j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2

Available online at www.sciencedirect.com

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

Original Article

Dynamic response of aluminium sheet 2024-T3

subjected to close-range shock wave: experimental
and numerical studies

Amin Bassiri Nia a, Ali Farokhi Nejad a,b, Li Xin c, Amran Ayob a,
Mohd Yazid Yahya a,**, Seyed Saeid Rahimian Koloor d,e,*, Michal Petr

u d,
Shukur Abu Hassan a
a
Center for Advanced Composite Materials, School of Mechanical Engineering, University Technology Malaysia,
Malaysia
b
Department of Mechanical and Aerospace Engineering, Politecnico di Torino, Torino, Italy
c
National Laboratory of Solid State Microstructures & Department of Material Science and Engineering, Nanjing
University, Jiangsu, China
d
Institute for Nanomaterials, Advanced Technologies and Innovation (CXI), Technical University of Liberec (TUL),
Studentska 2, Liberec, 46117, Czech Republic
e
Department of Aerospace Engineering, Faculty of Engineering, Universiti Putra Malaysia, Serdang, 43400,
Malaysia

article info abstract

Article history: This present study investigates experimentally and numerically the behaviour of 1 mm
Received 16 October 2020 thick aluminium 2024-T3 alloy sheets from near field shock waves. A comparison and
Accepted 7 December 2020 examination are undertaken with respect to global deformation and plastic damage for-
Available online 13 December 2020 mation from two different stand-off distances of 4 mm and 50 mm that were exposed to a
constant charged mass. A 4-cable instrumented pendulum blast set-up was used to carry
Keywords: out and monitor the blast test. The results of the blast test were subsequently used to
Near-field shock wave simulate the pressure history for different stand-off distances. The simulation involved
Stand-off distance implementing a user subroutine in ABAQUS/Explicit solver to model non-uniform pressure
Dynamic behaviour fields for use in finite element simulation. The results provided a strong alignment of the
Finite element method aluminium numerical method when compared with the experimental data. The main outcome of this
alloy sheet study is to show the significant effect of the changing damage from highly localised
Failure mechanics perforation to global deformation when the stand-off distance is changed from 4 mm to
50 mm.
© 2020 The Author(s). Published by Elsevier B.V. This is an open access article under the CC
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/).

* Corresponding author.
** Corresponding author.
E-mail addresses: [email protected] (M.Y. Yahya), [email protected] (S.S. Rahimian Koloor).
https://doi.org/10.1016/j.jmrt.2020.12.029
2238-7854/© 2020 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://
creativecommons.org/licenses/by-nc-nd/4.0/).
350 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2

1. Introduction

Aluminium alloys are widely used in automotive, aviation,

and naval industries as a result of their high specific strength,
effective energy absorption, excellent corrosion resistance,
and good formability. These alloys can bear dynamic shock
loadingdas in cases of road and air crashes. Moreover,
aluminium sheets are commonly used in functionally graded
material as a face-sheet, where its cladding ability, good
formability, and adhesive bonding all provide manufacturers
incentives to use aluminium sheets in their products.
Many structures have been proposed to protect against
blast loading such as polymeric composite panels, fibre metal
Fig. 1 e The general form of pressure-time history after the
laminates (FML), functionally graded material, etc. However,
explosion [18].
the most dominant responses from structures, in terms of
energy absorption and permanent deflection, can be nar-
rowed down to the metallic panels in protective walls [1,2]. In the validity of such models and employing these models as a
conducting experimental or numerical investigations, many predictive tool for metallic sheets under impulsive loading is
variables may contribute to different effects. Hence, to still limited. This present research investigates the effect of
examine the effect of test variables several tests are required. different parameters in a blast test on aluminium sheet 2024-
Impact testing on metal alone is a challenging task due to the T3. Two values of SOD were considered as close-range
various failure modes that are observed, depending on its impulsive loading on square aluminium panels with all four
ductility [3,4]. Some impulsive experiment tests were con- edges fully clamped. The main objective of this study is to
ducted to investigate the response on steel sheets [5] and the demonstrate that the change of SOD from 4 mm to 50 mm
observations were large inelastic deformation (mode I), tensile causes a significant effect on the damage variation, from a
tearing (mode II) and transverse shear rapture (mode III). highly localised perforation to a global deformation. However,
These failure modes were defined for beams, circular and regarding the case with a constant impulse, the pressure
square plates [6e8]. A reliable theoretical or numerical anal- reduction factor on exposed areas and the pressure-time
ysis on the dynamic response of structures made of different reduction factor plays a crucial role and shows a more
materials is still a formidable task. destructive effect on the target that is subjected to a SOD of
To understand the damage processes that are caused by 50 mm. The tests employed a four-cable ballistic pendulum
localised high rate explosive wave on metallic structures, with corresponding laser displacement to record the impulse.
numerical simulation is an interesting approach [9,10]. Nu- The experiment provided material, geometrical, loading, and
merical models have been used to predict structure behaviour boundary conditions data for the subsequent phase of the
against impulsive loading across numerous investigations. In finite element (FE) model in numerical simulation. The model
the earliest models to simulate localised blasts, a constant was used to simulate different loading conditions and high-
pressure load was applied to an area equivalent to the charged lights dynamic responses and blast pressure distributions
or burn diameter. In this method, the constant pressure was parameters in close-range explosion.
applied in a few milliseconds which was usually taken as the
time for the explosion wave to pass the charge length. This
method proved successful when the geometry was large and 2. Experimental procedure
stand-off distance (SOD) to the sample plates was sufficiently
far enough [11]. A variation of this method was presented by 2.1. Blast overview
Bimha [12] who modelled the impulsive load acting as a con-
stant pressure over the charge area with the decay function to When a solid explosive material is charged, it is converted to a
the edge of the plate. Based on a comparison with the exper- hot and high-pressure gas with a transient maximum tem-
imental plate deflection profiles, the decay constant was perature of about 3000
C and a peak pressure of up to 40 GPa
found for variations in charge radius as a function of plate [16,17]. The high velocity (6000e8000 m/s) compressed air
radius/width. In an extended method by Balden and Nurick produces a shock wave with discontinuities in pressure,
[13] the AUTODYN commercial software was used to obtain density, temperature, and velocity [8,17]. The instantaneously
the pressure profile for a localised blast load. The pressure expanded gas contains extreme levels of compressed air
profile was based on the profile of the maximum pressure that particles which apply pressure on all the surfaces they
was obtained from the simulations. The pressure magnitude encounter. In generating the high pressure, the gas from the
for the profile was adjusted so that the applied impulse cor- charged mass expands rapidly to about 4000 times the original
responded with the measured experimental ballistic volume of the explosive material [18]. The shock front is a
pendulum impulse [14]. This method is one of the most reli- characteristic of a blast wave that changes all the gas dynamic
able methods to simulate the impulsive pressure profile conditions, including static pressure, velocity temperature,
[6e8,15]. Despite valuable efforts to generate consistent density, and flows [19,20]. The maximum amplitude of the
theoretical or mathematical models in overcoming obstacles, shock front pressure can be observed immediately after
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2 351

Fig. 2 e The blast test set up (a) four-cable pendulum blast system. (b) Schematic of the four-cable pendulum.

reaching the encountered surface. Subsequently, it decays the structure in far-field or weak explosions, the peak pressure
exponentially with time and is followed by a negative phase. is calculated in terms that are dependent on the weight (W) and
The shock waves can produce a high impulsive amplitude SOD of the explosive charge (TNT equivalent), by:
over a very short time, typically within 0.01e1 s, depending on

0:33 1:13
the mass and SOD of explosive material [18]. The air shock W
Pmax y53:9 (2)
wave, in turn, applies a high strain rate (typically between D
~100 and 10,000 s1) loading on the structure which can cause Dependent on the SOD position, the air pressure that re-
severe dynamic deformation, vibrations, and damage [21,22]. sults from the shock wave decays over time and with
The peak point of air shock wave followed by a sharp decay, is increasing distance from the blast source. Some parts of air
represented in Fig. 1. pressure are reflected when the shock wave collides with the
The positive region of the overpressure curve decays surface of the specimen and other rigid walls such as the ex-
exponentially over time, and can be calculated by the modi- plosion chamber. This phenomenon generates a negative
fied Friedlander expression [17,23]: pressure below the ambient pressure after the initial positive

 pressure phase. The negative phase tends to be longer in
t at=td
P ¼ Pmax 1  e (1) duration than the positive phase yet at a much lower ampli-
td
tude. Therefore, it needs the additional air particles contained
where P is the impulse pressure at time t, Pmax is the peak in the shock wave to be brought to rest and further com-
pulse when t is zero (i.e. arrival time of the wave front), and td pressed, resulting in a higher reflected overpressure on rigid
is the time period after the wave front. surfaces than the incident overpressure [25]. Generally, posi-
Generally, explosive tests are categorised into three groups tive blast loads are far more damaging to materials and
based on stand-off distance i.e. near-field, mid-range field, and structures than negative phase loads [26]. The loading of a
far-field explosive tests. Many studies have been carried out for shock wave is estimated in terms of the impulse (I) or energy
mid-range [17] and far-field explosive tests [24]. The previous flux density (E), which are respectively defined as [8]:
research investigated complex dynamic structural responses
for far-field air blast and different materials. However, less Zt
attention has been paid to the near-field tests. When an explo- I¼ pðtÞdt (3)
sion occurs in a near-field condition, the shock wave propagates 0

from the source to the structure front face and is reflected after and
the collision. When the shock wave has enough time to reach
352 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2

The impulse intensity for time period (T),

the total mass ðMT Þ and b are specified for the present setup.
b ¼ 0:0268, MT ¼ 165 kg, and T ¼ 3:24. Fig. 3 shows the
recorded oscillating motion data from the pendulum mecha-
nism from the spherical emulsion detonation (10 g) at 4 mm
SOD on the aluminium 2024-T3 sheet. Here, x1 ¼ 20:1848 mm
and x2 ¼ 19:3266 mm. Therefore, from Eq. (8) the impulse is
6.6 N.s.

2.3. Aluminium 2024-T3 under the blast loading

Fig. 3 e Pendulum motion data recorded by laser for 10 g In the present study, aluminium sheets with a 1 mm thickness
spherical emulsion explosive at 4 mm stand-off distance. were subjected to 10 g mass of explosive at different stand-off
distances. Fig. 4 shows the blast test setup of the clamped
Al2024-T3 sample with a 4 mm SOD.
The material properties of AL 2024-T3 from previous
Zt research is taken into account [28,29]. For simulation,
1  
E¼ 1  A Pmax  B P2max P2 ðtÞdt (4) considering the high strain rate loading, the empirical
r0 c0
0 JohnsoneCook model was used in the stress equation, as
expressed in Eq. (9).
where A and B are constant parameters for post-flow velocity
correlation. ro and co are the initial density and speed of sound 
n 
pl 
ε_
is propagating in the air. s ¼ A þ B εpl 1 þ Cln ð1  ðT* Þm Þ (9)
ε_ 0

2.2. Experimental data acquisition where s is the rate-dependent yield stress, εpl is the equivalent
plastic strain, A, B and n are material parameters, C and ε_ 0 are
The blast load was generated by detonating the emulsion the strain-rate constant and the reference strain rate,
explosive which is spherical in shape, resulting in an impulse respectively. These material constants are extracted below, at
pl
velocity of 5200 m/s. The four-cable pendulum blast system transient temperature (T*). In Eq. (10), εD is the equivalent
(Fig. 2a) was used to measure the blast intensity. The plastic strain at damage initiation, and d1 to d4 are failure
aluminium sheets were clamped in front of the ballistic parameters. The JohnsoneCook model parameters for the AL
pendulum, exposing an area of 260 mm  260 mm facing the 2024-T3 alloy are summarised in Table 1.
charge. The explosive was located in line with the centre of 

_ 
aluminium sheet at 4 mm and 50 mm stand-off distances. pl d3 P εpl
εD ¼ d1 þ d2 exp 1 þ d4 ln (10)
This setup is designed to provide different blast pressure q ε_ 0
distribution on the exposed area of the sheet panels (Fig. 2b).
The impulse is calculated from the oscillating movement of
the pendulum from the explosion. The movement of the 3. Numerical model
pendulum was recorded using laser Micro-Epsilon LD1607-
200. Fig. 2b represents the block diagram of input variables
3.1. Shock wave modelling
from experiment to analytical solution.
A significant structural deformation occurs during an infini-
Furthermore, the motion formula for a four-cable
tesimal time-period and it undergoes two response phases.
pendulum blast system [26,27], is used to calculate the im-
The first phase is the compression through-thickness and the
pulse from detonation. The general solution for beam move-
ment is shown by Eq. (5).

X ¼ ebt x0 = u sinðutÞ (5)

where X is global displacement, b is the time constant, and x_0

is local initial speed. Two specific solutions, for time t1 ¼ T=4
and t2 ¼ 3T=4, are used as shown by Eqs. (6) and (7)
respectively.
bT
X1 ¼ e 4 x_0 T=2p (6)

3bT
X2 ¼ e 4 x_0 T=2p (7)

and finally, the impulse (I) can be calculated by the following

expression,

2 ln xx12 x1 2p bT=4 Fig. 4 e Aluminium sheet 2024-T3 clamped in the fixture

b¼ ; I ¼ MT x_0 ¼ MT e (8)
T T for the blast test.
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2 353

Fig. 5 e The pressure distribution pattern over different positions of the Al 2024-T3 sheet sample.

second is the overall time response. Therefore, tracing the locations to find the pressure-time history is essential. Gauges
pressure pulse over time should be considered as an impor- 1 to 5 are 4 mm apart, and Fig. 7a shows that the pressure
tant factor to model shock waves. The pressure history is distribution between the gauges is almost uniform. Gauges 6
highly dependent on the SOD position and mass of charge. In to 12 cover 84 mm, from 6 mm to 90 mm (Fig. 7b). The
this study, in order to model the localised impulsive (blast) approximate pressure distribution at a radial distance from
loading an explicit form of AUTODYN was used to find the the centre of the plate and related acting time factor are
pressure distribution as a function of position and time [14]. shown in Fig. 8a and b [30].
The pressure can be a function of location which means for Figs. 7 and 8 both show that by increasing the time and
different positions of the exposed areas the pressure differs. distance from the centre of panel, the pressure degrades
The pressure distribution pattern over different positions of exponentially by a function which has been used by different
the sample is shown schematically in Fig. 5. Moreover, the researchers [10,31]. For higher SOD the first part of pressure
pressure is a function of time which means for close-field acting time is negligible. Fig. 9 shows the results of the pres-
explosions the first part of the pressure time-factor in- sure history for 50 mm SOD and 10 g explosive. The maximum
creases and in the second part the pressure time-acting factor peak reflected pressure of 60 MPa was reached and the time
decreases by an exponential function. reduction factor was obtained from AUTODYN solution. The
Hence, the basic input to the numerical simulation is to find time duration was 55 ms (from 20 ms to 75 ms). Figs. 7e9 show
the pressure distribution as a function of time and location, the approximation functions of applied pressure histories for
from the explosion of the charged mass. The blast test is SODs 4 mm and 50 mm. The maximum standard error for all
modelled in AUTODYN using the 2D explicit nonlinear model cases was less than ±4% which are indicated with error bars.
by couple method (modelling air and explosive separately). The Furthermore, Eqs. (11) and (12) present the pressure-
air is modelled as an ideal gas. The material properties for air history function and time-history function for 10 g emulsion
and aluminium 2024-T3 are listed in Table 1. A 14.6 mm radius explosive at 4 mm SOD. In the following equations, the two
spherical shape of the explosive is modelled, and the EOS-JWL parameters xk and z0 have a significant role in influencing the
parameters for explosive are selected from [29]. The individual pressure distribution. The pressure decay factor on position
gauges are defined in the software to provide the pressure-time (xk ) which influences the pressure degradation on the exposed
history and pressure-position history. Fig. 6a illustrates the area is sensitive to stand-off distance SOD. The pressure decay
model of spherical emulsion (10 g) in 4 mm stand-off distance factor with time (z0 ) is used to control the effect of time on the
and Fig. 6b shows the pressure contour after 150 ms. pressure after the explosion [10].
From the pressure acting-time history for individual 8
gauges the approximation pressure pattern from AUTODYN < P0 if r  r0
PðrÞ ¼ P0 exkðrr0Þ if r0 < r  rb (11)
can be found [11]. When the mass is charged in very close :
0 if r > rb
range, the impulse pressure does not have enough time to
convert to a uniform distribution over the sheet blank. where, r0 ¼ 4 mm, xk ¼ 0:0675 1
mm
, rb ¼ 90 mm are found
Therefore, applying different individual gauges in different from approximation function.

Table 1 e AL 2024-T3 Johnson Cook parameters.

Elastic properties JohnsoneCook parameters Damage parameters
parameters r E w A B C n m d1 d2 d3 d4
unit (kg/m3) (GPa) e (MPa) (MPa) e e e e e e e
value 2690 74 0.29 369 684 0.0083 0.73 1.7 0.13 0.13 1.5 0.011
354 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2

Fig. 6 e (a) AUTODYN 2D model for 10 g emulsion and (b) pressure distribution for after 150 ms.

where arrival time (treach ) ¼ 3 ms, the time at peak reflected

8 pressure (t peak ) ¼ 5.5 ms, and t death ¼ 25 ms, z0 ¼ 7.8 (unitless) are
> 0 if time  t death
< timet
>
if t reach < time  t peak extracted from the AUTODYN output.
reach
PðtÞ ¼ z0ðtimet peak Þ=t death
(12)
>
> e if t peak < time  t death For 50 mm SOD, the general form of pressure distribution
:
0 if t > t death function on exposed areas is similar to 4 mm SOD with the

Fig. 7 e AUTODYN pressure history for different gauges (a) for center to 4 mm (b) for 6 mme90 mm.
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2 355

Fig. 8 e Pressure approximation pattern for 10 g emulsion in 4 mm SOD (a) time (b) radial distance.

difference in exponential decay functions, r0 , and rb . From the which effects time degradation on the pressure after the ex-
approximation pattern of pressure time history at 50 mm SOD plosion is chosen at 4.2, tpeak ¼ 20 ms and tdeath ¼ 75 ms. By
(Fig. 10) the pressure-time function is changeddas repre- comparing the two SODs, it is evident for the same mass of
sented by Eq. (13). explosive the reflected peak pressure in 4 mm SOD is higher
8 than 50 mm SOD but the duration time and exposed area is
>
> 0; time  tpeak
>
>

higher for 50 mm SOD. The total impulse loading can be found
>
>
< timetpeak by the following equation [32]:
k2
pðtÞ ¼ tdeath (13)
>
> e ; tpeak < time  tdeath
>
> Z∞ Z r
>
>
: I ¼ 2p rPðr; tÞdrdt (14)
0; time > tdeath
0 0
From AUTODYN, the pressure is constant for 30 mm from
where I is total impulse, which can be found from the exper-
the center of the plate (r0 ¼ 30 mm) and rb ¼ 150 mm, which
iment, t is time and r is the distance from the centre of plate.
means the blast pressure uniformly covers the whole surface
When the explosive is located at very close range, the impulse
of the aluminum sheet at 50 mm SOD. The degradation factor
pressure does not have enough time to convert to a uniform
in pressure distribution (K1 ) is chosen at 0.027/mm, K2 factor
356 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2

Fig. 9 e Pressure-time distribution for SOD 50 mm and 10 g emulsion.

distribution over the sheet blank. Therefore, applying uniform pressure history as a function of time and distribu-
different individual gauges in different locations to find the tion, a FORTRAN-based computer code was implemented in
pressure-time history is essential. By substituting the pres- ABAQUS from a user-defined subroutine VDLOAD [33]. Using
sure and time history functions for corresponding stand-off Eqs. (11)e(14), the results were used to estimate the impulse
distances in Eq. (13), the values of maximum reflected pres- energy for different SODs and charged mass. To reduce the
sure for different SODs can be found. The maximum reflected computational time a one-quarter symmetry model was
pressure for 10 g spherical emulsion at 4 mm and 50 mm SODs generated and analysed. Fig. (10) shows the 3D finite element
are 1160 MPa and 60 MPa, respectively. model analysed in this study.
The model was built up with continuum 3D 8-noded ele-
3.2. Finite element model ments with reduced integration ability (C3D8R). A mesh
convergence study, regarding the stable plastic strain, was
To simulate the high-speed deformation process, a solid carried out and the minimum element size of 1.5 mm was
three-dimensional model was generated. To produce the non- selected for the exposed area. The JohnsoneCook material

Fig. 10 e One quarter FE model for AL2024-T3 under impulsive loading.

 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2 357

Fig. 11 e Comparison between FE modelling and experimental results of Al-10-4 exposed to impulse 6.6 N.s.

model in Table 1 was assigned to the FE model. The damage the cracks around perforation hole grow and the outward
initiation and element deletion processes were adhered to tulips are clearly observed. The permanent deformation and
according to Eq. (10). The energy balance analysis was per- damage of the aluminium sheet is shown in Fig. 12.
formed to justify the numerical procedure [34e36]. The usual
boundary conditions were applied at the planes of symmetry. 4.3. Effect of mid-range impulse
The outer edges were fixed in all degrees of freedom and the
inner edges were constrained along x- and y-axes respectively. When the SOD is increased from 4 mm to 50 mm the impul-
The model was free to move in the z-direction. sive energy from the charged mass of 10 g is reduced to 3.6 N.s.
The reduced impulsive energy as well as the larger SOD helps
the specimen from being damaged by permanent plastic
4. Results deformation only. From previous research on the SOD value of
either 4 mm or 50 mm are considered as near-field range ex-
4.1. Effect of close range impulse plosions. However, changing the SOD from very close range to
mid-range, the shock wave the structure behaves differently.
The shape deformed from 4 mm SOD impulse, from the test Therefore, for a complete structural integrity assessment a
and FE simulation are compared in Fig. 11. The figure shows further study is necessary on different ranges of shock waves.
hexagon-shaped perforation with petalling in 8 tulips. The Fig. 13 compares the permanent deformed shape from
overall damage, in terms of deformation and perforation experimental test and from FE modelling when the structure
shape, shows good correlation between the experimental and is placed at 50 mm SOD. The close correlation of maximum
the FE simulation results. It was observed that the release of deflection from FE modelling and blast tests show that the
impulsive energy in very close range to the specimens caused applied shock wave from the blast test has been selected
all the energy to fracture the sample. There is no sign of global correctly and the FE model is validated.
lateral deformation of the sample. In a very short time interval
after detonation (less than 5 ms) the crack initiation process
began and propagated until the shock wave is nullified. 5. Discussion

4.2. Transient response in close-range impulse The results for the different test conditions are summarised in
Table 2. Two main types of damage have been observ-
The transient response of sample Al-10-4 exposed to the eddglobal deformation with cap, and petalling. In the case of
close-field impulse of 6.6 N.s at 4 mm SOD is difficult to obtain 4 mm SOD, the samples do not show any wrinkles around the
experimentally. The difficulty of monitoring the transient re- boundary, but by changing the SOD to 50 mm many winkles
sponses of samples in blast tests is the main reason to appeared. Moreover, for 50 mm SOD, in-plane bulking appears
consider a solution via FE modelling. The transient response is and the presence of plastic hinge lines from the corner of the
demonstrated as vertical displacement counterplots. At time boundary to the centre of the aluminium sheet (along a 45-
step 4 ms after detonation the acting pressure triggers perfo- degree direction). No perforation is observed for this sample
ration. When the incrementation time is increased to 800 ms, and the maximum back-face deflection reaches 34.8 mm. In
358 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2

Fig. 12 e The transient response of aluminum sheet in impulse of 6.7 N.s on 4 mm SOD.
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2 359

Fig. 13 e Comparison between FE and experimental results of Al-10-50 expose to impulse 3.6 N.s.

metal forming applications, the present study can be applied times higher than SOD 50 mm, the damage in SOD 50 mm is
to calculate the optimum amount of explosive and SOD so as much more severe (comparing the results from Figs. 11 and
to avoid the presence of boundary wrinkles. 15). The reason is that, the pressure degradation factor on
the exposed area and the degradation time factor. At a SOD of
5.1. Effect of different impulse energies 50 mm, the degradation time factor is 46%, and the pressure
degradation factor is 60% less than corresponding factors at
The initial pressure (P0) is one of the effective parameters in SOD 4 mm. The degradation from these two parameters cause
impulse calculation. Substitution of different values of P0 in more exposed area and more time-engagement with the
Eq. (11) leads to different values of impulse, and for the same impulsive loading and finally leads to more damage to the
SOD the maximum deflection and panel response can be structure. Therefore, it cannot be directly interpreted that, by
different. Fig. 14 compares the different values of plastic increasing the stand-off distance the structure can be pro-
deformation with different P0 and impulse energies. The tected against impulsive loading. As observed and explained,
values of P0 ¼ 60 MPa and I ¼ 3.72 N.s were the closest cases to these two parameters have crucial roles and should be
the experimental test. The results from the experimental considered, especially in cladding structure design. For the
repetition (five replicates for each SOD) in Fig. 14 are presented explosive in the cylinder shape, the researchers investigated
with ±3% error that is shown with error bars. the blast pressure characterisation at a constant stand-off
By increasing the impulse to 6.6 N.s in 50 mm SOD, the distance of 14 mm [10] and 50 mm [31]. By comparing the
maximum reflected peak pressure reaches 108 MPa. The pressure degradation factor for cylinder explosives [31] and
aluminium sheet petals in 8 tulips. The results for the corre- spherical explosives in this work, it is evident the pressure
sponding sample are shown in Fig. 15. degradation factor for the cylinder explosive is 0.067 1/mm
When the impulse for SODs at 4 mm and 50 mm is kept at a and for spherical explosive is 0.027 1/mm, which means in a
constant of 6.6 N.s, the peak reflected pressures for corre- mid-range explosion (SOD 50 mm) the pressure degradation
sponding SODs are 1160 MPa and 108 MPa, respectively. for spherical explosive is less than the cylinder shap, and will
Despite the peak reflected pressure for SOD 4 mm is around 10 engage an increased exposed area.

Table 2 e Stand-off distance, weight of explosive, the calculated impulse for the corresponding mass of explosive,
maximum back face deflection and failure mode.
AL sheet code Mass of charge (g) SOD (mm) Impulse (N.s) Failure mode deflection (mm)
Al-10-4 10 4 6.6 Tulip (petalling) e
Al-10-50 10 50 3.6 Global deformation 34.8
360 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2

Fig. 14 e Comparison between FE modelling and experimental results of Al-10-50 expose to different impulse energies.

Fig. 15 e The results for SOD 50 mm when expose to impulse 6.6 N.s.

material constitutive models and failure criteria were able to

6. Conclusion capture some high strain-rate failure features in the
aluminium alloy sheets, such as petalling and perforation
Aluminium 2024-T3 alloy sheets were used in a blast test with failure. By using the FE model more attention was focussed on
different stand-off distances and a constant charged mass, to the transient deformation process to analyse sheet behaviour.
investigate the effect of near-field shock waves. The transient For stand-off distance, the effect of an increase from 4 mm to
behaviour was measured through permanent plastic defor- 50 mm becomes significant when the damage changes from a
mation and perforation damage. A 4-cable instrumented highly localised perforation to a global deformation. This im-
pendulum blast set-up was prepared to carry out the test. The plies that an instant failure model can be a useful tool to
results of the blast test were used to simulate the pressure assess blast failures of metallic panel face-sheets in different
history from different SODs. A numerical finite element kinds of composite stacking. In small stand-off distance the
analysis was then carried out using commercial code in peak reflected pressure is a critical parameter, while by
ABAQUS/Explicit. A user-defined subroutine was imple- increasing the SOD, the duration time and exposed area are
mented in the ABAQUS solver to model non-uniform pressure increased and these parameters are more critical than peak
field. The FE modelling prediction was then compared with reflected pressure and need to be considered in structure
the experimental results. The comparisons indicate that the design.
 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2 361

The findings are summarised as follow: and fundamental grant scheme (Grant No. RJ
130000.7351.4B553). The authors wish to thank National Natural
 In the experiment section: Science Foundation of China for their support (Grant
o The effect of SOD on the aluminium sheet, subjected to No.11772215) as the technical and experimental assistance. The
the constant mass of a spherical shape emulsion explo- research was funded by the Ministry of Education, Youth, and
sive was studied from close-to mid-range explosions. Sports of the Czech Republic and the European Union (European
The results indicate that by increasing the SOD, the Structural and Investment Funds Operational Program
damage form is dominated by global deformation rather Research, Development, and Education) in the framework of the
than localised perforation. project “Modular platform for autonomous chassis of specialized
o The effect of spherical explosive geometry on the blast electric vehicles for freight and equipment transportation”, Reg.
pressure characterisation were highlighted by a com- No. CZ.02.1.01/0.0/0.0/16_025/0007293, as well as the financial
parison with available results. support from internal grants in the Institute for Nanomaterials,
 In the FE section: Advanced Technologies and Innovations (CXI), Technical Uni-
o The blast pressure distribution and pressure-time func- versity of Liberec (TUL). It was also supported by Universiti Putra
tions for close- and mid-range explosions were found Malaysia under Putra Grant No. GP/2018/9635100.
from AUTODYN and implemented into ABAQUS through
a VDLOAD subroutine to validate the experimental
results.
references
o The blast pressure characterisation such as peak reflected
pressure, pressure-radial position degradation factor, and
[1] Li Y, Lv Z, Wang Y. Blast response of aluminum foam
pressure-time degradation factor were highlighted for
sandwich panel with double V-shaped face plate. Int J Impact
close-range and mid-range explosions. The results indi-
Eng 2020 Jul 15:103666.
cate that for the same mass of spherical explosive (10 g), [2] Nia AB, Yahya MY, Ayob A, Nejad AF. Optimization of graded
the reflected peak pressure at 4 mm SOD is higher than metallic foam subjected to impulsive loading through DOE
50 mm SOD, but the duration time and exposed area is approach. In: In 2018 9th international conference on
higher for 50 mm SOD and causes the damage to shift mechanical and aerospace engineering (ICMAE). IEEE; 2018
from localised perforation to global deformation. Jul 10. p. 295e9.
[3] Pineau A, Benzerga AA, Pardoen T. Failure of metals I: brittle
o The effect of blast pressure characterisations at mid-
and ductile fracture. Acta Mater 2016 Apr 1;107:424e83.
range explosion was discovered by promoting the im- [4] Sadighi M, Alderliesten RC, Benedictus R. Impact resistance
pulses. The results indicate that by increasing the im- of fiber-metal laminates: a review. Int J Impact Eng 2012 Nov
pulse, the in-plane buckling disappears and the damage 1;49:77e90.
of the target is dominated by increased global deforma- [5] Jacob N, Nurick GN, Langdon GS. The effect of stand-off
tion the effect of SOD was studied by exposing the target distance on the failure of fully clamped circular mild steel
to the same impulse load. The roles of pressure degra- plates subjected to blast loads. Eng Struct 2007 Oct
1;29(10):2723e36.
dation and time-degradation factors in mid-range ex-
[6] Langdon GS, Lemanski SL, Nurick GN, Simmons MS,
plosions are more crucial as the results indicate a higher Cantwell WJ, Schleyer GK. Behaviour of fibreemetal
SOD will cause the target to suffer more damage. laminates subjected to localised blast loading: part I e
experimental observations and failure analysis. Int J Impact
Eng 2007;34:1202e22.
[7] Lemanski SL, Nurick GN, Langdon GS, Simmons MS,
Cantwell WJ, Schleyer GK. Behaviour of fibreemetal
Data availability statement
laminates subjected to localised blast loading: part II e
quantitative analysis. Int J Impact Eng 2007;34:1223e45.
The raw/processed data required to reproduce these findings [8] Langdon GS, Nuric GN, Lemanski SL, Simmons MS,
cannot be shared at this time as the data also forms part of an Cantwell WJ, Schleyer GK. Failure characterisation of blast-
ongoing study. The relevant data can be made available on loaded fibreemetal laminate panels based on aluminium and
request. glass-fibre reinforced polypropylene. Compos Sci Technol
2007;67:1385e405.
[9] Mehreganian N, Louca LA, Langdon GS, Curry RJ, Abdul-
Karim N. The response of mild steel and armour steel plates
Declaration of Competing Interest to localised air-blast loading-comparison of numerical
modelling techniques. Int J Impact Eng 2018 May 1;115:81e93.
The authors declare that they have no known competing [10] Sitnikova E, Guan ZW, Schleyer GK, Cantwell WJ. Modelling
financial interests or personal relationships that could have of perforation failure in fibre metal laminates subjected to
high impulsive blast loading. Int J Solid Struct 2014 Sep
appeared to influence the work reported in this paper.
1;51(18):3135e46.
[11] Bonorchis D, Nurick GN. The influence of boundary
conditions on the loading of rectangular plates subjected to
localised blast loading - importance in numerical
Acknowledgements
simulations. Int J Impact Eng 2009;36:40e52.
[12] Bimha RE. Response of thin circular plates to central blast
This work was supported by Universiti Teknologi Malaysia loading. Doctoral dissertation. University of Cape Town;
collaborative research grant (Grant No. Q.J130000.2409.07G99) 1996.
362 j o u r n a l o f m a t e r i a l s r e s e a r c h a n d t e c h n o l o g y 2 0 2 1 ; 1 0 : 3 4 9 e3 6 2

[13] Balden VH, Nurick GN. Numerical simulation of the post- [27] Senthil K, Iqbal MA, Arindam B, Mittal R, Gupta NK. Ballistic
failure motion of steel plates subjected to blast loading. Int J resistance of 2024 aluminium plates against hemispherical,
Impact Eng 2005 Dec 1;32(1e4):14e34. sphere and blunt nose projectiles. Thin-Walled Struct 2018
[14] ANSYS/Autodyn-2D and 3D, Version 6.1. User Documentation. May 1;126:94e105.
[15] Syed ZI, Mohamed OA, Rahman SA. Non-linear finite [28] Peirovi S, Pourasghar M, Nejad AF, Hassan MA. A study on
element analysis of offshore stainless steel blast wall under the different finite element approaches for laser cutting of
high impulsive pressure loads. Procedia Eng 2016 Jan aluminum alloy sheet. Int J Adv Manuf Technol 2017 Oct
1;145:1275e82. 1;93(1e4):1399e413.
[16] Mahmood Y, Dai K, Chen P, Zhou Q, Bhatti AA, Arab A. [29] Castedo R, Natale M, Lo  pez LM, Sanchidria n JA, Santos AP,
Experimental and numerical study on microstructure and Navarro J, et al. Estimation of Jones-Wilkins-Lee parameters
mechanical properties of Ti-6Al-4V/Al-1060 Explosive of emulsion explosives using cylinder tests and their
Welding. Metals 2019 Nov;9(11):1189. numerical validation. Int J Rock Mech Min Sci 2018 Dec
[17] Mouritz AP. Advances in understanding the response of 1;112:290e301.
fibre-based polymer composites to shock waves and [30] Karagiozova D, Langdon GS, Nurick GN, Yuen SC.
explosive blasts. Compos Appl Sci Manuf 2019 Oct Simulation of the response of fibreemetal laminates to
1;125:105502. localised blast loading. Int J Impact Eng 2010 Jun
[18] Cimpoeru SJ, Ritzel DV, Brett JM. Physics of explosive loading 1;37(6):766e82.
of structures. In: In explosion blast response of composites. [31] Langdon GS, Karagiozova D, Von Klemperer CJ, Nurick GN,
Woodhead Publishing; 2017 Jan 1. p. 1e22. Ozinsky A, Pickering EG. The air-blast response of sandwich
[19] Dewey JM. Measurement of the physical properties of blast panels with composite face sheets and polymer foam cores:
waves. In: In experimental methods of shock wave research. experiments and predictions. Int J Impact Eng 2013 Apr
Cham: Springer; 2016. p. 53e86. 1;54:64e82.
[20] Needham CE. Blast wave propagation. In blast waves. Cham: [32] Bassiri Nia A, Xin L, Yahya MY, Ayob A, Farokhi Nejad A,
Springer; 2018. p. 97e120. Rahimian Koloor SS, et al. Failure of glass fibre-reinforced
[21] Hunter KS, Geers TL. Pressure and velocity fields produced by polypropylene metal laminate subjected to close-range
an underwater explosion. J Acoust Soc Am 2004 explosion. Polymers 2020;12:2139.
Apr;115(4):1483e96. [33] Abaqus V. 6.14 documentation, vol. 651. Dassault Systemes
[22] Kinney GF, Graham KJ. Explosive shocks in air. Springer Simulia Corporation; 2014. 6-2.
Science & Business Media; 2013 Nov 11. [34] Farokhi Nejad A, Alipour R, Shokri Rad M, Yazid Yahya M,
[23] Sanborn MJ. Experimental methods for understanding the Rahimian Koloor SS, Petr
u M. Using finite element approach
behavior and residual capacity of bolts and steel bolted for crashworthiness assessment of a polymeric auxetic
connections under impulsive loads. Doctoral dissertation. structure subjected to the axial loading. Polymers
Georgia Institute of Technology; 2018. 2020;12:1312.
[24] Tekalur SA, Bogdanovich AE, Shukla A. Shock loading [35] Farokhi Nejad A, Chiandussi G, Solimine V, Serra A. Study of
response of sandwich panels with 3-D woven E-glass a synchronizer mechanism through multibody dynamic
composite skins and stitched foam core. Compos Sci Technol analysis. Proc Instit Mech Eng Part D: J Auto Eng
2009 May 1;69(6):736e53. 2019;233(6):1601e13.
[25] Ngo T, Mendis P, Gupta A, Ramsay J. Blast loading and blast [36] Nejad Farokhi A, Chiandussi G, Solimine V, Serra A.
effects on structuresean overview. Electron J Struct Eng 2007 Estimation of the synchronization time of a transmission
Jan;7(S1):76e91. system through multi body dynamic analysis. Int J Mech Eng
[26] Jing L, Wang Z, Zhao L. Measurement of impulse acted on a Robot Res 2017;6(3):232e6.
structure subjected to blast loading. J ExpMech
2009;24:151e6.