions of the confined pressure, the incident energy shows a positive relationship
single-flawed specimens under different loading conditions with the strain rate since the higher impact velocity inputs

Fig. 15  Energy dissipation density and fragmentation effect: a energy dissipation density of the single-flawed specimens in different loading
conditions; b the location parameter in terms of incident energy and radial confining pressure

13
Effect of Confining Pressure and Strain Rate on Mechanical Behaviors and Failure Characteristics…

more energy to the loading system. The confining pressure shows the coupled effects of confining pressure and incident
makes rocks deform more difficult. As the radial confining energy on the GEV location parameter. When the incident
pressure increases, specimens require more incident energy energy exceeds 250 J, the value of the location parameter
to obtain a similar strain rate. The reflected energy, trans- increases with the radial confining pressure, indicating the
mitted energy, and dissipated energy are normalized by the fragment size of the single-flawed specimen is smaller under
incident energy, which are shown in Fig. 14 in the form of a low confining pressure. As the incident energy is lower than
percentage. Note that the proportion of transmitted energy 250 J, however, the single-flawed specimen under medium
of the unbroken specimen is significantly higher while the confining pressure has the smaller rock fragment size. It can
proportion of dissipated energy is apparently lower than that be seen that even a lower incident energy can produce a
of other specimens in the same group. small rock fragment size under medium confining pressure.
Under the same radial confining pressure, the higher the Therefore, in practical underground rock engineering, it is
strain rate, the higher is the proportion of reflected energy not necessary to pursue too high input energy to achieve
and the lower is the proportion of transmitted energy. The an ideal rock fragmentation effect. Selecting an appropriate
higher incident energy generally enhances the fracture input energy according to the confining pressure conditions
degree of the specimen, which makes the specimen more can reduce unnecessary energy dissipation and further make
difficult to transmit the stress wave and thus a greater part the cost reduction.
of the stress wave is reflected. Though more energy is con-
sumed for rock breakage and producing a large number of 4.5 Microscopic Observation on the Crack Surfaces
new cracks with the increase of strain rate, the proportion of
dissipated energy, i.e., the energy utilization efficiency, ed, To further reveal the microscopic fracture mechanism of
has no obvious change, indicating that the energy utilization triaxial confined single-flawed sandstone, a Hitachi SU3500
efficiency is hardly improved by adjusting the strain rate scanning electron microscope (SEM) is used in our tests
under external dynamic loading. to observe the microstructure and morphology of the rock
Under a similar strain rate, the proportion of reflected crack surface. The surface morphology of the three main
energy decreases, while both the proportion of transmitted cracks (i.e., OSc, TSc and QCc) mentioned in Sect. 4.3
energy and the energy utilization efficiency increase as the is observed by SEM. Under a similar strain rate around
radial confining pressure increases. For example, under the 140 ­s−1, Fig. 16 exhibits the micro morphology of OSc or
strain rate of about 140 ­s−1, as the radial confining pressure TSc of the specimen under confining pressure from 5.7 to
increases from 5.7 to 22.8 MPa, the proportion of reflected 22.7 MPa, and Fig. 17 presents the microscopic observa-
energy changes from 48 to 20%, while the proportion of tion results of QCc surface of the specimen under confining
transmitted energy and the energy utilization efficiency pressure from 11.4 to 22.7 MPa. Considering that the aver-
increase from 12 to 31% and from 40 to 48%, respectively. age size of mineral particles is about 0.2–0.3 mm, the first
The higher radial confining pressure improves the bear- picture, for each radial confining pressure group, shows the
ing capacity of the specimen, and more energy is used to overall morphology of the crack surface at a lower magni-
fracture the rock. The residual strength is also enhanced fication (100 times), and the rest of the pictures show the
and the specimen is relatively more complete after failure, detailed view at higher magnification (500 or 1000 times).
and thus the proportion of transmitted energy is positively For OSc or TSc, on the whole, the crack surface is rough
correlated with the radial confining pressure. and the fluctuation is small, and the edges and corners of
Figure 15a illustrates the effect of strain rate and radial the mineral crystal are groud resulting in large quantities
confining pressure on the dynamic energy dissipation den- of debris gathered on the crystal surface. There are many
sity of confined single-flawed specimens. With the increas- obvious friction traces on the crack surface. For example,
ing strain rate, the energy dissipation density of confined sin- the biotite flake minerals are bent (see a3 and b3 in Fig. 16),
gle-flawed specimen exhibits a rising trend under the same indicating the sandstone is pulled along the crack surface
radial confining pressure. At a similar strain rate, the energy under the action of shear stress. Meanwhile, a lot of crys-
dissipation density features a positive correlation with the tal grains are disintegrated (see c3 in Fig. 16), and many
radial confining pressure increase. For example, with the scratched traces (e.g., ripple shape texture in c2 and moiré
strain rate of about 140 ­s−1, the corresponding energy dis- pattern in d3 of Fig. 16) are generally observed on the frac-
sipation density for each radial confining pressure group is ture surface of grains. Both the intergranular (IG) fracture
1.59, 1.96, 2.26, and 2.66 J/cm3, respectively. In the triaxial and transgranular (TG) fractures are commonly observed on
dynamic compression tests, the fragmentation characteristics OSc or TSc (see a2 and d2 in Fig. 16). Under higher radial
of rock specimens are influenced by both confining pres- confining pressure, the cementation is tighter and the shear
sure and incident energy. Since the average fragment size is dominant crack surface is flattened more.
directly reflected by the GEV location parameter, Fig. 15b

13
 W. You et al.

Fig. 16  SEM images for the surface of the oblique shear crack (OSc) or through shear crack (TSc) under different radial confining pressures: a
5.7, b 11.4, c 17.0 and d 22.7 MPa

For QCc, in general, the crack surface is relatively clean (see a3 in Fig. 17). From the above analysis, the QCc is thus
and undulating high and low. The edges and corners of the mainly caused by the tensile mechanism. Most micro-cracks
exposed crystal grain are obvious, and the fracture surface of are IG fracture, and TG fracture is relatively rare. With the
the crystal grain is flat and smooth with few friction powders rising of radial confining pressure, the TG fracture and the
(e.g., a2 and c3 in Fig. 17). The sandstone seems to be pulled friction trace appear more.
away from the crack surface, and the cleavage fracture char-
acteristics are evident, such as the river valley terrace shape
fracture (see b3 in Fig. 17). Meanwhile, the fracture of the
biotite flake minerals is uneven without an obvious crease

13
Effect of Confining Pressure and Strain Rate on Mechanical Behaviors and Failure Characteristics…

Fig. 17  SEM images for the surface of the quasi-coplanar crack (QCc) under different radial confining pressures: a 11.4, b 17.0 and c 22.7 MPa

5 Conclusions Quantified by the mean value, both the average dynamic

elastic modulus and dynamic failure strain also increase
In this study, a series of triaxial impact loading tests are with the rising radial confining pressure.
conducted on single-flawed sandstone specimens confronted 2. Three main cracks, i.e., oblique shear crack (OSc),
with three-dimensional static stress using an SHPB appara- quasi-coplanar crack (QCc), and through shear crack
tus modified with axial and radial confining systems. The (TSc), are classified in our tests. Three main failure
effect of the radial confining pressure and strain rate on types I–III are OSc dominant, OSc–QCc mixed and
dynamic strength, deformation properties, fragment distri- TSc–QCc mixed, respectively. The specimen failure is of
bution, failure pattern, and energy dissipation of the single- type I when the radial confining pressure and strain rate
flawed sandstone are systematically investigated. Moreover, are rather low. More specimen failures are of type III
the microscopic observation using scanning electron micro- as the radial confining pressure rises. The higher strain
scope (SEM) further reveals the fracture mechanism of new rate induces a smaller fragment size and wider fragment
cracks. The following major conclusions can be drawn: distribution range, while a higher confining pressure has
the opposite effect.
1. At a given radial confining pressure, the dynamic 3. Except the transmitted energy, the incident, reflected,
strength shows a positive correlation with strain rate. and dissipated energy apparently increase with the
Meanwhile, the higher radial confining pressure increasing strain rate. As the radial confining pressure
enhances the dynamic strength of rock specimens. rises, the proportion of reflected energy decreases, while
both the proportion of transmitted energy and dissi-

13
 W. You et al.

pated energy increase. The energy utilization efficiency Frew DJ, Forrestal MJ, Chen W (2001) A split Hopkinson pressure
increases with the rising radial confining pressure, but bar technique to determine compressive stress-strain data for rock
materials. Exp Mech 41(1):40–46
it is insensitive to the strain rate. Both the strain rate and Fu Q, Bu M, Su L et al (2021) Dynamic triaxial compressive response
confining pressure will increase the energy dissipation and failure mechanism of basalt fibre-reinforced coral concrete.
density of the single-flawed sandstone. Int J Impact Eng 156:103930
4. From the microscopic observation results in SEM tests, Gong FQ, Si XF, Li XB, Wang SY (2019) Dynamic triaxial compres-
sion tests on sandstone at high strain rates and low confining pres-
the fracture mechanisms of three main cracks are further sures with split Hopkinson pressure bar. Int J Rock Mech Min
revealed. The OSc and TSc are mainly produced by the 113:211–219
shear slip mechanism, while the QCc is generally ten- Hoek E, Brown E (1980) Underground excavations in rock. Inst. Min.
sile dominated. Under higher radial confining pressure, 537 Metall, London
Hoek E, Bieniawski ZT (1984) Brittle-fracture propagation in rock
the cementation is tighter and the shear dominant crack under compression. Int J Fract 26(4):276–294
surface is more flattened on OSc or TSc, while both the Hokka M, Black J, Tkalich D et al (2016) Effects of strain rate and
TG fracture and the friction trace increase on QCc. confining pressure on the compressive behavior of Kuru granite.
Int J Impact Eng 91:183–193
Hou TX, Xu Q, Yang XG, Lu PY, Zhou JW (2015) Experimental study
of the fragmentation characteristics of brittle rocks by the effect
Acknowledgements The authors thank the financial support from the of a freefall round hammer. Int J Fract 194:169–185
National Natural Science Foundation of China (No. 52039007) and Huang D, Gu DM, Yang C, Huang RQ, Fu GY (2016a) Investigation
the Youth Science and Technology Innovation Research Team Fund on mechanical behaviors of sandstone with two preexisting flaws
of Sichuan Province (2020JDTD0001). under triaxial compression. Rock Mech Rock Eng 49:375–399
Huang YH, Yang SQ, Zhao J (2016b) Three-dimensional numerical
Declarations simulation on triaxial failure mechanical behavior of rock-like
specimen containing two unparallel fissures. Rock Mech Rock
Conflict of interest All authors declare no conflict of interest. Eng 49(12):4711–4729
Huang C, Yang W, Duan K, Fang L, Wang L, Bo C (2019) Mechani-
cal behaviors of the brittle rock-like specimens with multi-non-
persistent joints under uniaxial compression. Constr Build Mater
References 220:426–443
Jiang R, Dai F, Liu Y, Li A (2021) Fast marching method for microseis-
Bobet A, Einstein HH (1998) Fracture coalescence in rock-type materi- mic source location in cavern-containing rockmass: Performance
als under uniaxial and biaxial compression. Int J Rock Mech Min analysis and engineering application. Engineering. https://d​ oi.o​ rg/​
35(7):863–888 10.​1016/j.​eng.​2020.​10.​019
Cao RH, Cao P, Fan X, Xiong XG, Lin H (2016a) An experimental and Kolsky H (1949) An investigation of the mechanical properties of
numerical study on mechanical behavior of ubiquitous-joint brittle materials at very high rates of loading. Proc Phys Soc Sect B
rock-like specimens under uniaxial compression. Rock Mech Rock 62(11):676–700
Eng 49(11):4319–4338 Lee H, Jeon S (2011) An experimental and numerical study of fracture
Cao RH, Cao P, Lin H, Pu CZ, Ou K (2016b) Mechanical behavior coalescence in pre-cracked specimens under uniaxial compres-
of brittle rock-like specimens with pre-existing fissures under sion. Int J Solids Struct 48(6):979–999
uniaxial loading: experimental studies and particle mechanics Li YP, Chen LZ, Wang YH (2005) Experimental research on
approach. Rock Mech Rock Eng 49(3):763–783 pre-cracked marble under compression. Int J Solids Struct
Chen R, Yao W, Lu F, Xia K (2018) Evaluation of the stress equilib- 42(9–10):2505–2516
rium condition in axially constrained triaxial SHPB tests. Exp Li XB, Zhou T, Li DY (2017) Dynamic strength and fracturing behav-
Mech 58(3):527–531 ior of single-flawed prismatic marble specimens under impact
Christensen RJ, Swanson SR, Brown WS (1972) Split-Hopkinson-bar loading with a split-Hopkinson pressure bar. Rock Mech Rock
tests on rock under confining pressure. Exp Mech 12(11):508–513 Eng 50(1):29–44
Dai F, Huang S, Xia KW, Tan ZY (2010) Some fundamental issues Li XF, Li HB, Zhang QB, Jiang JL, Zhao J (2018) Dynamic fragmenta-
in dynamic compression and tension tests of rocks using split tion of rock material: characteristic size, fragment distribution and
Hopkinson pressure bar. Rock Mech Rock Eng 43(6):657–666 pulverization law. Eng Fract Mech 199:739–759
Du HB, Dai F, Liu Y, Xu Y, Wei MD (2020a) Dynamic response and Li DY, Han ZY, Sun XL, Zhou T, Li XB (2019) Dynamic mechanical
failure mechanism of hydrostatically pressurized rocks subjected properties and fracturing behavior of marble specimens contain-
to high loading rate impacting. Soil Dyn Earthq Eng 129:105927 ing single and double flaws in SHPB tests. Rock Mech Rock Eng
Du HB, Dai F, Xu Y, Yan ZL, Wei MD (2020b) Mechanical responses 52(6):1623–1643
and failure mechanism of hydrostatically pressurized rocks Liu Y, Dai F (2021) A review of experimental and theoretical
under combined compression-shear impacting. Int J Mech Sci research on the deformation and failure behavior of rocks sub-
165:105219 jected to cyclic loading. J Rock Mech Geotech. https://​doi.​org/​
Feng P, Dai F, Liu Y, Xu NW, Fan PX (2018) Effects of coupled static 10.​1016/j.​jrmge.​2021.​03.​012
and dynamic strain rates on mechanical behaviors of rock-like Liu Y, Dai F, Dong L, Xu NW, Feng P (2018) Experimental inves-
specimens containing pre-existing fissures under uniaxial com- tigation on the fatigue mechanical properties of intermittently
pression. Can Geotech J 55(5):640–652 jointed rock models under cyclic uniaxial compression with dif-
Feng P, Xu Y, Dai F (2021) Effects of dynamic strain rate on the energy ferent loading parameters. Rock Mech Rock Eng 51(1):47–68
dissipation and fragment characteristics of cross-fissured rocks. Liu K, Zhang QB, Wu G, Li JC, Zhao J (2019) Dynamic mechani-
Int J Rock Mech Min 138:104600 cal and fracture behaviour of sandstone under multiaxial

13
Effect of Confining Pressure and Strain Rate on Mechanical Behaviors and Failure Characteristics…

loads using a triaxial Hopkinson bar. Rock Mech Rock Eng Yang Y, Zhang Z (2020) Dynamic fracturing process of fissured rock
52:2175–2195 under abrupt unloading condition: a numerical study. Eng Fract
Liu Y, Dai F, Pei P (2021) A wing-crack extension model for tensile Mech 231:107025
response of saturated rocks under coupled static-dynamic load- Yang SQ, Tian WL, Liu XR, Huang YH, Yang J (2021) An experi-
ing. Int J Rock Mech Min 146:104893 mental study on failure mechanical behavior and cracking
Sagong M, Bobet A (2002) Coalescence of multiple flaws in a rock- mechanism of rectangular solid sandstone containing two
model material in uniaxial compression. Int J Rock Mech Min non-coplanar fissures under conventional triaxial compression.
39(2):229–241 Theor Appl Fract Mec 114:102975
Si X, Gong F, Li X, Wang S, Luo S (2019) Dynamic Mohr-Coulomb Yuan FP, Prakash V, Tullis T (2011) Origin of pulverized rocks
and Hoek-Brown strength criteria of sandstone at high strain during earthquake fault rupture. J Geophys Res Solid Earth
rates. Int J Rock Mech Min 115:48–59 116(B6):B06309
Wasantha PLP, Ranjith PG (2014) Water-weakening behavior of Zhang QB, Zhao J (2014) A review of dynamic experimental tech-
Hawkesbury sandstone in brittle regime. Eng Geol 178:91–101 niques and mechanical behaviour of rock materials. Rock Mech
Wei MD, Dai F, Liu Y, Li A, Yan ZL (2021) Influences of loading Rock Eng 47(4):1411–1478
method and notch type on rock fracture toughness measure- Zhou YX, Xia K, Li XB et al (2012) Suggested methods for deter-
ments: from the perspectives of T-stress and fracture process mining the dynamic strength parameters and mode-I fracture
zone. Rock Mech Rock Eng 54:4965–4986 toughness of rock materials. Int J Rock Mech Min 49:105–112
Wong RHC, Chau KT (1998) Crack coalescence in a rock-like mate- Zhou T, Zhu JB, Ju Y, Xie HP (2019) Volumetric fracturing behav-
rial containing two cracks. Int J Rock Mech Min 35(2):147–164 ior of 3D printed artificial rocks containing single and double
Wong LNY, Einstein HH (2009) Systematic evaluation of cracking 3D internal flaws under static uniaxial compression. Eng Fract
behavior in specimens containing single flaws under uniaxial Mech 205:190–204
compression. Int J Rock Mech Min 46(2):239–249 Zhou T, Zhu JB, Xie H (2020) Mechanical and volumetric fracturing
Xia KW, Yao W (2015) Dynamic rock tests using split Hopkinson behaviour of three-dimensional printing rock-like samples under
(Kolsky) bar system - A review. J Rock Mech Geotech 7(1):27–59 dynamic loading. Rock Mech Rock Eng 53(6):2855–2864
Xiao P, Li DY, Zhao GY et al (2020) Mechanical properties and fail- Zou CJ, Wong LNY (2014) Experimental studies on cracking processes
ure behavior of rock with different flaw inclinations under cou- and failure in marble under dynamic loading. Eng Geol 173:19–31
pled static and dynamic loads. J Cent South Univ 27:2945–2958 Zou CJ, Wong LNY, Loo JJ, Gan BS (2016) Different mechanical and
Yan ZL, Dai F, Liu Y, Du HB, Luo J (2020) Dynamic strength cracking behaviors of single-flawed brittle gypsum specimens
and cracking behaviors of single-flawed rock subjected to under dynamic and quasi-static loadings. Eng Geol 201:71–84
coupled static-dynamic compression. Rock Mech Rock Eng
53(9):4289–4298 Publisher's Note Springer Nature remains neutral with regard to
Yan Z, Dai F, Zhu J, Xu Y (2021) Dynamic cracking behaviors jurisdictional claims in published maps and institutional affiliations.
and energy evolution of multi-flawed rocks under static pre-
compression. Rock Mech Rock Eng. https://​doi.​org/​10.​1007/​
s00603-​021-​02564-2
Yang SQ, Jing HW (2011) Strength failure and crack coalescence
behavior of brittle sandstone samples containing a single fissure
under uniaxial compression. Int J Fract 168(2):227–250

13