You are on page 1of 8

Defence Technology 16 (2020) 609e616

Contents lists available at ScienceDirect

Defence Technology
journal homepage: www.keaipublishing.com/en/journals/defence-technology

Growth model of cavity generated by the projectile impacting liquid-


filled tank
Bei-lei Zhao a, Ji-guang Zhao b, Cun-yan Cui a, *, Yong-sheng Duan a, Yan Wang a
a
Department of Aerospace Science and Technology, Space Engineering University, Beijing 101416, China
b
Department of Electronic and Optical Engineering, Space Engineering University, Beijing 101416, China

a r t i c l e i n f o a b s t r a c t

Article history: The high-speed impact of a projectile on a liquid-filled tank causes the hydraulic ram, in which a cavity is
Received 21 June 2019 formed. To study the growth characteristics of the cavity, the formation mechanism of the cavity is
Received in revised form analyzed. The effect of Reynolds number and Mach number on drag coefficient is considered, the axial
11 September 2019
and radial growth models of the cavity are established respectively. The relative errors between the
Accepted 29 September 2019
Available online 3 October 2019
cavity length calculated by the axial growth model, the cavity diameter calculated by the radial growth
model and Ma L. Y. test results are less than 20%, which verifies the effectiveness of the axial and radial
growth models. Finally, numerical simulation is carried out to study the growth characteristics of the
Keywords:
Projectile
cavity caused by the projectile impacting the satellite tank at the velocity of 4000 m/s. The cavity length
Impact and diameter calculated by the axial and radial growth models agree well with those obtained by
Liquid-filled tank simulation results, indicating that the cavity length and diameter in satellite tank can be accurately
Cavity calculated by the axial and radial growth models.
Axial growth model Copyright © 2020 China Ordnance Society. Publishing Services by Elsevier B.V. on behalf of KeAi
Radial growth model Communications Co. Ltd. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).

1. Introduction was considered as the first person to study water entry, recorded
the phenomenon of spherical projectiles impacting different liquids
The impact of a projectile on the liquid-filled tank results in a by means of instantaneous photography [2]. Lundstrom E. A. and
hydrodynamic ram, which includes the shock stage, drag stage, Fung W. K. (1976) performed a series of tests where the armour
cavity stage and exit stage [1]. During the shock stage, the hemi- piercing projectiles with the diameter of 12.7 mm and 14.5 mm
spherical shock wave is generated and the pressure loading is entered the water at the velocity of 800 m/s~1070 m/s [3]. They
exerted on the walls. As the projectile propagates forward, its ve- analyzed the formation, closure and collapse of the cavity, and
locity attenuates due to the viscous drag. Part of projectile kinetic proposed an empirical formula for predicting the cavity radius. Lee
energy is converted into liquid kinetic energy, a cavity forms behind M. et al. (1997) deduced formulations of cavity radius and cavity
the projectile. The exit stage occurs when the projectile penetrates wall velocity for the spherical projectile entering water vertically
the back wall of the tank. Among the four stages, the cavity stage [4]. Ma Q.P. (2014) carried out numerical simulation in which cyl-
plays a key role. On the one hand, the growth process of the cavity inders with different head cones entered the water at a velocity of
is affected by the shock wave. On the other hand, the cavity 500 m/s [5]. He concluded that the head cone angle and drag co-
expansion squeezes the liquid and imposes the pressure loading on efficient were the main factors affecting the cavity radius. Unlike
the walls. Therefore, it is necessary to further study the growth the cavity induced by water entry, the cavity in hydrodynamic ram
characteristics of the cavity. is formed in a closed container, its growth is constrained by the
Studies on cavity are originated from the naval weapons, which container walls. Therefore, the conclusions in water entry cannot
aim at analyzing the cavity caused by the water entry of high speed accurately describe the cavity characteristics in hydrodynamic ram.
torpedoes, missiles and projectiles. Worthington A. M.(1900), who Crater caused by the penetration of shaped charge jet has also
become the focus of many studies. Szendrei T. (1983) first proposed
a radial growth equation for crater [6], and Held M. (1995) made a
* Corresponding author. slight modification [7], which constituted the classical Szendrei-
E-mail address: ccy6655@126.com (C.-y. Cui). Held equation. Held M. (1996) also conducted a test in which the
Peer review under responsibility of China Ordnance Society

https://doi.org/10.1016/j.dt.2019.09.013
2214-9147/Copyright © 2020 China Ordnance Society. Publishing Services by Elsevier B.V. on behalf of KeAi Communications Co. Ltd. This is an open access article under the
CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).
610 B.-l. Zhao et al. / Defence Technology 16 (2020) 609e616

shaped charge jet with a diameter of 42 mm was fired at the ve-


locity of 7.9 km/s against a plexiglass container filled with water
[8]. The radial expansion of the crater was recorded as a function of
time, and the experimental results agreed well with the calculation
results of Szendrei-Held equation. Unlike the projectile with initial
kinetic energy, the shaped charge jet has sustained energy input
and stronger penetration capability. Thus, there are some differ-
ences between the cavity characteristics in hydrodynamic ram and
crater characteristics in jet penetration.
Since the 1970s, a large number of tests and numerical simu-
lations on hydrodynamic ram have been carried out. Nishida M.
et al. (2006) conducted a test in which the spherical projectile was
accelerated to 40~200 m/s to impact a water-filled vessel [9].
Disimile P. J. et al. (2008) performed tests in which the 12.7-mm-
diameter projectiles made of tungsten, steel and aluminum were
Fig. 1. Schematic diagram of a cavity caused by a spherical projectile impacting liquid-
shot into a water-filled tank at the velocity of 341 m/s, 389 m/s and filled tank.
455 m/s [10]. Varas D. et al. (2012) used the ALE simulation
approach to reproduce a projectile impacting the 60%, 75%, 100%
water-filled tank at the velocity of 600 m/s and 900 m/s [11,12]. moment when the projectile begins to impact the liquid is t ¼ 0.
These study mainly focus on the deformation of the tank wall, the The distance from the impact point to the cavity head is the cavity
peak pressure of shock wave and the attenuation of projectile ve- length in the axial direction, which is expressed as l. The maximum
locity. Lecysyn N. et al. (2010) analyzed the cavity growth theo- radius at the cavity bottom is the cavity radius in the radial direc-
retically and deduced the formula of cavity radius as a function of tion, which is expressed as r.
time [13]. However, the effect of the shock wave on the cavity
growth is neglected, and the axial growth law of cavity has not been
explored. 3. Cavity growth model
In this paper, the formation mechanism of the cavity after a
projectile impacting the liquid-filled tank is analyzed. The effect of 3.1. Axial growth model of cavity
the shock wave on the axial growth of cavity and liquid motion on
the radial growth of the cavity are considered, and the axial and In the studies of hydrodynamic ram [15], the length of cavity is
radial growth models of the cavity are established respectively. The assumed to be equal to the projectile axial displacement, which
validity of the axial and radial growth models were verified based means the axial growth velocity of cavity ua is the same as the
on Ma L. Y. test results. Finally, with the method of numerical projectile velocity up.
simulation and cavity growth models, the growth characteristics of
the cavity caused by the projectile impacting the satellite tank at ua ¼ up (1)
the velocity of 3000 m/s are analyzed.
When the projectile moves in the liquid hydrazine, it is only
subjected to drag force. From Newton’s second law, the following
2. The formation mechanism of cavity
can be deduced [16]:

When a high-speed projectile enters a liquid-filled tank, a dup 1


boundary layer forms around the projectile. As the fluid flows back rp Vp ¼  CD r1 Ap u2p (2)
dt 2
along the boundary layer, the kinetic energy in the boundary layer
is dissipated by the liquid viscosity, and the velocity of the fluid where, rl is the density of the liquid; rp is the density of the pro-
decreases. Due to the continuity of the flow, the boundary layer jectile; Vp is the projectile volume; Ap is the cross section area of the
becomes thicker, and more low-speed fluids flow through at the projectile; CD is the drag coefficient. For spherical projectile, Vp ¼ 4/
same time, which results in the adverse pressure gradient. Under 3pr3,
p Ap ¼ prp.
2

the action of the adverse pressure gradient, the flow slows down Two parameters on which the drag coefficient can be very
further. Finally the kinetic energy of the fluid in the boundary layer
is insufficient to maintain the downstream flow. Thus, boundary
layer separation occurs.
Then, double-line vortices with opposite rotation direction and
regular arrangement appear on both sides of the projectile tail.
These vortices, called Karman Vortex Street, fall off periodically.
Part of the projectile kinetic energy is transformed into the liquid
kinetic energy, the liquid is accelerated and the Reynolds number
enlarges. If the Reynolds number is large enough, the randomness
of vortices shedding rises, finally Karman vortex street turns into a
turbulent wake [14]. Meanwhile, the external air enters through the
perforation at the impact point, the turbulent wake and liquid
behind the projectile are squeezed. The liquid flows perpendicular
to the impact direction and the cavity is formed.
Fig. 1 presents the schematic diagram of a cavity caused by a
spherical projectile impacting liquid-filled tank. The projectile is
shot vertically into the tank along the x-axis, the length and height
of the tank are d and 2h, respectively. It is assumed that the Fig. 2. The curve of drag coefficient CD changing with Reynolds number Re [14].
B.-l. Zhao et al. / Defence Technology 16 (2020) 609e616 611

dependent are the Reynolds number (Re) and the Mach number 8
(Ma). The accurate dependence of the drag coefficient on Re and Ma > u
>
>  p0   ð0 < Ma < 0:5Þ
is generally quite complex. Reasonable simplifications are made in < 1 þ 0:15 r1 up0 rp rp t
Ref. [14]. For Ma<0.5, the drag coefficient is essentially independent up ¼ (5)
>
> u
of Ma, it is strongly dependent on Re, as is indicated in Fig. 2 [14]. >
:  p0   ð3 < Ma  4Þ
1 þ 0:351 r1 up0 rp rp t
When the projectile impacts the liquid-filled tank, the turbulent
wake is generated at the tail of the projectile, and the Reynolds
where, up0 is the initial velocity of the projectile.
number Re > 103. Within this range, the drag coefficient basically
Eq. (3) can also be expressed as:
remains unchanged. For spherical projectile, CD ¼ 0.4.
For Ma  0.5, the drag coefficient is a strong function of Ma, and  2
the influence of Re is secondary. The relationship between drag up up
CD ¼  0:23122 þ 1:55868  0:64417
coefficient and Mach number is shown in Fig. 2 [14]. For spherical C1 C1
projectile, with the increase of Mach number, the drag coefficient
 3
up
first increases rapidly and then decreases slowly. For 0.5  Ma3, þ 0:08478 ð0:5  Ma  3Þ (6)
C1
the function between drag coefficient and Mach number is fitted, as
shown in Fig. 3(a). Substituting Eq. (6) into Eq. (2), the following can be gained:

dup 3r1 
CD ¼  0:23122 þ 1:55868Ma  0:64417Ma2 ¼   0:23122C 31 þ 1:55868C 21 up  0:64417C1 u2p
dt 8rp rp C 31

þ 0:08474Ma3 ð0:5  Ma  3Þ (3)
þ 0:0847u3p ð0:5  Ma  3Þ
For spherical projectile, the drag coefficient remains approxi- (7)
mately constant when 3 < Ma4. With the range of 0 < Ma4, the
drag coefficient can be expressed as: If the length of the cavity is l, the projectile velocity up is defined
as:

8
< 0:4 ð0 < Ma < 0:5Þ
CD ¼ 0:23122 þ 1:55868Ma  0:64417Ma2 þ 0:08474Ma3 ð0:5  Ma  3Þ (4)
:
0:936 ð3 < Ma  4Þ

up
Mach number is defined as Ma ¼ C1 , Cl is the sound velocity of dl
the liquid. For 0 < Ma<0.5 and 3 < Ma4, the drag coefficient is a up ¼ ua ¼ (8)
dt
constant, the projectile velocity up can be solved by an integral of
Eq. (2): Then, the length of the cavity can be solved as follows:

Fig. 3. (a) Fitting curve of drag coefficient for 0.5  Ma3; (b) Curves of drag coefficient varying with Mach number for different shapes.
612 B.-l. Zhao et al. / Defence Technology 16 (2020) 609e616

8 u
>
>  p0   ð0 < Ma < 0:5Þ 1 r l up 2
>
< 1 þ 0:15 r1 up0 rp rp t Pr ¼ r1 u2r ¼  qffiffiffiffiffiffiffiffiffiffiffi
 ffi 2 (17)
dl 2 2 1þ r r
¼ (9) l p
dt > > u
>
:  p0   ð3 < Ma  4Þ
1 þ 0:351 r1 up0 rp rp t The cavity radius at the bottom is r. Substituting Eqs. (15) and
(17) into Eq. (12), the following can be obtained for spherical
"  2 projectile:
d2 l 3r1 3 2 dl dl
¼  0:23122C 1 þ1:55868C 1 0:64417C 1 2 3
dt 2 8rp rp C 31 dt dt u2p0 u2p
 3 # 4u2
p0 þ qffiffiffiffiffiffiffiffiffiffiffi 5 2
 ffi 2 r p ¼  qffiffiffiffiffiffiffiffiffiffiffi
 ffi 2 r
2
(18)
dl 1 þ rl rp 1 þ rl rp
þ0:0847 ð0:5Ma3Þ
dt
As the cavity radius is related to the velocity of the projectile at a
(10)
certain time, equation (5), equation (7) and equation (18) consti-
The differential Eq. (9) and Eq. (10) constitute the axial growth tutes the radial growth model of the cavity, in which the initial
model of the cavity, in which the initial condition is t ¼ 0, l ¼ 0, and condition is t ¼ 0, r ¼ rp, and the boundary condition is 0  r  h.
the boundary condition is 0  l  d.
4. Verification of cavity growth model
3.2. Radial growth model of cavity
Ma L. Y. et al. (2010) performed a test where a steel projectile
In the analysis of the radial growth of the cavity, the static with the diameter of 9.5 mm was shot into an aluminum vessel
pressure is neglected and the dynamic pressure produced by the filled with water at the velocity of 686 m/s [15]. The length and
liquid motion is considered. The pressure P at the cavity bottom can diameter of the cavity at 441 ms, 774 ms and 1107 ms was recorded by
be expressed as: a high-speed camera, as illustrated in Fig. 4. The vessel is a cuboid
with 50, 40, 40 and 0.4 cm in length, width, height and wall
1
P ¼ r1 u21 (11) thickness, correspondingly.
2 Based on Ma L. Y. test conditions, the axial and radial growth
models are used to calculate the length and diameter of the cavity
where, ul is the liquid velocity at the cavity bottom.
at 441 ms, 774 ms and 1107 ms, 1440 ms, 1773 ms and 2106 ms. The
As the pressure changes with the increase of the cavity radius,
related parameters of water [17], vessel and projectile are shown in
Szendrei T. [6] supposed that the product of cavity pressure and
Table 1.
cross section area at the cavity bottom always remain constant.
The impact velocity of projectile is 686 m/s and the Mach
P0 , Ap ¼ Pr ,Ar (12) number in water is less than 0.5, so Eq. (9) is used to calculate the
cavity length. The cavity length lt in the test, the cavity length lc
where, P0 is the initial stagnation pressure; Ap is the area of pro- calculated by axial growth model, the cavity diameter Dt in test, the
jectile at the cavity bottom; Pr is the dynamic pressure at the cavity cavity diameter Dc (Dc ¼ 2r) calculated by radial growth model and
bottom; Ar is the cross section area at the cavity bottom. the relative errors ε are shown in Table 2. It is concluded that the
In the process of the projectile getting into the liquid, The liquid calculated cavity length is larger than the cavity length in the test.
in contact with the projectile produces both axial and radial mo- With the increase of time, the calculated cavity length is close to the
tion. The axial velocity of liquid ula is equal to the initial velocity of test results, and the relative errors between them decreases. When
projectile up0: t < 1440 ms, the calculated cavity diameter is smaller than the test
results. When t > 1773 ms, the calculated cavity diameter is larger
u1a ¼ up0 (13) than the test results. The reason for the error is that a series of
simplifications have been made in the derivation of the axial and
The radial velocity of liquid ulr is proportional to the initial ve- radial growth models. The influence of shock wave are neglected.
locity of projectile up0 [13]: Overall, the relative errors between test results and calculation
up0 results are less than 20%, which verifies the effectiveness of the
ulr ¼ qffiffiffiffiffiffiffiffiffiffiffi
 ffi (14) axial and radial growth models.
1þ rl rp
5. Cavity growth characteristics in the satellite tank
The initial stagnation pressure P0 is as follows:
2 3 5.1. Simulation model of the projectile impacting satellite tank
1 2 1 2 1 4 2 u2p0
P0 ¼ rl u1a þ rl u1r ¼ rl up0 þ  qffiffiffiffiffiffiffiffiffiffiffi 5
2 2 2  ffi 2 (15)
1þ r r l p
The satellite may be impacted by a high speed projectile in the
launch process. The satellite tank filled with liquid hydrazine,
The projectile continues to move forward, generating a cavity which accounts for a large part of the satellite mass, is one of the
behind it. The liquid at the bottom of the cavity moves radially. The most vulnerable components [18]. The hypervelocity impact of the
radial velocity at the bottom ur is proportional to the velocity of projectile on a satellite tank generates a cavity in liquid hydrazine.
projectile up at a certain time. Compared with the impact test, numerical simulation has the ad-
vantages of low cost and good repeatability to study the cavity
up characteristics. Numerical simulation is carried out to study the
ur ¼ qffiffiffiffiffiffiffiffiffiffiffi
 ffi (16)
1þ rl rp axial and radial growth characteristics of the cavity in the satellite
tank.
The dynamic pressure Pr at the cavity bottom can be expressed The satellite tank consists of a straight cylindrical section, a
as: upper hemisphere and a lower hemisphere. It is 50 cm in diameter,
B.-l. Zhao et al. / Defence Technology 16 (2020) 609e616 613

Fig. 4. The length and diameter of cavity recorded by a high-speed camera at t ¼ 441 ms, 774 ms, 1107 ms, 1440 ms, 1773 ms and 2106 ms.

Table 1 120 cm in height and 2 mm in wall thickness. The tank is made of


Parameters of water, vessel and projectile. titanium and is filled with liquid hydrazine. ANSYS/LS-DYNA soft-
Material r/(kg$m3) Cl/(m$s1) up0/(m$s1) rp/mm h/cm ware is used to simulate the situation in which a spherical steel
Water 1000 1483 e e e
projectile with a diameter of 4 cm impacts the satellite tank at the
Vessel 2700 e e e 40 velocity of 4000 m/s. The finite element model of a projectile
Projectile 7800 e 686 4.75 e impacting the satellite tank is displayed in Fig. 5.
The simulation model consists of the projectile, satellite tank,
liquid hydrazine and air. The interaction between the projectile and
Table 2 satellite tank is achieved through the eroding surface-to-surface
Test and calculation results of axial and radial growth models. contact. Lagrange grids are constructed for the projectile and
t/ms lt/cm lc/cm ε1/% Dt/cm Dc/cm ε2/% tank, yet Euler grids are constructed for liquid hydrazine and air.
These two grids are coupled through Arbitrary
441 16.43 19.75 16.81 4.29 3.56 17.02
774 24.50 28.34 15.67 5.71 5.05 11.56
LagrangianeEulerian (ALE) algorithm [19]. The air domain is a
1107 31.43 34.70 10.40 7.00 6.53 6.71 cylinder with the diameter of 80 cm and the height of 150 cm. The
1440 38.09 39.75 4.36 8.09 8.01 0.99 projectile and satellite tank are located in the air domain. To ensure
1773 42.87 43.94 2.50 8.97 9.49 5.80 the air flows into satellite tank and forms a cavity, air and liquid
2106 47.21 47.53 0.68 9.7 10.97 13.09
hydrazine meshes share the same nodes at their interfaces. The
mesh sizes of the projectile, satellite tank, liquid hydrazine and air
are 0.4 cm, 1 cm, 1 cm and 2 cm, respectively. The computing time
step is 0.6 ms. The steel projectile is modelled using the constitutive
equation MAT_JOHNSON_COOK and the state equation EOS_-
GRUNEISEN. The titanium tank is modelled using the material
model MAT_PLASTIC_KINEMATIC. Since the density, boiling point
and critical temperature of liquid hydrazine are close to the values
of water (within 2%) [20], the constitutive equation MAT_NULL and
the state equation EOS_GRUNEISEN of water are used to describe
liquid hydrazine. Air is modelled using the constitutive equation
MAT_NULL and the state equation EOS_LINEAR_POLYNOMIAL. The

Table 3
Material parameters and state equation parameters.

Material r/(kg$m3) C4 C5 A/MPa B/MPa E/GPa ss/MPa sb/MPa


Steel 7800 e e 496 434 206 e e
Titanium 4500 e e e e 118 820 902
Fig. 5. Finite element model of the projectile impacting satellite tank; the right side is Hydrazine 1008 e e e e e e e
the model cut in half. Air 1.293 0.4 0.4 e e e e e
614 B.-l. Zhao et al. / Defence Technology 16 (2020) 609e616

t ¼ 0 indicates the moment when the projectile begins to impact


liquid hydrazine. When t < 100 ms, the projectile velocities obtained
by simulation are larger than the calculated velocities. t > 125 ms,
the projectile velocities obtained by simulation are smaller than the
calculated velocities. The velocity of projectile decreases rapidly
due to the viscous resistance. In numerical simulation, the kinetic
energy lost by the projectile is converted into the kinetic energy of
liquid hydrazine and the deformation energy of the tank walls. The
attenuation trend of projectile velocities obtained by simulation is
consistent with the theoretical curve. The solution to equation (10)
is t ¼ 234 ms, l ¼ 0.5 m, which demonstrates that the projectile
passes through the satellite tank at t ¼ 234 ms. The residual velocity
of projectile is 1345.6 m/s. The simulation results indicate that the
residual velocity of projectile is 1284.3 m/s at t ¼ 234 ms. The rela-
tive error between them is 4.6%, which verifies the effectiveness of
the simulation model.

5.3. Cavity length and diameter in the satellite tank


Fig. 6. Theoretical curve and simulation results of projectile velocity.

Fig. 7 presents the evolution of cavity shape when the spherical


material parameters and state equation parameters used are listed projectile penetrates the satellite tank. With the expansion of the
in Table 3 [21,22]. Where, S1 is the slope coefficient of the us-up cavity, the axial growth velocity is higher than the radial growth
curve; C4 and C5 are the polynomial equation coefficients, A is the velocity, which results in a slender cavity. The cavity length ob-
quasi-static yield stress, B is the strain hardening modulus, E is the tained by simulation is 46.9 cm at t ¼ 234 ms, and the cavity radius
elastic modulus, ss is the yield strength, and sb is the tensile obtained by simulation is 9.05 cm at t ¼ 234 ms.
strength. The length and diameter of the cavity at different times are
obtained by simulation results. Meanwhile, the axial and radial
growth models are used to calculate the length and diameter of the
5.2. Verification of the simulation model cavity. The growth curves of the cavity length and diameter are
plotted, the growth model results are compared with the simula-
The impact velocity of projectile is 4000 m/s and the Mach tion results and the relative errors between them are calculated, as
number is Ma ¼ 2.67, so Eq. (7) is used to calculate the projectile shown in Figs. 8 and 9.
velocity theoretically. Moreover, the projectile velocities at As shown in Figs. 8 and 9, the growth trend of the simulation
different times are obtained according to the simulation results. results agrees well with the growth curve. Due to the penetration of
The theoretical curve and simulation results are plotted in Fig. 6. the projectile, the growth velocity of the cavity in the axial direction

Fig. 7. The cavity shape at different times after the projectile enters the satellite tank.
B.-l. Zhao et al. / Defence Technology 16 (2020) 609e616 615

Fig. 8. (a) Growth curve of cavity length and simulation results; (b) the relative errors between growth curve and simulation results.

Fig. 9. (a) Growth curve of cavity diameter and simulation results; (b) the relative errors between growth curve and simulation results.

is obviously greater than that in the radial direction. The growth indicated that the cavity length and diameter at different times in
velocity of the cavity in the axial direction slows down gradually as the satellite tank can be accurately calculated by the axial and radial
a result of the attenuation of projectile velocity. In the numerical growth models.
simulation, the velocity attenuation of projectile is more obvious, so
the cavity length is smaller than the theoretical results when
t > 125 ms. The expansion of cavity squeezes the liquid hydrazine, 6. Conclusions
which generates a radial pressure field in the hydrazine. Hydrazine
has inertia and tries to restore its original state, the radial pressure The axial and radial growth models are used to calculate the
field compresses the cavity in turn. The growth velocity of cavity in cavity length and diameter in Ma L. Y. test. The relative errors
radial direction is smaller, therefore, the cavity is slender, which is between calculations and test results are less than 20%, which
consistent with the cavity shape in Fig. 7. verifies the effectiveness of the axial and radial growth models.
Theoretical calculations indicate that the maximum volume of Theoretical results indicate that the projectile passes through
the cavity is reached at t ¼ 234 ms, the maximum length and the satellite tank at t ¼ 234 ms, and the residual velocity of pro-
diameter of the cavity are 50 cm and 17.45 cm respectively. There jectile is 1345.6 m/s. The simulation results indicate that the
are differences between theoretical calculations and simulation residual velocity of projectile is 1284.3 m/s at t ¼ 234 ms. The
results. The reason is that the growth models are based on a series relative error between them is 4.6%.
of assumptions and simplifications. In the numerical simulation, The cavity shape is slender. The cavity length obtained by
the cavity squeezes liquid hydrazine radially, and the liquid hy- simulation is 46.9 cm and the cavity radius is 9.05 cm at
drazine squeezes tank walls, causing the deformation of tank walls, t ¼ 234 ms. Theoretical calculations indicate that the maximum
so the cavity radius is larger than the theoretical results. In general, volume of the cavity is reached at t ¼ 234 ms, the maximum
the relative errors between simulation results and theoretical re- length and diameter of the cavity are 50 cm and 17.45 cm. The
sults are within the range of (10%, 10%), which are acceptable. It is relative errors between the simulation and theoretical results
are within the range of (10%, 10%).
616 B.-l. Zhao et al. / Defence Technology 16 (2020) 609e616

Declaration of competing interest [10] Disimile PJ, Swanson LA, Toy N. The hydrodynamic ram pressure generated by
spherical projectiles. Int J Impact Eng 2009;36(6):821e9.
[11] Varas D, Zaera R, Lopz-Puente J. Numerical modelling of the hydrodynamic
All the authors declare that they have no conflict of interest. ram phenomenon. Int J Impact Eng 2009;36(3):363e74.
[12] Varas D, Zaera R, Lopz-Puente J. Numerical modeling of partially filled aircraft
References fuel tanks submitted to hydrodynamic ram. Aeroesp Sci Technol 2012;16(1):
19e28.
[13] Lecysyn N, Bony DA, Aprin L, Heymes F, Slangen P, Dusserre G, Munier L,
[1] Varas D, Zaera R, Lopz-Puente J. Experimental analysis of fluid-filled Gallic CL. Experimental study of hydraulic ram effects on a liquid storage tank:
aluminium tubes subjected to high-velocity impact. Int J Impact Eng analysis of overpressure and cavitation induced by a high-speed projectile.
2009;36:81e91. J Hazard Mater 2010;178:635e43.
[2] Worthington AM, Cole RS. Impact with a liquid surface studied by means of [14] Munson BR, Young DF, Okiishi TH, Huebsch WW. Fundamentals of fluid me-
instantaneous photography. Proc R Soc Lond 1900;194:175e99. chanics. sixth ed. 2009. p. 500e3. Hoboken(US).
[3] Lundstrom EA, Fung WK. Fluid dynamic analysis of hydrodynamic ram III. [15] Ma LY, Li XD, Zhou LW, Zhang GF. Study on the characteristics of cavity
Washington DC: Technical Coordinating Group for Aircraft Survivability, AD; formed by fragment impacting liquid filling container. Explos Shock Waves
1976, 031644. 2018;38(6):235e41.
[4] Lee M, Longoria RG, Wilson DE. Cavity dynamics in high-speed water entry. [16] Stepka FS, Morse CR, Dengler RP. Investigation of characteristics of pressure
Phys Fluids 1997;9(3):540e50. waves generated in water filled tanks impacted by high-velocity projectiles.
[5] Ma QP, Wei YJ, Wang C, Zhao CG. Numerical simulation of high-speed water NASA TN D-3143. 1965.
entry cavity of clinders. J Harbin Inst Technol 2014;46(11):24e9. [17] Li Y, Zhang L, Zhu HQ, Zhang W, Zhao PD. Velocity attenuation of blast frag-
[6] Szendrei T. Analytical model of crater formation by jet impact and its appli- ments in water tank. Shipbuild China 2016;57(1):127e37.
cation to calculation of penetration curves and hole profiles. In: Proceedings of [18] Shen HR, Wang WJ, Li YY, Shao QL. Modeling and analysis of spacecraft
the 7th international symposium on Ballistics.Hague, Netherlands; 1983. impact. Beijing: Science Press; 2014. p. 138e40.
p. 575e83. [19] Seddon CM, Moodie K, Thyer AM, Moatamedi M. Preliminary analysis of fuel
[7] Held M. Verification of the equation for radial crater growth by shaped charge tank impact. Int J Crashworthiness 2004;9(3):237e44.
jet penetration. Int J Impact Eng 1995;17(1e3):387e98. [20] Garcia BO, Chavez DJ. Shock compression of liquid hydrazine. Off SciTech Inf
[8] Held M, Huang NS, Jiang D, Chang CC. Determination of the crater radius as a Tech Rep 1995;370(1):831e4.
function of time of a shaped charge jet that penetrates water. Propellants, [21] Zhang KF, Liang MZ, Lu FY, Li XY. Mechanics of plate fracture from detonation
Explos Pyrotech 1996;21(2):64e9. wave interaction. Propellants, Explos Pyrotech 2019;44:188.
[9] Nishida M, Tanak K. Experimental study of perforation and cracking of water- [22] Li YF, Jin QC, Liu ZD, Li G. Finite element analysis of a propellant satellite tank
filled aluminum tubes impacted by steel spheres. Int J Impact Eng based on ANSYS. Aerosp Manuf Technol 2012;12(6):10e5.
2006;32(12):2000e16.

You might also like