Journal of Physics: Conference Series

PAPER • OPEN ACCESS You may also like

- Historical evolution and new trends for
Dynamic spherical cavity expansion analysis of soil-intruder interaction modeling
Serena R M Pirrone, Emanuela Del
concrete/rock based on Hoek-Brown criterion Dottore and Barbara Mazzolai

- Modeling the self-penetration process of a

bio-inspired probe in granular soils
To cite this article: H Dong et al 2020 J. Phys.: Conf. Ser. 1507 032012 Yuyan Chen, Ali Khosravi, Alejandro
Martinez et al.

- Experimental and numerical investigation

of the Rayleigh-Taylor instability of the
Newtonian and dilatant fluid system
View the article online for updates and enhancements. A N Doludenko

This content was downloaded from IP address 194.154.201.2 on 10/01/2023 at 08:46

The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

Dynamic spherical cavity expansion analysis of concrete/rock

based on Hoek-Brown criterion

H Dong, H J Wu*, J Zh Li, A G Pi and F L Huang

State Key Laboratory of Explosion Science and Technology, Beijing Institute of
Technology, Beijing 100081

Email:3120170117@bit.edu.cn

Abstract. Cavity expansion theory has been widely applied to predict the penetration depth of
projectiles into geological materials，in which the yield criterion and the constitutive model are
combined to describe the shear and compaction of the material, separately. For a precise
description of the material behaviors in plastic for geological materials with varied
compressive strength, in this present manuscript, the Hoek-Brown yield criterion and the
Dilatant-Kinematic equation were introduced into the plastic region. A response region called
elastic-cracking-dilatant-compaction region for concrete and rock was established. Then the
effects of various parameters, including brittleness coefficient, material integrity coefficient
and dilatant coefficient, on the relationship between cavity pressure and normal expansion
velocity were discussed. Finally, a penetration depth formula of rigid projectiles into
semi-infinite concrete and rock targets was established. It showed that, compared with the
classical Forrestal model, the proposed model had a well-being as well as higher accuracy
prediction of the penetration depth of concrete and rock materials with various strengths.

1. Introduction
The research of resistance models for projectiles penetrating into concrete or rock targets has great
significance in the field of impact and defense. In the past 200 years, numerous empirical and
theoretical resistance models [1-4] have been developed, and detailed information can be found in the
review papers [5-7]. Among the numbers of theoretical models, the dynamic cavity expansion model
has been widely applied in impact engineering for its simplicity, as the normal stress on the surface of
a cavity is expressed as a quadratic function of the particle velocity. The pioneering work can be
dating back to 1945, Bishop [8] firstly calculated the static pressure required to enlarge a spherical
hole in a plastic-incompressible material. Subsequently, this theory was developed as dynamic cavity
expansion theory(DCE) by applying the Hugonit jump conditions and the conservation equations [9].
Since the constitutive model and yield criterion are combined to describe the material response in
the plastic region in DCE, selecting a precise yield criterion and a constitutive model to capture the
material response in penetration case is vital. In terms of the constitutive or EOS models, the

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

traditional HJC [10, 11] and P-alpha [12] model were selected to describe the compressive behavior of
concrete materials. Recently, some new models have been proposed to meet the demand of the
development of high-speed projectiles penetration. Wang [13] used the Bingham liquid constitutive
model to characterize incompressibility for concrete material suffering ultra-high pressures, such as in
the case of long rod hypervelocity penetration. Wu [14] introduced the Murnaghan equation to
describe the plastic behavior of concrete material within the pressure of 50GPa. As for the yield
criterion, the linear yield criteria were adopted in early stage to describe the material shear behaviors,
such as Tresca [15], Mohr-Coulomb [16] and Griffith [17] yield criterion. Lately, Bavdekar [18]
developed a generalized non-linear for brittle ceramic, named Extended-Mohr-Coulomb model. Wu
[14] improved the dynamic cavity expansion model by using a hyperbolic yield criterion for concrete
material. Those all above-mentioned constitutive models and yield criteria provide a good description
of the brittle material mechanical characteristics in impacting conditions. However, some
shortcomings, as shown below, should be considered when suing the DCE theory.
(1) The linear yield criteria, as depicted in [26], are not sufficient to capture the response of the
brittle materials. Although some non-linear yield criteria have been developed, the physical meaning
of the parameters in those models are not explicit, such as the parameter ‘k’ in
Extended-Mohr-Coulomb model [26]. This makes it difficult to be used in other brittle materials.
(2) The constitutive model and yield criterion are combined to describe the compaction and shear
behaviors of concrete material, separately. It means that the geological materials were treated as the
metals (no dilatancy flow in plastic region) referencing the methods in classical elastic-plastic
mechanics. However, As the material under high pressure has become granular or pulverized, the
concrete and the rock will undergo dilation and compaction.
(3) The existing models rarely verify its accuracy by the penetration experiments using geological
materials with varied compressive strengths. On the other hand, the concrete and rock with different
compressive strength have been widely used in practice, especially the development of the ultra-high
performance concrete [19].
Considering the deficiencies of the existing models and the realistic demand, a new DCE model
was established in this paper based on the Hoek-Brown yield criterion [20] and the
Dilatant-Kinematic [21] equation in the plastic region in DCE. A numerical solution of cavity wall
pressure is obtained by solving the dynamic spherical cavity expansion resistance. Then, a resistance
model and a calculation formula of penetration depth are established. The reliability of the model is
verified by comparison with test data.

2. Basis equations and division of the response regions

2.1. Description of Hoek-Brown yield criterion and Dilatant-Kinematic equation

In traditional dynamic cavity expansion theory, the Mohr-Coulomb yield criterion and Tresca-limit
shear condition were combined to describe the shear behaviors of geological materials [9], as shown
in figure 1(a). However, It is not sufficient to capture the non-linear relationship between p and  of
material as observed in experimental [22]. To overcome this deficiency, the Hoek-Brown [20] model,
a hyperbolic yield criterion, was introduced in DCE. It was a statistic results based on a significant

2
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

quantity of laboratory triaxial test data for various rock types, as depicted in figure 1(b). The equations
selected to describe the yield criterion of the concrete/rock is

3
1   3  Y m s
Y (1)

where  1 and  3 are respectively the first and third principle stresses of concrete or rock materials;
Y is the compressive strength; m and s are material constants.
In [23], it has been derived theoretically from fracture mechanics that the constant m is not
simply an empirical constant but represents the brittleness of rock-like materials; s  1 for intact rock.
The relationship between Y / f t and m is

Y
 8.62  0.7m
ft
(2)

where ft is the uniaxial tensile strength of the rock material.

However, note that equation (2) can only be used for brittle rocks and common strength concretes,
as the statistic results did not include the experimental data of high strength concrete under high
hydrostatic pressure. Thus, for the high strength concrete, the value of m should be modified
according to the triaxial test results. In figure 2, the yield envelopes of Hoek-Brown model shows
great agreement with the test data [22], in which m are separately 2.5 and 5.5 for the 140 MPa and
48 MPa concrete. Then adding the test data from Li [24], a quadratic function relation between
concrete compressive strength and m is obtained by fitting, as shown in figure 3. It suggested that
the value of m for common strength concrete is in the range of 5-10, which is consistent with the
conclusions in [20]. While, for the high-strength concrete, the value of m dramatically decreases
with the increasing of the concrete compressive strength, this result will be used in section 4.3.

(a) Mohr-Coulomb yield criterion.

3
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

(b) Hoek-Brown yield criterion.

Figure. 1 Yield criterion for concrete/rock.

Figure 2. Experimental data for yield criterion. Figure 3. The relationship of Y and m for concrete.

In addition, as mentioned in section 1, geological materials will undergo dilatancy in compression

due to the effect of nucleation and growth of micro-cracks. Thus the volumetric compression and
expansion of geological materials in compressive tests should be reflected in constitution model to
make a precise resistance prediction. Based on a large number of explosion experiments in rocks, a
relationship between volumetric strain rate and shear strain rate of geological materials was proposed
[25],
v 2v 2  k  v v 
    
r r k  1  r r  (3)

where v is particle velocity, V  v r  2v r is volumetric strain rate,   v r  v r is

shear strain rate; k is dilatant coefficient.

4
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

It can be seen from equation (3) that k determines the compression or dilation of material volume:

if k  2 , the material is compressible; if k  2 , the material is incompressible; and if k  2 , the

material is dilated[26]. Moreover, equation (3) suggested that the effects of shear dilatancy and
compressibility are combined together, which is the typical property of geological materials.

2.2. Response regions

According to the given Hoek-Brown yield criterion and Dilatant-Kinematic equation, the material
outside the cavity is divided into two parts, i.e., an elastic part (elastic, cracked) and a comminuted

part (dilatant and compacted), as shown in figure4, where V , C3 , C 2 , C 1 , C d and t are the cavity
expansion velocity, compacted-dilatant boundary velocity, dilatant-cracked boundary velocity,
cracked-elastic boundary velocity and time.

Figure 4. Response regions.

Equations of mass conservation and momentum of material particles in Eulerian coordinate system
with spherical symmetry are as follows[13]

 v 2v     
      v  (4)
 r r   t r 

 r 2  r      v v 
     v  (5)
r r  t r 

where  is the material density, v is the particle velocity of the material, and r is the radical

coordinate;  r and   are respectively the radical stress and circumferential stress in material.
With cavity expansion caused by the radical pressure on the cavity surface, the stress wave will
transfer in each response region. Thus, the following Hugoniot jump conditions must be satisfied on
the interface of each response area[13],
   v  cn      v  cn  (6)

     v  v  cn        v  v  cn  (7)

5
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

where   , v and   are density, particle velocity and stress of the material before the wave;   , v

and   are density, particle velocity and stress of the material after the wave, and cn is boundary
velocity.
In section 3, the radial stress and particle velocity in different regions is solved with the similarity
transformation method and a fitting quadratic function relation of cavity expansion velocity and
expansion stress on the surface of the cavity is obtained.

3. Dynamic spherical cavity expansion analysis

3.1. elastic region

In this region, generalized Hooke’s rule is applied to describe the radial stress and circumferential
stress of the material in spherical coordinates,
E  u u
r    1    2 
1   1  2   r r
(8)

E  u u 
    
 
1   1  2   r r  (9)
where E and  are separately the Young’s model and Poisson’s ratio of the material, u is the
radical displacement.
The displacement of the material in elastic is obtained by using Eqs. (8, 9, 5), in which the
convective term is neglected as the particle velocity is extremely small.
 2u 2 u 2u 1  2u
  
r 2 r r r 2 cd 2 t 2
(10)

E 1   
cd 
1   1  2   0 (11)

By using 1  r / c1t , u1  u / c1t in equation (10) and similarity transformations[13], the equation
(10) is transformed into non-dimensionless forms,

d 2u1 2 du1 2u1 1 2 d 2u1

   1
d12 1 d1 1  2 d12 (12)

where   cd / c1 ,and the non-displacement in the elastic region is

 2  312
u1  A1  B , 1  1  
312 (13)

6
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

in which, the constants A and B are determined by using the boundary conditions at the inner
surface of the elastic region and the elastic–cracked interface.
u1 1   0
(14)

   1   f t
1
(15)

where f t is the tensile strength of the material

3
B A
2 (16)
 f  2 1   
A t 2 
  0 cd  2 1      2  3   
3 3
(17)

3.2. Cracked region

For geological materials, its tensile strength is lower than compressive strength, and materials are easy
to produce tensile cracks under compression. In this region, the circumferential stress of materials is
set as zero and the momentum conservation equation and radial stress are
 r 2 r  2u
   0 2
r r t (18)
u
 r  E
r (19)
Substituting equation (18) into equation (19) and using the transformation in equation (20). Thus
r u c c
2  , u2  ,  2,   1
c2 t c2 t cp c2
(20)

d 2 u 2 2 du 2 2
2 2 d u2
    2
d 2 2  2 d 2 d 2 2 (21)

cp  E / 0
where . The solution of equation (20) is

1   22 2
u2  C  D , 1  2  
2 (22)

in which, the parameters of C and D are evaluated from the continuity conditions of displacement
at the elastic-cracked interface and the yield criterion at the cracked-dilatant interface.
u2 2    u1 1 1
(23)
du 2
 r  2 1   E Y
d 2 (24)

7
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

Y
D
E   2  1
(25)

 A 3 2  1   2 2
C     A  1   D
 2  3  
(26)
With equation (19, 22, 25, 26) and Hook’s law, the radical stress and volumetric strain in the
cracked region at the elastic-cracked interface are derived as
 1 
2  2 
  ED  2   2 
   (27)

 1 
2  2 
  D  2   2  1 2 
  (28)
Similarly, the particle velocity, radical stress, volumetric strain and density of the material in the
cracked region at the cracked-dilatant interface are
du 2
v2 2 1  c22  c2 u2  c2 C  2D 
d 2 (29)

2  Y
2 1
(30)

Y 1  2 
2  2 1

E (31)

0
2 
1  2 2 1
(32)

With increasing cavity expansion velocity, the interface velocity of C 2 will catch up the interface

velocity of C1 , which means the cracked region disappeared [27]. In this case, the elastic region is
directly adjacent to the dilatant region, and the boundary conditions in elastic region at the
elastic-dilatant interface is equation (31). Thus the constants A and B in equation (13) becomes
3
B A
2 (33)
 Y   1   
A   2 
 0 cd  1      (1  2 )  3
3
(34)

3.3. dilatant region

In this region the material satisfies both the Hoek-Brown yield criterion and Dilatant-Kinematic

8
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

v  c2t , t   v3
equation. Appling the boundary condition and integrating equation (3), the particle
velocity can be expressed as
k1
c t
v  v3  2  (35)
 r 

where v3 is the particle velocity in the dilatant region at the cracked–dilatant interface, k1 is the
dilatant coefficient in dilatant region.
With equation (35), the mass conservation equation (1) becomes

1    
  2  k1  v3  c2t 
k1
k1 1
   v  (36)
r  t r 

Using the transformations in equation (37) and an ordinary differential equation form equation (36)
is obtained.
r r c c 
3  or ,   2 or 2 , S3  r (37)
Vt c3t V c3 Y

1
 2  k1   v3   k 1

c3 d
 d3  (38)
k1  2 v 
3  3   k1  3
c3
Equation (38) integrates to
M1
 N  N1 1 k1 
   c   k111  1 (39)
 3 

1
M1   2  k1   v3   k1  (40)
c3

v3 k1
N1   (41)
c3

To describe the compaction or dilation state of the material, the density of the concrete material in

comminuted region must be known. Combining the boundary condition  (3   )  3 ,  c is derived
from equation (39).
3 3   
c  M1
(42)
 N1  N1 1 k1 
   k1 1  1
 
With equation (1, 35), the conservation equation (5) becomes

9
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

4 r mY   mY   4Y 2 s  mY
2
 r  c k1 t k1 1 v 
     v3 k1 2 k1  v  (43)
r r  r r 

Introducing equation (37) to equation (43), an ordinary differential equation is obtained,

M1 S s
 k1
2 k1  m 4 3 1 4  m
dS3  c  N1  N1 1 k1      m m
   1 v3 k1   c3  v32 k1 2 k 1   (44)
d3 Y  3k1 1    
  3 
1
3 3

The boundary conditions at the cracked-dilatant interface are as follows when the cracked region
exists:
 2  v2  2 1
 c2 c
v3 3   2 (45)
3

3  3 
  2 2 1   2 v1  2 1 v
2  2 1  
 c2  3 v3 3  v3 3   c2  (46)
The boundary conditions at the elastic-dilatant interface are as follows when the cracked region
vanishes:
 2  v1  1  c2 
v3 3   1
 c2
3
(47)

 
 3 3  1 11  2 v1 11 v1 11  c2  3  v3 3 v3 3  c2   (48)

Here, according to the division the response regions, two different boundary conditions should be

considered for the inner surface 

3  1
of the dilatant region. If there is no compacted-region, the
cavity expansion velocity is equal to the particle velocity as in equation (45) or (46). If the
compacted-region exists, the particle velocity and material density in dilatant region at the
dilatant-compacted interface are deduced from equation (35) and (39).

v3 3  1  v3 3    k1 (49)

M1
3 3  1  c 1  N1  N1 1 k1 
(50)

3.4. Compacted region

In this region, material is also described by both the Hoek-Brown yield criterion and
Dilatant-Kinematic equation. Different with the dilatant region, the compacted coefficient k 2 is
applied in the Dilatant-Kinematic equation for the compacted region. Following dimensionless
transformations are introduced,
r  c
4  , S4  r ,   3
Vt Y V (51)
Similar to the derivation process of the dilatant region, the dimensionless stress differential

10
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

equation of compact region is

M2 S s
 k2
2 k2  m 4 4 1 4  m
dS4  c  N2  N 2 1 k 2      m m
   1 v4 k2   V  v4 2 k2 2 k 1  
d4 Y  4 k2 1   
 4  4
2
  4
(52)
1
M 2  (2  k 2 )  v4   k2 
V (53)
v4
N2    k2
V (54)
  4  4   
Combining the boundary condition and the Hugoniot jump conditions, The
boundary conditions at the dilatant-compacted interface are
3 3 1  v3 3 1  c3 
v4 4    c3
4 (55)

4  4 
  
  3 3 1  3 3 1 v3 3 1 v3 3 1  c3   4 4  v4 4  v4 4   c3  (56)

The ordinary differential equations, equation (44) and (52) are solved by using the Runge-Kutta
method to get the radical stress of the material in the dilatant and compacted regions. The boundary

velocities V and C3 are solved by iteration method until the given boundary conditions are

satisfied.
Based on the analysis in above, the relationships between the interface velocities and the expansion
velocity of the cavity for the concrete with a uniaxial compressive strength of 48MPa are shown in
figure5. It can be seen that the compacted region appears when v is above of 350 m/s. The response
region is compacted-dilatant-cracked-elastic for the material at this point. With increasing V, the
cracked region diminishes and is eliminated eventually, and the response region is
compacted-dilatant-elastic. figure 6 presents a dimensionless quadratic curve of the normal stress on
the cavity surface versus the cavity expansion velocity.
2
r  V   V 
 a1    a2    a3
Y  Y    Y  
 0   0  (57)

a1 a2 a3
where, , , and are inertial, viscosity and strength coefficients[2] with corresponding values

1.23, 3.06 and 9.32, separately. Dilatant coefficients k1  1.8 and k2  2.1 are adopted in this
model[26].

11
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

Figure 5. The propagation of region interfaces while the cavity expansion velocity increased.

Figure 6. Dimensionless radial stress at the cavity surface vs. cavity expansion velocity.

4. Results and validation

4.1. The influences of material parameters

The influences of different material parameters on the relationships of the radical stress  r and the

cavity expansion velocity V are discussed. In figure 7, it suggested that  r decreases with increasing

k1 and k 2 , which means that dilatancy of the material in comminuted region will increase the cavity

expansion resistance, and the compressibility of the material will inversely decrease the cavity

expansion resistance. Under the same cavity expansion velocity, radical stress  r increases as
brittleness coefficient m increases, as shown in figure 8 (a). It can be seen that m is a sensitive

parameter forth radical stress  r , and the conclusions in section 2 will be used to determine the value

of m for each geological material. figure 8 (b) shows that the radical stress  r increases with
increasing intact coefficient s . When the dimensionless cavity expansion velocity on the cavity

12
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

surface is lower than 2.5, the distinction of the radical stresses under different intact coefficients are
obvious; While, the influence of intact coefficient on radical stress can be neglected when the
dimensionless expansion velocity is larger than 2.5. The existence of initial flaws reduces the strength
of materials: With low cavity expansion velocity, the radial stress is mainly determined by the static
term in equation (57); As cavity expansion velocity increases, the percentage of resistance induced by
material inertial increases, thus in this condition the influence of the intact coefficient on radical stress
is not obvious.

Figure 7. The influence of dilatant coefficient.

(a) The influence of m (b) The influence of s

Figure 8. Dimensionless radial stress vs. cavity expansion velocity for different coefficients m and s.

4.2. Rigid projectile penetration model

The projectile penetration process in geological target is described by the two-stage model[27]: cavity
stage and tunnel stage. The penetration resistance at the cavity stage is

Fz  cz , 0  z  4R , (58)

where c is the parameter in [27], z is the penetration depth, R is the radius of the projectile. At
the tunnel stage the resistance force is determined by assuming that the normal resistance on the
surface of the projectile is equal to the radical stress in cavity expansion theory.

13
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

Fz    (vn )cos d 4R  z ,


vn  v cos  , (59)
  arccos(e z , n),

ez
where is the unit vector along the z-axis, n is the unit vector of interior normal on the projectile

(e z , n)  is the angle between ez and n , v is the

surface, is the inner product of vectors ,
projectile velocity along the z-axis, and  is the lateral surface of the projectile. Here, the friction on
projectile surface is neglected.
Combining the Newton’s law ,the kinematic equation of the projectile with a mass m is

d 2z cz , 0  z  4R
m 2
  Fz    2 . (60)
dt  A1v  A2 v  A3 , z  4R

in which, A1 , A2 , A3 are the integral coefficients from equation(59). Then, the expression for

penetration depth P is
m  Q1 2Q  Q v  Q2  
P ln  2  arctan 2  arctan 1    4 M0
Q3  Q3  
2
2 A1  (v1  Q1 )  Q3 Q3

m mv 2
 A2  A22  4( A1  )( A3  s )
4M 0 4M 0
v1 
m
2( A1  )
4M 0
A3 A
Q1  , Q2  2 , Q3  Q1  Q22
A1 2 A1 (61)

where vs and v1 are the initial velocity at the first stage and the initial velocity at the second stage

separately, V is the volume of projectile, and m is the projectile mass.

4.3. Model validation

In this section, based on the tests of rigid projectiles penetration into concrete and rock targets, the
penetration depths predicted by equation (61) as well as the classical model[28] suggested by
Forrestal are compared, respectively.
The penetration depths of rigid projectile penetrating into 36 MPa、51 MPa and 62.5 MPa common
strength concrete [28, 29]are depicted in figure 9 (a). Predictions from the proposed model and the
classical model show good agreement with the test data for those three kinds of concrete. When the
initial impact velocity is lower than 800 m/s, the proposed model is more accurate than the classical
model. When the initial impact velocity is higher than 800 m/s, the predictions form both models are
distinct: the prediction results from the classical model are more accurate for the 36 MPa concrete,

14
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

while the proposed model gives a better prediction for 51 MPa and 61 MPa concretes. As for the
high-strength compressive concrete (87 MPa and 112.5 MPa) [30], the predictions from both model
integrating the experiments data are shown in figure 9(b). It can be observed that the present model
based on the Hoek-Brown yield criterion and Dilatant-Kinematic equation gives better predictions
than the classical model
The penetration data of limestone with compressive strength of 60 MPa and granite with
compressive strength of 154 MPa are selected as typical data of rigid projectiles penetration into low
strength rock targets and high strength rock targets respectively[31, 32] , as shown in figure 10 (a and
b). It can be seen that the prediction results of the proposed model are consistent with the
experimental data than that of Forrestal model, especially for the granite targets. It should be point

that as the rocks tensile strength and Poisson’s ratio are unavailable in primary literature, Y  12 f t

and   0.26 are adopted in our model.

（a） （b）
Figure 9. Depth of penetration vs. initial impact velocity for concrete target.

（a） （b）
Figure 10. Depth of penetration vs. initial impact velocity for rock target.

15
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

5. Conclusion
In this paper, the dynamic cavity expansion theory is applied for the depth prediction of rigid
projectiles penetrating into concrete and rock targets, the main works and conclusion are as follows:
(1) The Hoek-Brown yield criterion and Dilatant-Kinematic equation are introduced to describe the
plastic behavior of concrete and rock materials, and a precise dynamic cavity expansion model is
proposed.
(2) A depth of penetration equation for rigid projectile is proposed combing the Newton’s law, the
present model with higher accuracy has a well-being prediction of the penetration depth of
concrete/rock materials with different strength.
(3) The dilatancy of the material in comminuted region will increase the cavity expansion
resistance, and compressibility of the material will inversely decrease the cavity expansion resistance.

(4) The radical stress  r increases as brittleness coefficient m increases, and the value of m
decreases with increasing uniaxial compressive strength for concrete material.

(5) When cavity expansion velocity is low, the radical stress  r increases with increasing intact
coefficient s . However, when the dimensionless cavity expansion velocity on the cavity surface is
larger than 2.5, the influence of intact coefficient on radical stress can be neglected.

Acknowledgment
This work was supported by the National Natural Science Foundation of China grant No. 11390362,
11521062, and the Advanced Research in Equipment Foundation of China grant No. 61406570101.

References:
[1] Ben-Dor, G A Dubinsky and T Elperin 2003 Computers & Structures 81 9-14.
[2] Ben-Dor, G A Dubinsky and T Elperin 2012 Central European Journal of Engineering 2
473-82
[3] Tate A 1967 Journal of the Mechanics and Physics of Solids 15 387-99
[4] Tate A 1986 International Journal of Mechanical Sciences 28 535-48
[5] Ben-Dor, G A Dubinsky and T Elperin 2013 High-Speed Penetration Dynamics 696
[6] Walter T A 1984 Journal of Structural Engineering 110 39-55
[7] Yankelevsky D Z 2017 International Journal of Impact Engineering 106 30-43
[8] Bishop RF, R. Hill and N F Mott 2002 Proceedings of the Physical Society 57 147-59
[9] Forrestal M J and V K Luk 1992 International Journal of Impact Engineering 12 427-44
[10] Yinan W 2009 The Mechanism of High-speed Kinetic Energy Projectile Penetration into
Concrete (Beijing Institute of Technology)
[11] Peng Y 2018 Acta Mechanica Sinica 1-17
[12] Feng J 2015 International Journal of Impact Engineering 84 24-37
[13] Wang J 2018 International Journal of Impact Engineering 120 110-7
[14] Kong X Z 2017 International Journal of Impact Engineering 100 13-22.
[15] Luk V K and M J Forrestal 1987 International Journal of Impact Engineering 6 291-301

16
The 2020 Spring International Conference on Defence Technology IOP Publishing
Journal of Physics: Conference Series 1507 (2020) 032012 doi:10.1088/1742-6596/1507/3/032012

[16] Forrestal M J 1986 Penetration into dry porous rock. International Journal of Solids &
Structures 22 1485-500
[17] Rosenberg Z 1993 Journal of applied physics 74 752-3
[18] Bavdekar S and G Subhash 2018 International Journal of Impact Engineering 118 60-66
[19] Shao R 2019 Composite Structures 220 861-74
[20] Hoek E and C Martin 2014 Journal of Rock Mechanics and Geotechnical Engineering 6
287-300
[21] Guo X , T He and H. Wen 2012 Journal of Engineering Mechanics 139 1260-7
[22] Hanchak S 1992 International Journal of Impact Engineering 12 1-7.
[23] Zuo J P 2008 International Journal of Rock Mechanics & Mining Sciences 45 594-9
[24] Li Q and F Ansari 1999 Journal of engineering mechanics 125 1-10
[25] Nikolaevskij V N 1990 Mechanics of Porous and Fractured Media
[26] He T, H M Wen and X J Guo 2011 Acta Mechanica Sinica 27 1001-12
[27] Forrestal M J and D Y Tzou 1997 International Journal of Solids & Structures 34 4127–4146
[28] Forrestal M J 1994 International Journal of Impact Engineering 15 395-405
[29] Forrestal M J 1996 International Journal of Impact Engineering 18 465-76
[30] Zhang M 2005 International Journal of Impact Engineering 31 825-841.
[31] Forrestal M J, D J Frew and S J Hanchak 1999 Journal of Applied Mechanics 841
[32] Zhang 2005 Journal of rock mechanics and engineering 24 1612-8

17