Simulation Modelling Practice and Theory 140 (2025) 103079

Contents lists available at ScienceDirect

Simulation Modelling Practice and Theory

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

Peridynamics simulating of dynamics crack propagation in rock

mass under blasting load
Liping Li a,b, Xiaochu Chen a,b, Chenglu Gao a,b,* , Zongqing Zhou a,b,c,* ,
Minghao Li a,b , Daosheng Zhang a,b, Jinbo Chen a,b
a
School of Qilu Transportation, Shandong University, Jinan, Shandong, 250002, PR China
b
State Key Laboratory for Tunnel Engineering, Jinan, Shandong 250061, PR China
c
Research Institute of New Material and Intelligent Equipment, Shandong University, Dezhou, Shandong, 251114, PR China

A R T I C L E I N F O A B S T R A C T

Keywords: It is important to study the crack propagation and failure criterion of fractured rock masses under
Peridynamics dynamic blasting loads to ensure the safety and stability of tunnels constructed using the drill-
Rock mass and-blast method. Based on the idea of conventional peridynamics, a peridynamics constitutive
Blast load
model of the pull-compression composite failure of rock materials and a peridynamics simulation
Crack propagation
method for crack propagation in a fractured rock mass under a blasting load were established. In
addition, the criterion for the stress wave incident angle and in-situ stress state for crack prop-
agation in a rock mass were investigated. The incident angle of the stress waves was found to
affect the crack propagation. When the incident angle was 0◦ , new cracks propagated parallel to
prefabricated cracks. When the incident angle was 15◦ -75◦ , wing cracks formed at the tips on both
sides of a prefabricated crack, and the initiation and propagation angles of these new cracks
increased with the incident angle. The in-situ stress affected the propagation direction of an
explosive-generated crack and the range of the fracture zone. When unidirectional stress was
applied, the radial main crack expanded along the direction of the stress, and the range of the
fracture zone decreased with an increase in stress. The applicability and accuracy of the proposed
dynamic disturbance peridynamics simulation method for solving the problem of dynamic failure
and crack growth in a rock mass were verified based on previous physical and model test results.

1. Introduction

Tunnels and other transportation facilities in China have flourished with the implementation of major national strategies such as
the transportation power strategy, and the scale and difficulty of tunnel construction have garnered global attention. As one of the
main construction methods for tunnels and other underground engineering projects [1], the drilling and blasting method has ad-
vantages such as a high rock-breaking efficiency, low construction cost, and strong applicability to surrounding rocks. However, an
explosion impact load is generated during the construction process, which exerts a strong dynamic disturbance on the rocks sur-
rounding a tunnel. The propagation of an explosion stress wave in a fractured rock mass can cause crack propagation and failure in the
rock mass and may give rise to large-scale geological disasters such as rock bursts and collapse [2–4], which severely impact the safety
of tunnel construction and operation.

* Corresponding authors.
E-mail addresses: [email protected] (C. Gao), [email protected] (Z. Zhou).

https://doi.org/10.1016/j.simpat.2025.103079
Received 30 September 2024; Received in revised form 12 January 2025; Accepted 12 January 2025
Available online 12 January 2025
1569-190X/© 2025 Elsevier B.V. All rights are reserved, including those for text and data mining, AI training, and similar technologies.
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 1. Rock blasting zone and its characteristics.

The basic theory of peridynamics (PD) was proposed by Silling of Sandia National Laboratories [5]. Over 20 years of development,
it has been applied in various fields by numerous researchers. The core concept of the PD theory is non-local action. Objects are
dispersed into a finite number of material points containing physical and mechanical information, and their mechanical behaviour is
depicted by solving space integral equations. When addressing the continuum-discontinuity problem, cracks in materials can initiate
and expand naturally. Discontinuity problems can be solved with a high calculation accuracy at both the macroscopic and microscopic
scales [6–7].
The original PD model was the bond model [8], which has clear physical significance and straightforward calculations. However,
Poisson’s ratio is fixed. Thus, only the overall deformation of an object is reflected, without distinguishing a geometric shape change
from the volume change. To address the aforementioned issues, Silling et al. [9] proposed a modal PD model that considers the in-
fluence of the deformation between material points, which further enhanced the theoretical system of PD and expanded its application
scope.
In the exploration of the failure criterion, Silling et al. [10] initially suggested determining whether a bond is broken based on the
critical elongation rate. Lammi et al. [11] formulated a fracture criterion based on the integral of the bond energy density dissipation
rate in the PD of the state. Regarding the research progress of constitutive relations, Bobaru et al. [12] investigated the shear
deformation constitutive relationship of materials. Mitchell [13] explored the plastic deformation constitutive relationship of mate-
rials and examined the flow and viscosity constitutive relationship of materials based on the theory of PD for state forms. In the PD
model, the neighbourhood of material points near the boundary is frequently incomplete, and the force between material points is
imbalanced; that is, there is a boundary effect. To a certain extent, this boundary effect can influence the simulation accuracy.
Currently, the influence of the boundary effect cannot be eliminated, but three effective approaches to mitigate its influence have been
reported: (1) Madenic et al. [14] showed that it could be modified by the force density; (2) Oterkus [15] and Ganzenmüller et al. [16]
employed the energy density to weaken the boundary effect; and (3) Kilic [17] utilized the volume correction method, while Madenci
et al. [18] set a virtual boundary layer at the material boundary.
The impact and compression resulting from blasting damage a rock mass in two ways. First, the stress field within the rock mass
undergoes a significant alteration, which causes the further expansion of natural cracks, pulverisation, and destruction of the rock
masses in adjacent areas, along with initiating new cracks in a distant area of the rock mass [19]. Second, it weakens the mechanical
properties of the rock mass, such as its strength and elastic modulus [20]. Disturbances from blasting can cause various types of
damage to rocks [21]. Taking the blasthole as the origin and progressing outward, crushing, fracture, and elastic vibration damage
zones are formed [19]. The most serious damage is found in the crushing zone, followed by the fracture zone. The elastic vibration zone
only experiences elastic vibration without failure, as shown in Fig. 1.
Laboratory tests, in-situ tests, and numerical simulations are effective approaches for investigating the damage mechanisms of rock
blasting. Nevertheless, laboratory tests have certain drawbacks, including their small scale, high cost, and limited data. Meanwhile, the
deficiencies of in-situ tests include insecurity, the significant influence of the external environment on the test results, and operational
difficulties. Hence, numerical simulation [22–23] have emerged as a common method for studying the damage mechanism of rock
blasting, which has effectively promoted the development of rock dynamics. Ma et al. [24] incorporated the Johnson–Holmquist
constitutive model into the commercial software LS-DYNA to simulate crack propagation in rocks under blasting loads. Fakhimi et al.
[25] proposed a numerical model of rock blasting coupled with DEM-SPH, in which particles interact with each other at the contact
point between normal and tangential springs to simulate the elasticity of rocks. This method was successfully used to simulate crack
propagation in rock materials. Zaid et al. [26] employed the CONWEP method to study the responses of surface blasting in rocks with
different cross-sectional shapes surrounding a tunnel.
Since PD was proposed, it has been extensively utilised in research on general dynamic fractures. Silling [27] performed the first
simulation of the Kalthoff-Winkler dynamic fracture test using a PD model. Ha and Bobaru [28] investigated dynamic propagation and
crack branching based on the theory of PD. Zhou et al. [29] simulated crack propagation in brittle materials under dynamic biaxial
loads using the non-normal PD theory. Zhu et al. [30] proposed a non-ordinary PD model to simulate rock blasting in combination with
the JH2 constitutive model and simulated and compared a single-hole blasting test of granite rock to verify the accuracy of the nu-
merical model. Zhang et al. [31] established a PD model of rock fracturing under the effect of liquid carbon dioxide blasting and
analysed the criterion of fracture expansion in liquid carbon dioxide blasting, as well as the influences of the pressure from shock waves

2
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 2. Constitutive force function of PMB material.

and the number and radius of the blastholes on the blasting effect of liquid carbon dioxide. Kilic [32] defined the form of the rate of
change in PD bond elongation and constructed a corresponding microelastic constitutive force model, which can be employed to
simulate the impact failure problem.
Currently, research on the PD simulation of rock blasting damage mainly centres on two-dimensional circumstances, and there is an
issue with the oversimplification of the constitutive model of rock materials. Hence, it is crucial to predict the stability of a fractured
rock mass under dynamic loads to investigate the fracture propagation criterion and summarise the mechanical behaviour and failure
criterion of a rock mass under dynamic loads. Based on conventional state PD, a dynamic failure constitutive model of rock materials
and a PD approach to crack propagation in a fractured rock mass under a blasting load were investigated. The influence of the stress
wave incident angle and ground stress state on crack propagation was studied. This provided a design reference for controlling the
stability of the surrounding rock during construction blasting, such as in tunnel engineering.

2. PD constitutive function of rock materials

Since the establishment of the PD theory, numerous scholars have conducted extensive research on the constitutive functions of
various materials. Silling [33–34] had already proposed constitutive models of linear elastic and isotropic materials and proposed the
constitutive force function of a prototype microelastic brittle material (PMB). The effectiveness of the constitutive model of this PMB
material has been verified through numerical simulations and has been widely utilised. Under normal circumstances, rock materials
are damaged by small deformations and can be regarded as quasibrittleness. Therefore, this section first systematically introduces the
constitutive model of a PMB material. An appropriate constitutive model for rock materials is proposed based on the linear and
nonlinear mechanical responses during the failure of a rock material.

2.1. PD constitutive model for tensile and compressive composite failure of rock

The constitutive function of a microscopic elastic-brittle material employs the relationship between the periforce, f, between
material points and elongation s to characterise the linear mechanical behaviour during material failure. This is utilised to depict brittle
or quasi-brittle materials with approximately linear mechanical behaviours [35], as shown in Fig. 2.
A PMB material is an ideal substance possessing the following attributes.

(1) The periforces, f, generated between two material points are of equal magnitude and act along the line connecting the two
points in opposite directions.
(2) Only tensile and compressive deformations are considered in relation to the bond between two material points. During tensile
deformation, when elongation ratio s is greater than or equal to critical elongation ratio s0 , the bond between the material points
breaks and cannot be restored. Compression fractures are not considered in the model.
(3) The fracture of the bond between material points exhibits linear elastic characteristics, and the relationship between periforce f
(scalar) and elongation s is linear.

PD is used to discretise space domain R into a series of material points with volume V and mass density ρ, and analyse the motion of
an object by studying the interaction between centre point x and material point xʹ within a certain range. Material point x exists in

3
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 3. Interaction between material points in the neighbourhood.

Fig. 4. Tensile or compressive stress-strain curves of rock materials.

neighbourhood Hx with radius δ and interacts with other material points, xʹ, in the neighbourhood (as shown in Fig. 3). There is no
interaction between material points outside the neighbourhood. The material points in the neighbourhood of material point xi can be
expressed as follows:
{ }
Hxi = H(x, δ) = xj ∈ R : ‖ xi − xj ‖ ≤ δ (1)

Remember that x and xʹ are the position vectors of material points in the reference configuration, and u and uʹ are the displacement
vectors of material points x and xʹ, respectively, at time t. Then, the relative position vectors and relative displacement vectors of the
two material points can be expressed as ξ = xʹ − x and η = uʹ − u, respectively.
For PMB materials, the constitutive force function can be expressed as follows:
⎧
⎨ cs η + ξ s ≤ s0
⎪
f(η, ξ) = |η + ξ| (2)
⎪
⎩
0 other

6E
where c is the microscopic modulus, which is denoted as c = [πδ4 (1− 2υ)]
; υ is the Poisson’s ratio of a material; and s is the elongation of the
bond, as denoted by s = |η+ξ|− |ξ|
|ξ| .
Scalar function μ(t, ξ) is usually used to describe whether the bond is broken or not, taking a value of 1 or 0. When μ(t, ξ) = 1, the
bond is complete, but when μ(t, ξ) = 0, the bond is broken. The equation is as follows:
{
1 if s < s0
μ(x, t, ξ) = (3)
0 otherwise

In PD, the formation of cracks in a material represents a macroscopic manifestation of bond fracture between material points. The
occurrence of cracks is typically characterised by the ratio of the number of broken bonds, D, in the vicinity of a material point to the

4
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 5. Constitutive force function of rock material.

total number of bonds, N, in the vicinity, where φ can be used to express this ratio as follows:
∫
D μ(x, t, ξ)dVξ
φ(x, t) = = 1 − Hx ∫ . (4)
N Hx
dVξ

In the above equation, φ ∈ [0, 1], where 0 indicates no damage and 1 indicates complete damage.
Generally, whether a material behaves in a brittle or plastic manner depends on whether plastic deformation occurs and the extent
of its occurrence prior to crack propagation during the loading failure process of the material [36]. Rock materials are brittle on a
macroscopic scale, whereas plastic strain exists in microcracks formed by external forces on a microscopic scale. However, the plastic
strain on the microscopic scale is extremely small, especially under dynamic loading, and the assumption can be made that no plastic
strain exists on the macroscopic scale. Therefore, from an engineering perspective, it is scientifically rational to consider rocks as brittle
or quasibrittle materials with brittle macroscopic fractures [37]. In addition, the stress-strain curves of rocks under dynamic conditions
are similar to those under static conditions. The tensile or compressive stress-strain curves of rock materials are shown in Fig. 4.
As shown in Fig. 4, the stress-strain curves of rock materials under the influence of tension and compression can be approximately
divided into three phases.

(1) Linear elastic deformation stage (OA): The stress and strain are approximately linearly proportional, and the material undergoes
linear elastic deformation without cracks.
(2) Nonlinear strengthening stage (AB): The stress and strain exhibit a nonlinear positive growth relationship, and microcracks
gradually form and steadily expand.
(3) Nonlinear softening stage (BC): The stress decreases as the strain increases, which results in macroscopic cracks, intensified
unstable expansion, and even fracture failure.

Therefore, the damage and failure processes of rock materials encompass both linear and nonlinear mechanical behaviours.
However, the PMB material constitutive model is only applicable to an ideal brittle material, disregarding the nonlinear mechanical
behaviour and compression failure in the material, which is inconsistent with the actual mechanical characteristics of rock materials.
Therefore, this section presents a PD constitutive model of rock materials. Corresponding to the above distinct mechanical behaviour
stages of rock materials, the material point damage variable function is set as ψ (ψ ∈ [0, 1]), and its expression is as follows:
⎧
⎨0 (Linear elastic segment)
ψ = ψ (s) (Nonlinear segment) (5)
⎩
1 (Rupture failure)

where ψ is incorporated into force f(η, ξ) to delineate the distinct mechanical behaviours of rock materials under tensile and
compressive stresses. Then, periforce f can be revised as follows:
⎧
⎨ cs(1 − ψ ) η + ξ (|s| ≤ s0 )
⎪
f(η, ξ) = |η + ξ| (6)
⎪
⎩
0 (|s| > s0 )

5
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

In PD, the relationship between periforce f and elongation s characterises the failure behaviour of a material, whereas physical tests
show the relationship between tensile or compressive stress σ and strain ε. Hence, to acquire damage variable function ψ , the f-s
constitutive structure can be equated with the σ -ε curve, and the PD constitutive force function curve of a rock material can be obtained
as shown in Fig. 5.
In Fig. 5, sy represents the critical elongation at compression break, soy indicates the critical elongation of compressive strength-
ening with maximum f, sey represents the critical elongation of linear elasticity under compression, sl represents the critical elongation
at tensile failure, sol indicates the critical elongation of tensile strengthening with maximum f, and sel represents the critical elongation
of linear elasticity.
According to Fig. 5 and Eq. (5), the expression of the rock material behaviour in a nonlinear segment, ψ (s), can be obtained through
mathematical fitting as follows:
⎧
⎪
⎪ sey − s
⎪
⎪ soy
⎨ 1 − esy− sey sy ≤ s < sey
ψ (s) = (7)
⎪ sel − s
⎪ sol
⎩ 1 − esl − sel sel < s ≤ sl
⎪
⎪

Therefore, by integrating Eqs. (5) and (7), damage variable function ψ of a rock material in PD can be derived as follows:
⎧
⎪
⎪
⎪
⎪ 1 s < sy
⎪
⎪
⎪
⎪ sey − s
⎪
⎪ s
sy− sey oy
⎪
⎪
⎪ 1− e sy ≤ s < sey
⎨
ψ= 0 sey ≤ s ≤ sel . (8)
⎪
⎪
⎪
⎪ sel − s
⎪ s
⎪
⎪
⎪
⎪ 1− esl − sel ol sel < s ≤ sl
⎪
⎪
⎪
⎪
⎩ 1 sl < s

By substituting Eq. (8) into Eq. (6), the constitutive force function of the PD model for medium rock materials can be obtained as
follows:
⎧
⎪
⎪
⎪
⎪ 0 s < sy
⎪
⎪
⎪
⎪ η + ξ sey − s s
⎪
⎪ oy
⎪
⎪
⎪ cs esy− sey sy ≤ s < sey
⎪
⎪ |η + ξ|
⎪
⎪
⎨
η+ξ
f(η, ξ) = cs sey ≤ s ≤ sel . (9)
⎪
⎪ | η + ξ|
⎪
⎪
⎪
⎪
⎪
⎪ cs
η + ξ ssel− −s s sol
⎪
⎪
⎪ e l el sel < s ≤ sl
⎪
⎪ | η + ξ|
⎪
⎪
⎪
⎪
⎩ 0 sl < s

2.2. Fracture criterion for rock materials

In the PMB model, when the tensile elongation of the bond between two material points satisfies s > s0 , the bond is considered
broken. However, the damage and failure processes of rock materials include not only tensile failures but also compressive failures, and
the failure processes exhibit both linear and nonlinear mechanical behaviours. Therefore, it is necessary to establish fracture criterion
that can be applied to rock materials.
The previous section introduced various critical elongation rates to establish the constitutive force function of rock materials. The
following considers tensile fracture as an example to elaborate on the fracture process of rock materials.

(1) If s < sel , the material is in the linear elastic tensile stage, the bonds between the material points are undamaged, and no cracks
form.
(2) If sel ≤ s < sol , the material is in the nonlinear strengthening stage, and periforce f and bond elongation s show nonlinear
increased. The bonds gradually undergo damage, and microcracks occur in the material.
(3) If sol ≤ s < sl , the material is in the nonlinear softening stage, periforce f decreases, elongation s continues to increase, and
microcracks continue to occur in the material.
(4) If sl < s, the material reaches the failure stage, the bond between the material points breaks, periforce f vanishes, and micro-
cracks accumulate to form macroscopic cracks.

The compression fracture process is the same as that described above. The relationship between the amount of bond damage and

6
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 6. Relationship between damage and elongation.

the elongation is shown in Fig. 6. Thus, it can be observed that in the PD model of rock materials, when s < sy or sl < s is fulfilled, the
bonds between material points can be regarded as permanently damaged.

2.3. Program implementation

Initialising the parameters and discrete material points, solving the interaction force between material points, and determining the
damage of material points are indispensable when simulating the dynamic failure issue using PD, as shown in Fig. 7. The specific
procedure is as follows:

(1) Parameter initialisation: Geometric parameters such as the model size and calculation parameters such as the neighbourhood
range, time step, and material point size are determined based on the research objective. Then, the physical and mechanical
parameters are set, along with the basic variables of the material.
(2) Discrete material points: The research object is dispersed into a limited number of material points containing physical infor-
mation. The virtual boundary layer (typically three layers of material points) is set in accordance with the research re-
quirements, and the material points are numbered successively.
(3) Determine the neighbourhood range: The material points within the neighbourhood range of each material point are retrieved
individually. The information of the material points within the neighbourhood range of each material point is stored, and the
bond force between the material point and material points within the neighbourhood range is initialised.
(4) Define the fracture surface: The material points on both sides of the fracture surface are retrieved, and the material points
crossing the fracture surface are set as failures based on the fracture geometry information.
(5) The volume and surface of the material points are corrected.
(6) Setting boundary conditions: Load and boundary conditions are applied based on the operating condition design. Typically, a
load is applied to the virtual layer.
(7) Integrate the velocity, displacement, and other parameters of all the material points, and conduct continuous iteration.
(8) Calculate the relative deformation of each material point in its neighbourhood and obtain the elongation of the bond between
the material points.
(9) Determine whether the deformation between material points enters a nonlinear stage. Otherwise, the point-to-point force is
calculated directly. If so, we consider the linear elastic deformation and nonlinear deformation simultaneously, and then
calculate the point-to-point force between the material points.
(10) Calculate the elongation of the material and material points in the neighbourhood. If the critical elongation is exceeded, the
force-bond breaks. Otherwise, the force-bond does not break.
(11) Determine whether all the material points in the neighbourhood have undergone the procedures in steps 8 and 9. If so, proceed
to the next step. Otherwise, we return to step 8 and continue the calculations.
(12) Determine whether the calculations of the bonds between all the material points in the calculation domain have been
completed. If so, the next time-step cycle begins. Otherwise, we return to step 8 and continue the calculations.
(13) Check whether the calculation has attained the specified time step. If yes, the calculation is terminated. Otherwise, we return to
step 8 and continue the calculations.

7
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 7. Flow chart for calculating tensile and compressive compound failure in rock.

3. Effect of stress-wave incident angle on blasting crack propagation

The incident angle of the explosion stress wave generated during construction blasting significantly influences the propagation
paths of cracks. The influence of the incident angle of the stress wave on the initiation and expansion angles of crack propagation was
explored, with the goal of optimising the blasting design and ensuring construction safety. This section presents the results of a

8
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 8. Prefabricated crack blasting sample.

Fig. 9. Boundary conditions.

simulation study on the propagation of blasting cracks in fractured rock masses with various incident angles using plexiglass or
polymethyl methacrylate (PMMA) as the test material. The results of a PD numerical simulation were compared with those of physical
tests reported in the literature [38] to validate the efficacy of the PD rock material constitutive model established in this study under a
simulated dynamic load.
To determine whether plexiglass could substitute for rock materials, Rammanith et al. [39] conducted dynamic test studies on
PMMA and rock materials and demonstrated that the fracture behaviours of PMMA and rock materials were analogous under dynamic
loads. Additionally, PMMA possesses excellent light transmission, which enables the clear observation of crack growth when it is
utilised to conduct rock blasting tests.

3.1. Numerical simulation design

As shown in Fig. 8, the test material sample was 400 mm × 400 mm × 6 mm (length × width × height) in size. The length of the
prefabricated crack was 30 mm, and its width was 1 mm. The diameter of the blasthole was 8 mm, the distance between the centre of
the blasthole and the crack tip was 60 mm, and the angle, β, between the connection line and horizontal direction was the incident
angle.

9
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Table 1
Prefabricated crack simulation conditions.
Operating conditions The distance between the blasthole and right end of the crack (mm) Incident angle of explosion stress wave (◦ )

#1 60 0
#2 60 15
#3 60 30
#4 60 45
#5 60 60
#6 60 75
#7 60 90

Fig. 10. Definition of crack propagation parameters.

Fig. 11. Stress diagram of crack tip at different incident angles.

The material points formed a three-dimensional grid, with 200, 3, and 200 layers in the X, Y, and Z directions, respectively. After
removing the material points in the prefabricated crack and blasthole, the total number is 119,900. The distance between the material
points was 0.002 m, the radius of the neighbourhood was 0.006 m, tensile critical elongation sl = 0.00015, and compressive critical
elongation sy = − 0.0015. Each timestep was 0.1 μs. The material parameters included a Young’s modulus of 2940 MPa, material
density of 1200 kg/m3 , and Poisson’s ratio of 0.25. As shown in Fig. 9, a triangular blasting load was applied. The peak value of the

10
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 12. Blasting crack propagation with different incident angles.

blasting load was 100 MPa, the duration was 100 μs, and the peak value of the blasting load was reached when t = 10 μs.
A total of seven distinct conditions for incident angle β of the explosive stress wave were set, and the specific parameters are
presented in Table 1.

3.2. Simulation results of blasting crack growth

To facilitate the analysis of the numerical simulation results, the relevant parameters of the blasting crack were initially defined as
follows. As shown in Fig. 10, the crack that expanded towards the blasting hole was main burst crack a, and the crack generated by the
expansion near the crack tip of the blasting hole was near-end crack b. The crack far away from the blasting hole was far-end crack c.
Propagation angle φ was defined as the angle between the line passing through the crack point and first inflection point and the left side
of the horizontal line, whereas near-end crack angle γ was defined as the angle formed between the near-end crack and right side of the
prefabricated crack. The angle between the distal crack and left side of the prefabricated crack was far-end crack angle θ [38].

11
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 12. (continued).

As can be observed in Fig. 11, within the sample featuring prefabricated cracks, two types of stress were generated at the crack tip
owing to the blasting stress. First, when the explosion stress wave was parallel to the prefabricated crack, the stress wave underwent
reflection and diffraction at the crack, which produced tensile stress σ, as depicted in Fig. 11(a). Second, when there was an angle
between the explosion stress wave and prefabricated crack, the stress wave passed through and was reflected upon encountering the
crack. Shear stress τ formed in the passing portion, and tensile stress σ formed in the reflecting portion. Shear stress τ and tensile stress σ
jointly induced wing cracks at the crack tip, as illustrated in Fig. 11(b)–(d).
PD simulation damage diagrams of the blasting failure mode and crack growth path of the specimen under different operating
conditions are shown in Fig. 12.
Based on the above crack propagation graphs under different incident angles, it can be seen that when incident angle β was 0◦ , the
explosion stress wave propagated in the direction parallel to the prefabricated crack. The stress wave initially reached the proximal end
of the crack, and the presence of local cracks led to the reflection and diffraction of the stress wave. The reflection generated a reflected
tensile stress wave at the proximal end of the crack, which induced the proximal end crack, spread towards the explosion source, and
connected with the main crack. Diffraction caused the formation of a reflected tensile stress wave at the distal end of the prefabricated
crack and induced a crack at the distal end to propagate away from the explosion source. Eventually, penetrating explosive cracks,
proximal cracks, prefabricated cracks, and distal cracks formed.
When incident angle β was 15◦ or 30◦ , the blasting stress wave induced two wing cracks at the ends of the prefabricated crack.
Under the combined effects of shear stress τ and tensile stress σ , the proximal wing cracks propagated towards the explosion source,
whereas the distal wing cracks propagated away from the explosion source. As incident angle β increased from 15◦ to 30◦ , the influence
of shear stress τ on the initiation and propagation of the wing crack at the crack tip became increasingly significant. Therefore, near-
end crack angle γ, far-end crack angle θ, and propagation angle φ of the wing crack increased with incident angle β.
When incident angle β was 45◦ or 60◦ , the blasting stress wave induced three wing cracks at the ends of the prefabricated crack,
including one proximal end crack and two distal end cracks (one propagating in the positive direction and one propagating in the
negative direction of the Z axis). The distal crack extending in the negative Z direction was shorter and emerged at the end of blasting,

12
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 13. Comparison of results of sample blasting crack propagation test and PD simulation.

mainly as a result of the increase in incident angle β. This enhanced tensile stress σ , which was formed by reflection, resulting in the
failure of rock materials in the direction of tensile stress σ .
When incident angle β was 75◦ or 90◦ , four wing cracks were generated on the sides of the prefabricated crack. It can be observed
that the two distal cracks expanded in the positive and negative directions of the Z axis, with the length of the distal crack in the

13
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 13. (continued).

negative direction of the Z axis being shorter. In contrast to the previous results, when incident angle β ≥ 75◦ , the distal end of the crack
was closer to the damage area of the initiation point, and the explosive crack generated by blasting expanded to the crack and passed
through the distal crack. Similarly, one of the two proximal cracks was caused by tensile stress σ and shear stress τ and extended to the
explosion source, while the other proximal crack was mainly formed by the expansion of explosive cracks. Additionally, it can be
observed from the crack growth process that the crack growth length at the proximal end was very small. With an increase in the
incident angle, tensile stress σ and shear stress τ were rotated perpendicularly to the prefabricated crack until incident angle β = 90◦ .
Thus, tensile stress σ and shear stress τ were almost collinear, and the resultant force reached its minimum value. Therefore, the tip
stress was the least destructive.
Fig. 13 presents a comparison between the PD simulation results and physical test results based on the rock material constitutive
model proposed in this study. The effect of the PD method on crack propagation in the rock material was consistent with the
experimental results using a PMMA plate.
As shown in Fig. 14, the correlation between near-end crack angle γ and distal initiation angle θ varied with incident angle β. A
comparative analysis showed that the results of the PD simulation in this study essentially agreed with the physical test results reported
in the literature [38]. These results demonstrated a pattern where near-end crack angle γ initially rose and then subsequently dropped
with an increase in incident angle β. Far-end crack angle θ increased with incident angle β, but the increase in the distal initiation angle
gradually slowed down, and the graph became flat.
The correlation between incident angle β and propagation angle φ is shown in Fig. 15. It can be observed that the PD simulation
results obtained in this study essentially agreed with the physical test results in the literature [38], which demonstrated the following
relationships. Propagation angle φ increases with incident angle β. This increase was basically linear when incident angle β was less

14
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 14. Relationships between incident angle β and initiation angles γ and θ.

Fig. 15. Relationship between incident angle β and extended angle φ.

than 75◦ . When incident angle β exceeded 75◦ , the growth rate declined rapidly, and the graph became flat. Additionally, propagation
angle φ was consistently smaller than stress wave incident angle β.
In summary, this section compared the results of crack growth experiments under dynamic explosive loads from the literature [38]
in comparison to the results of a simulation using the PD method established in this study. The results of the PD simulation and physical
tests were very consistent in terms of the crack growth path, crack initiation angle, and crack expansion angle change criterion.
Consequently, it was demonstrated that the PD constitutive model of rock materials established in this study was effective at simulating
the dynamic failure of fractured rocks.

4. Effect of ground stress on crack propagation in rock blasting

A rock mass exists within a geological environment, and ground stress results from gravity, the geological structure, and other
environmental factors. Generally, as the buried depth of a rock mass increases, the ground stress level of the rock mass also increases
significantly [40]. Under high ground stress, the mechanical properties of the rock mass undergo significant changes such as an in-
crease in compressive strength. In deep underground tunnel engineering, rock is simultaneously subjected to a static load from the
initial ground stress and a dynamic load from excavation blasting. The initial ground stress often has a crucial influence on explosion
stress wave propagation, crack propagation, and rock mass failure [41]. Numerous studies have indicated that the main reason for the
damage zone of the surrounding rock in deep tunnel excavation are unloading of initial stress field and excavation methods [42–43].
Simultaneously, a considerable amount of engineering practice has revealed that the initial ground stress state has a significant impact

15
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 16. Schematic diagram of rock blasting simulation.

Fig. 17. Boundary conditions of rock blasting simulation.

on the dynamic disturbance results for a rock mass [44–45]. Hence, it is important to consider the initial ground stress state when
studying the criterion for rock blasting crack propagation.
This section shows how the criterion of rock blasting crack propagation under a unidirectional initial earth stress were simulated,
and the PD simulation results are compared with model test results from the literature [40] to verify the effectiveness of the PD
simulation method for rock mass damage under dynamic disturbances established in this study.

4.1. Numerical simulation design

As shown in Fig. 16, the dimensions of the model were 300 mm × 300 mm × 20 mm (length × width × height), and the diameter of
the blasthole was 4 mm. The material points were arranged in a three-dimensional grid with 150, 10, and 150 layers in the X, Y, and Z
directions, respectively. Three virtual boundary layers were set around the X–Z plane, with a total of 242,880 material points. The
distance between the material points was 2 mm, and the radius of the neighbourhood was 6 mm. The critical tensile elongation was

16
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Table 2
Simulated conditions of rock blasting simulation.
Loading conditions Z-direction confining pressure, X-direction confining pressure, Blasthole diameter
σv (MPa) σh (MPa) (mm)

#1 1.2 0 4
#2 3.1 0 4
#3 4.8 0 4

Fig. 18. Crack propagation under different ground stresses.

0.0001, critical compressive elongation was − 0.001, and time step was 0.1 μs. The elastic modulus of the model was 0.552 GPa, bulk
density was 12.3 kN/m3 , Poisson’s ratio was 0.28, tensile strength was 0.76 MPa, and compressive strength was 10.1 MPa.
As shown in Fig. 17, a triangular blasting load was applied. The peak value of the blasting load was 25 MPa, and its duration was
100 μs. The peak blasting load occurred at t = 20 μs. The initial ground stress was applied to the positive and negative virtual boundary
layers of the Z axis using a static load (red boundary in Fig. 14(b)). Three distinct confining pressure (σv ) loading conditions were
established, as listed in Table 2.

4.2. Analysis of blasting crack simulation results

Fig. 18 shows the crack propagation process and final failure patterns of the specimen when t = 10 μs, 20 μs, 60 μs, 100 μs and
confining pressure σ v = 1.2 MPa, 3.1 MPa, 4.8 MPa. As can be observed from the figures, when the initial unidirectional ground stress
was applied, most of the principal radial cracks in the model propagated in the direction of this initial ground stress loading (the σ v
direction), with few generated in the horizontal direction (the σ h direction). This implied that the maximum principal stress induced
crack propagation and that the main cracks typically propagated along the direction of the maximum principal stress. This was

17
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 19. Comparison of crack propagation in PD simulation and physical test.

consistent with the results of previous studies [46]. Additionally, with an increase in the initial vertical ground stress (σ v ), the sec-
ondary cracks gradually diminished, and the angle between the main crack and z axis gradually decreased until σ v = 4.8 MPa, at which
time the main cracks basically spread in the vertical direction.
Fig. 19 presents a comparison between the PD simulation results and model test results for crack growth under blasting dynamics
when σ v = 1.2 MPa, σ v = 3.1 MPa, and σv = 4.8 MPa. According to Table 2 and Fig 19, it can be observed that under loading
conditions #1 and #2, three radial principal cracks were formed in both the PD simulation and model test results. Moreover, the angle
between the principal cracks under loading condition #2 was smaller than that under loading condition #1. However, only two radial
principal cracks were observed under loading condition #3. The results of the PD simulation fundamentally agreed with those of the
model tests.
Fig. 20 shows a comparison of the influences of the unidirectional initial ground stress on the size of the crushing zone under the
three loading conditions. In the model test results, the blue line indicates the crushing zone and red line represents the fracture zone.
Based on the statistics from the PD simulation results, when confining pressure σv was 1.2 MPa, 3.1 MPa, and 4.8 MPa, the diameter of
the crushing zone was 4.11 cm, 4.32 cm, and 4.69 cm, respectively. In other words, the size of the crushing zone gradually increased
with the increase of the initial ground stress. This was consistent with the conclusions derived from model tests in the literature [47].

18
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Fig. 20. Fracture and crushing zones of rock mass.

Various factors may have contributed to this result. On one hand, the application of initial stress σ v causes changes in the internal stress
of the specimen, and more energy generated by explosive initiation is used to form the crushing zone. On the other hand, because the
specimen belonged to a plane model and could not be blocked, a large amount of energy escaped. Thus, although the crushing zone was
broken, it did not collapse. Additionally, it can be observed from Fig. 20 that with an increase in ground stress, the length and number
of secondary cracks generated around the explosion source decreased, which may have been due to the effect of the initial ground
stress. The mechanical properties of the rock mass will have been improved to a certain extent, and its integrity may have been
enhanced. Therefore, the initial ground stress had an inhibitory effect on the initiation and spread of blasting cracks in the rock mass,
and a greater ground stress strengthened this inhibitory effect on rock mass cracks.
Based on the PD simulation results, the diameters of the fracture zone along the directions of σ v and σh were computed and
compared with the model test results, as presented in Table 3.
As can be observed from Table 3, under the three loading conditions, the diameter of the fracture zone in the direction of σ v was
consistently smaller than that in the direction of σ h . Moreover, with an increase in the ground stress, the extent of the fracture zone
gradually contracted. This was attributed to the fact that the application of ground stress in the direction of σv mainly restricted the
expansion of the fracture zone in the direction of σ v and had an impact on the expansion in the direction of σ h . This agreed with the
results of the model tests in the literature.
To summarise, this section compared the results of model tests of blasting crack growth under the influence of an initial ground
stress from the literature [45] with those obtained when the PD method established in this study was employed for simulations. The
results validated the efficacy of the PD constitutive model of rock materials established to simulate the dynamic failure of fractured

19
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

Table 3
Comparison of fracture zone diameters.
Loading conditions In the direction of σv (cm) In the direction of σh (cm)

Model experiments [47] PD simulation Model experiments [47] PD simulation

#1 8.20 8.30 8.40 8.51

#2 7.30 7.36 7.50 7.60
#3 7.20 7.08 7.40 7.28

rocks.

5. Conclusion

(1) The mechanism of rock blasting failure, mechanical behaviour characteristics of rock, and classical PMB constitutive model
based on the early PD theory were initially introduced. The linear and nonlinear mechanical behaviours of rock materials during
the failure process were depicted by introducing damage variable function ψ, and compression failure was considered. An
appropriate PD constitutive model for analysing rock material failure was established. Subsequently, the criterion for the
growth of a blasting crack in a fractured rock mass induced by the incident angle of the stress wave and the criterion for the
growth of a blasting crack in rock caused by the in-situ stress state were investigated.
(2) When the incident angle of an explosion stress wave impacting a prefabricated crack was zero, the new cracks that were
dynamically induced propagated parallel to the prefabricated crack. When the incident angle of the explosion stress wave was
between 15◦ and 75◦ , wing cracks were formed on both sides of the prefabricated crack under the action of the stress wave, and
the crack initiation angle and expansion angle increased with the incident angle.
(3) The in-situ stress in the rock affected the propagation path of the explosive cracks and range of the fracture zone. When a
unidirectional in-situ stress was applied, the radial main crack propagated along the direction of the in-situ stress, and the range
of the fracture zone decreased with an increase in the in-situ stress. The results demonstrated that the PD simulation results were
consistent with previous experimental results. This verified the effectiveness and accuracy of the PD method in simulating
fractured rock damage and failure under dynamic disturbances.

CRediT authorship contribution statement

Liping Li: Writing – original draft, Methodology, Data curation, Investigation, Software, Funding acquisition. Xiaochu Chen:
Writing – review & editing, Supervision, Conceptualization. Chenglu Gao: Conceptualization, Supervision, Investigation, Funding
acquisition. Zongqing Zhou: Data curation, Visualization, Validation, Funding acquisition. Minghao Li: Software, Data curation.
Daosheng Zhang: Resources, Visualization. Jinbo Chen: Formal analysis, Validation.

Declaration of competing interest

The authors declare the following financial interests/personal relationships which may be considered as potential competing
interests.

Acknowledgments

Much of the work presented in this paper was supported by the National Key R&D Program of China (2022YFC3005603), the
National Natural Science Foundation of China (52025091, 52279104, 52309135), Natural Science Foundation of Shandong Province
(ZR2023JQ019, ZR2022QE180), the “20 New Universities” Project of Jinan City (202333049), Taishan Scholars Program, and Young
Talent of Lifting engineering for Science and Technology in Shandong, China.

Data availability

Data will be made available on request.

References

[1] Y Zhang, J Zhang, S Wang, Y Chen, Controlling vibration speed in tunnel excavation using fine blasting method under complex environmental conditions, Shock
Vib. 2021 (2021) 6060485, https://doi.org/10.1155/2021/6060485.
[2] S. Li, S. Li, Q. Zhang, Y. Xue, B. Liu, M. Su, Z. Wang, S. Wang, Predicting geological hazards during tunnel construction, J. Rock Mech. Geotech. Eng. 2 (2010)
232–242, https://doi.org/10.3724/SP.J.1235.2010.0023.
[3] Y. Fan, J. Chen, Z. Leng, G. Yang, X. Liu, B. Tian, Punching mechanism of air-deck stemming for drilling blasting and its influence on rock fragmentation, Rock
Mech. Rock Eng. (2024), https://doi.org/10.1007/s00603-024-03773-1.
[4] Y. Fan, G.D. Yang, X.Z. Cui, X.H. Zhao, J.G. Wu, L.H. Wang, Low-frequency characteristics of vibrations induced by transient unloading of in situ stress and its
influence on safety of a deeply buried tunnel, Int. J. Geomech. (2024), https://doi.org/10.1061/IJGNAI.GMENG-819.

20
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

[5] S.A Silling, Reformulation of elasticity theory for discontinuities and long-range forces, J. Mech. Phys. Solid. 48 (2000) 175–209, https://doi.org/10.1016/
S0022-5096(99)00029-0.
[6] C. Gao, Z. Zhou, Z. Li, L. Li, S. Cheng, Peridynamics simulation of surrounding rock damage characteristics during tunnel excavation, Tunn. Undergr. Space
Technol. 97 (2020) 103289, https://doi.org/10.1016/j.tust.2020.103289.
[7] C. Gao, L. Li, Z. Zhou, Z. Li, S. Cheng, L. Wang, D. Zhang, Peridynamics simulation of water inrush channels evolution process due to rock mass progressive
failure in karst tunnels, Int. J. Geomech. 21 (2021) 04021028, https://doi.org/10.1061/(ASCE)GM.1943-5622.0001980.
[8] H Ouchi, Development of Peridynamics-Based Hydraulic Fracturing Model For Fracture Growth in Heterogeneous Reservoirs, University of Texas Libraries,
2016. http://hdl.handle.net/2152/40966.
[9] S.A Silling, M Epton, O Weckner, J Xu, Peridynamic states and constitutive modeling, J. Elast. 88 (2007) 151–184, https://doi.org/10.1007/s10659-007-9125-
1.
[10] S.A Silling, E Askari, A meshfree method based on the peridynamic model of solid mechanics, Comput. Struct. 83 (2005) 1526–1535, https://doi.org/10.1016/j.
compstruc.2004.11.026.
[11] T Vogler, C.J Lammi, A Nonlocal, Peridynamic Plasticity Model For the Dynamic Flow and Fracture of Concrete, Sandia National Lab.(SNL-CA), Livermore, CA
(United States), 2014, https://doi.org/10.2172/1159446.
[12] F Bobaru, J.T Foster, P.H Geubelle, S.A Silling, Handbook of Peridynamic Modeling, CRC press, 2016, https://doi.org/10.1201/9781315373331.
[13] J.A Mitchell, A nonlocal, ordinary, State-Based Plasticity Model For Peridynamics, Sandia National Laboratories (SNL), Albuquerque, NM, and Livermore, CA
(United States), 2011, https://doi.org/10.2172/1018475.
[14] B Kilic, E Madenci, Coupling of peridynamic theory and the finite element method, J. Mech. Mater. Struct. (2014) 191–202, https://doi.org/10.1007/978-1-
4614-8465-3_11.
[15] E Oterkus, Peridynamic Theory for Modeling Three-Dimensional Damage Growth in Metallic and Composite Structures, The University of Arizona, 2010.
https://repository.arizona.edu/handle/10150/145366.
[16] G.C Ganzenmüller, S Hiermaier, M May, Improvements to the Prototype Micro-Brittle Model of peridynamics, Meshfree methods For Partial Differential
Equations VII, Springer International Publishing, 2015, pp. 163–183, https://doi.org/10.1007/978-3-319-06898-5_9.
[17] B Kilic, Peridynamic Theory For Progressive Failure Prediction in Homogeneous and Heterogeneous Materials, The University of Arizona, 2008. https://api.
semanticscholar.org/CorpusID:135671908.
[18] E Madenci, E Oterkus, Peridynamic theory, Peridynamic theory and Its Applications, Springer New York, New York, NY, 2013, pp. 19–43, https://doi.org/
10.1007/978-1-4614-8465-3_2.
[19] J.B. Martino, N.A. Chandler, Excavation-induced damage studies at the Underground Research Laboratory, Int. J. Rock. Mech. Min. 41 (2004) 1413–1436,
https://doi.org/10.1016/j.ijrmms.2004.09.010.
[20] M Kononenko, O Khomenko, New theory for the rock mass destruction by blasting, Min. Miner. Deposits. 15 (2021) 111–123, https://doi.org/10.33271/
mining15.02.111.
[21] P.A Persson, R Holmberg, J Lee, Rock Blasting and Explosives Engineering, CRC press, 2018, https://doi.org/10.1201/9780203740514.
[22] X. Huo, Y. Jiang, W. Wei, X. Qiu, Z. Yu, J. Nong, Q. Li, Three-dimensional finite element simulation and reconstruction of jointed rock masses for bench blasting,
Simul. Model. Pract. Theory 135 (2024) 102975, https://doi.org/10.1016/j.simpat.2024.102975.
[23] Z. Zhang, W. Gao, K. Li, B. Li, Numerical simulation of rock mass blasting using particle flow code and particle expansion loading algorithm, Simul. Model.
Pract. Theory 104 (2020) 102119, https://doi.org/10.1016/j.simpat.2020.102119.
[24] G.W Ma, X.M An, Numerical simulation of blasting-induced rock fractures, Int. J. Rock Mech. Min. Sci. 45 (2008) 966–975, https://doi.org/10.1016/j.
ijrmms.2007.12.002.
[25] A Fakhimi, M Lanari, DEM–SPH simulation of rock blasting, Comput. Geotech. 55 (2014) 158–164, https://doi.org/10.1016/j.compgeo.2013.08.008.
[26] M Zaid, S Mishra, K.S Rao, in: Stability of different shapes of Himalayan tunnels under blast loading, 8th Indian Rock Conf., New Delhi, 2019, pp. 375–380.
[27] S.A Silling, Dynamic Fracture Modeling With a Meshfree Peridynamic code, Computational Fluid and Solid Mechanics 2003, Elsevier Science Ltd, 2003,
pp. 641–644, https://doi.org/10.1016/B978-008044046-0.50157-3.
[28] Y.D Ha, F Bobaru, Studies of dynamic crack propagation and crack branching with peridynamics, Int. J. Fract. (2010) 229–244, https://doi.org/10.1007/978-
90-481-9760-6_18.
[29] X Zhou, Y Wang, Q Qian, Numerical simulation of crack curving and branching in brittle materials under dynamic loads using the extended non-ordinary state-
based peridynamics, Eur. J. Mech. A-Solid. 60 (2016) 277–299, https://doi.org/10.1016/j.euromechsol.2016.08.009.
[30] F Zhu, J Zhao, Peridynamic modelling of blasting induced rock fractures, J. Mech. Phys. Solid. 153 (2021) 104469, https://doi.org/10.1016/j.
jmps.2021.104469.
[31] Y. Zhang, J. Deng, H. Deng, B. Ke, Peridynamics simulation of rock fracturing under liquid carbon dioxide blasting, Int. J. Damage Mech. 28 (2019) 1038–1052,
https://doi.org/10.1177/1056789518807532.
[32] B Kilic, Peridynamic Theory for Progressive Failure Prediction in Homogeneous and Heterogeneous Materials, The University of Arizona, 2008. https://api.
semanticscholar.org/CorpusID:135671908.
[33] S.A Silling, M Epton, O Weckner, J Xu, Peridynamic states and constitutive modeling, J. Elast. 88 (2007) 151–184, https://doi.org/10.1007/s10659-007-9125-
1.
[34] S.A Silling, E Askari, Peridynamic modeling of impact damage, in: ASME pressure vessels and piping conference 46849, 2004, pp. 197–205, https://doi.org/
10.1115/PVP2004-3049.
[35] Y.D Ha, F Bobaru, Studies of dynamic crack propagation and crack branching with peridynamics, Int. J. Fract. (2010) 229–244, https://doi.org/10.1007/978-
90-481-9760-6_18.
[36] J Lemaitre, A Course On Damage Mechanics, Springer science & business media, 2012. https://api.semanticscholar.org/CorpusID:136066146.
[37] Q.B. Zhang, J. Zhao, A review of dynamic experimental techniques and mechanical behaviour of rock materials, Rock Mech Rock Eng 47 (2014) 1411–1478,
https://doi.org/10.1007/s00603-013-0463-y.
[38] H. Rong, Z. Zheming, H. Zheyuan, K. Jiming, Experimental study of crack propagation under explosive dynamic load, Chin. J. Rock Mech. Eng. 32 (2013)
1476–1481, https://doi.org/10.3969/j.issn.1000-6915.2013.07.024 (in Chinese).
[39] H.P. Rossmanith, A. Daehnke, R.E.K. Nasmillner, N. Kouzniak, M. Ohtsu, K. Uenishi, Fracture mechanics applications to drilling and blasting, Fatigue Fract. Eng.
Mater. Struct. 20 (1997) 1617–1636, https://doi.org/10.1111/j.1460-2695.1997.tb01515.x.
[40] C Zhang, X.T Feng, H Zhou, Estimation of in situ stress along deep tunnels buried in complex geological conditions, Int. J. Rock Mech. Min. Sci. 52 (2012)
139–162, https://doi.org/10.1016/j.ijrmms.2012.03.016.
[41] G. Jinjin, Study On Dynamic Mechanical Properties of Cemented Sand Similar Materials Under Confining Pressur, M.S., Huainan: Anhui University of Science
and Technology, 2017 (in Chinese).
[42] S. Kwon, C.S. Lee, S.J. Cho, S.W. Jeon, W.J. Cho, An investigation of the excavation damaged zone at the KAERI underground research tunnel, Tunn. Undergr.
Space Technol. 24 (2009) 1–13, https://doi.org/10.1016/j.tust.2008.01.004.
[43] R.S Read, 20 years of excavation response studies at AECL’s underground research laboratory, Int. J. Rock Mech. Min. Sci. 41 (2004) 1251–1275, https://doi.
org/10.1016/j.ijrmms.2004.09.012.
[44] W. Lu, J. Yang, P. Yan, M. Chen, C. Zhou, Y. Luo, L. Jin, Dynamic response of rock mass induced by the transient release of in-situ stress, Int. J. Rock Mech. Min.
Sci. 53 (2012) 129–141, https://doi.org/10.1016/j.ijrmms.2012.05.001.

21
L. Li et al. Simulation Modelling Practice and Theory 140 (2025) 103079

[45] J. Yang, K. Liu, X. Li, Z. Liu, Stress initialization methods for dynamic numerical simulation of rock mass with high in-situ stress, J. Cent. South Univ. 27 (2020)
3149–3162, https://doi.org/10.1007/s11771-020-4535-3.
[46] T. Haofan, L. Zhen, B. Tai, Y. Shuai, D. Zhizhong, Study on rock blasting crack propagation under initial in-situ stress, Non-Ferr. Metal Eng. 12 (2022), https://
doi.org/10.3969/j.issn.2095-1744.2022.03.018, 138-146+159 (in Chinese).
[47] G. Jinjin, Model Test Study of Rock Blasting Crack Propagation Under Initial Stress state, Ph.D, Anhui University of Science and Technology, Huainan, 2020 (in
Chinese).

22