You are on page 1of 29

See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/222999972

Constitutive behaviour of anisotropic materials under shock loading

Article  in  International Journal of Plasticity · January 2008


DOI: 10.1016/j.ijplas.2007.02.009

CITATIONS READS

47 311

1 author:

Alexander A. Lukyanov
Harvard Medical School, Boston, United States
70 PUBLICATIONS   369 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Deflated Krylov methods View project

A Fully Implicit Slug Capturing Algorithm for Multiphase Pipe Flows View project

All content following this page was uploaded by Alexander A. Lukyanov on 01 July 2018.

The user has requested enhancement of the downloaded file.


Available online at www.sciencedirect.com

International Journal of Plasticity 24 (2008) 140–167


www.elsevier.com/locate/ijplas

Constitutive behaviour of anisotropic


materials under shock loading
Alexander A. Lukyanov *
Crashworthiness, Impact and Structures Mechanics (CISM) Group, School of Engineering,
Cranfield University, Cranfield, Bedford MK43 0AL, UK

Received 2 August 2006; received in final revised form 1 February 2007


Available online 12 March 2007

Abstract

Thermodynamically and mathematically consistent constitutive equations suitable for shock


wave propagation in an anisotropic material are presented in this paper. Two fundamental tensors
aij and bij which represent anisotropic material properties are defined and can be considered as gen-
eralisations of the Kronecker delta symbol, which plays the main role in the theory of isotropic mate-
rials. Using two fundamental tensors aij and bij, the concept of total generalised ‘‘pressure” and
pressure corresponding to the thermodynamic (equation of state) response are redefined. The equa-
tion of state represents mathematical and physical generalisation of the classical Mie–Grüneisen
equation of state for isotropic material and reduces to the Mie–Grüneisen equation of state in the
limit of isotropy. Based on the generalised decomposition of the stress tensor, the modified equation
of state for anisotropic materials, and the modified Hill criteria, combined with the associated flow
rule, a system of constitutive equations suitable for shock wave propagation is formulated. The
behaviour of aluminium alloy 7010-T6 under shock loading conditions is considered. A comparison
of numerical simulations with existing experimental data shows good agreement of the general pulse
shape, Hugoniot Elastic Limits (HELs), and Hugoniot stress levels, and suggests that the constitutive
equations are performing satisfactorily. The results are presented and discussed, and future studies
are outlined.
Ó 2007 Elsevier Ltd. All rights reserved.

Keywords: A. Shock waves; B. Anisotropic material; B. Constitutive behaviour; C. Plate impact; Equation of
state

*
Tel.: +44 0 7840 355 383; fax: +44 0 1235 532 543.
E-mail address: aaluk@mail.ru

0749-6419/$ - see front matter Ó 2007 Elsevier Ltd. All rights reserved.
doi:10.1016/j.ijplas.2007.02.009
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 141

1. Introduction

Investigation of anisotropic material behaviour (such as aluminium alloys, composite


materials) has found significant interest in the research community due to the widespread
application of anisotropic materials in aerospace and civil engineering problems. For
example, aluminium alloys are one of the main materials in the construction of modern
aircraft and rockets. The strain rate dependent mechanical behavior of anisotropic mate-
rial (e.g. aluminium alloys) in air and space vehicles is important for applications involving
impact. These applications cover a wide range of situations such as crashworthiness and
protective armours in air and space vehicles and other applications. Since shock wave
and high-strain-rate phenomena are involved in many physical phenomena, we are inter-
ested in understanding the material mechanical properties under these non-trivial condi-
tions (Wallace, 1981; Chidambaram and Sharma, 1991).
The purpose of this paper is the investigation of the shock wave propagation in aniso-
tropic solids, and more specifically, for anisotropic elastic–plastic solids. This is a subject
which has received considerable attention in the isotropic solid-state physics and mechan-
ics literature in recent decades (e.g. Wackerle, 1962; Zel’dovich and Raizer, 1966; Davi-
son and Graham, 1979; Eliezer et al., 1986; Asay and Shahinpoor, 1993; Meyers, 1994;
Drumheller, 1998). Some attempts have been made by Anderson et al. (1994) to describe
the constitutive relationships for an anisotropic material. To describe the anisotropic
material response under shock loading the following general aspects need to be investi-
gated: appropriate constitutive equations to describe the strength effect (e.g. Lukyanov,
2006) and constitutive equations to describe the equation of state. Mechanical yielding
and strength behavior in shock waves show complexities that are not understood yet,
especially in anisotropic materials. The determination of the equation of state for aniso-
tropic materials is an important problem in metallurgy, geophysics, aerospace, and also
in other areas where the behaviour of anisotropic materials at high pressures is of
interest.
Modern, high-resolution methods to monitoring the stress and particle velocity histo-
ries in shock waves and equipment have been created (e.g. Barker and Hollenbach,
1972; Kanel, 1998; Kanel et al., 1998; Millett and Bourne, 2001; Bourne and Stevens,
2001; Bourne, 2003; Gu and Ravichandran, 2006); numerous investigations into the
mechanical properties of different classes of materials have been undertaken (e.g. Stein-
berg, 1991; Meyers, 1994; Johnson et al., 1994; Kanel et al., 1997; Millett et al., 2002; Lop-
atnikov et al., 2004; and Zaretsky et al., 2005; Gebbeken et al., 2006; Gu and
Ravichandran, 2006; Bronkhorst et al., 2006), and numerous phenomenological as well
as microscopic models have been developed (e.g. Wallace, 1981; Steinberg et al., 1980;
Swegle and Grady, 1985; Meyers, 1994; Kanel et al., 1995; Nellis et al., 2003; Bourne
and Gray, 2003; Krüger et al., 2003; and Chijioke et al., 2005; Boidin et al., 2006; Petit
and Dequiedt, 2006). However, in spite of a perfectly adequate general understanding,
experimental methodology, and theory, material models do not agree in detail, especially
for anisotropic materials.
For many years, it has been assumed that the response of materials to shock loading is
isotropic, and only recently has anisotropy in the shock response attracted the attention of
researchers (e.g. Gray et al., 2000, 2002; Vignjevic et al., 2002). It was shown in an inves-
tigation of cold rolled and annealed zirconium (see Gray et al., 2000) that the value of
stresses varies in different directions in the quasi-static test and plate impact test. Gray
142 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

et al. (2000) showed that under shock loading conditions (one-dimensional strain space),
the variation of the Hugoniot elastic limit (HEL) or the yield strength of annealed zirco-
nium was consistent with the quasi-static experimental data. Later on, Butcher (1968)
worked on aluminum alloy 6061-T6 and predicted that spall strength should vary in accor-
dance with the one-dimensional stress yield strength depending on material orientation.
Here it is important to mention the work done by Johnson and Barker (1969), Stevens
and Tuler (1971), Rubin (1990) on aluminum alloy 6061-T6 and Rosenberg et al. (1983,
1980) on the shock response of the alloy 2024-T86.
A technique for the study of the rate-sensitive behaviour of metals at high rates-of-
strain is the planar plate impact test (one-dimensional shock wave propagation). The
shock wave experiment has certain potential advantages associated with the level of strain
rate which can be induced and the commonly accepted belief that no geometrical disper-
sion effects occur. It has frequently provided the motivation for the construction of mate-
rial constitutive relations and has been the principal means for determining material
parameters for some of these relations (Steinberg, 1991; Meyers, 1994; Bourne and Gray,
2003; Chijioke et al., 2005; Gu and Ravichandran, 2006).

2. Existing model of anisotropic material (orthotropic material)

In this section a conventional approach for anisotropic material (orthotropic material)


is presented. It is well known that elastic isotropic materials are characterized by two con-
stants (Lamé parameters). In the case of elastic orthotropic material, nine constants are
required to describe the constitutive relationship. The strain–stress relationship for an elas-
tic orthotropic material is as follows:
0 m 1
2 3 1
Ex
 Eyxy  mEzxz 0 0 0 2 3
exx B mxy C rxx
B 1 m C6
6 eyy 7 B Ex Ey
 Ezyz 0 0 0 C6 ryy 7
6 7 B C6 7
6 7 m
C6 r 7
6 ezz 7 B  mExzx  Eyzy 1
0 0 0 C6 zz 7
6 7¼B Ez
7; ð1Þ
6 2e 7 B C6
C6 rxy 7
6 xy 7 B
1
0 0 0 0 0 7
6 7 B
B
Gxy C6
C4 r 7
4 2exz 5 B 0 0 0 0 1
0 C xz 5
@ Gxz A
2eyz 1 ryz
0 0 0 0 0 Gyz

where Ex ; Ey ; Ez are the Young’s moduli; Gxy ; Gxz ; Gyz are the shear moduli; and mij are the
six Poisson ratios. Due to symmetry conditions, the equations may be used to reduce the
number of constitutive constants to nine:
mij mji
¼ ; ðno sum on i 6¼ j and j 6¼ iÞ: ð2Þ
E i Ej

2.1. Conventional decomposition

Speaking generally, any second order tensor can be decomposed into the spherical part
and the deviatorical part. In the case of continuum mechanics, the decomposition of the
stress tensor and the strain tensor into their volumetric and deviatoric components (e.g.
Malvern, 1969; Wilkins, 1969; Sedov, 1972; and Anderson et al., 1994) has certain physical
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 143

justification. This step is done in order to distinguish between thermodynamic (equation of


state) response and the ability of the material to carry shear loads (strength) (e.g. Malvern,
1969; Wilkins, 1969; Sedov, 1972). The distortional change can be represented by the devi-
atoric strain tensor edij ; edij ¼ eij  13 edij , where e ¼ e11 þ e22 þ e33 is the volumetric part of the
strain tensor (Anderson et al., 1994).
It is a well known fact that for isotropic material the stress tensor can be described in
terms of two quantities: the hydrostatic stress (or pressure), which only induces a change
of scale, and the deviatoric stress, which only induces a change of shape (e.g. Malvern,
1969; Wilkins, 1969; Sedov, 1972). This detailed description of such well known facts will
be used later in the construction of a generalised decomposition of the stress tensor which
will take into account physical properties of anisotropic materials.
The pressure is defined as the negative of the three normal stress components. For iso-
tropic materials in the elastic region, the pressure is directly linked with volumetric strain,
and the decomposition of the stress tensor has the conventional form:
d
rij ¼ pdij þ S ij ; p ¼ Ke; S ij ¼ 2lðee Þij ; epkk ¼ 0; ð3Þ
where p is the hydrostatic pressure, Sij is the deviatoric part of the stress tensor, K is the
conventional bulk modulus, and l is the shear modulus. However, for anisotropic mate-
rials (e.g. for an orthotropic material), the decomposition of the stress and strain tensors
into spherical and deviatoric parts in stress space and strain space results in stress and
strain components which do not correspond to each other due to the material properties’
anisotropy. For an anisotropic material the mean stress depends on the deviatoric strains,
therefore, the decomposition used for isotropic materials is not applicable (Anderson
et al., 1994). The conventional expression for the pressure is:
1
p ¼  ðrxx þ ryy þ rzz Þ: ð4Þ
3
In the case of anisotropic material, using the stress-strain relation for an orthotropic mate-
rial Eq. (1), the above equation can be rewritten in the following form (Anderson et al.,
1994):
1
p¼ fEx ð1  myz mzy Þ þ Ey ð1  mxz mzx Þ þ Ez ð1  mxy myx Þ þ 2  ½Ex ðmyx þ mzx myz Þ
9b
1
þ Ex ðmzx þ myx mzy Þ þ Ey ðmzy þ mzx mxy Þg  e  fEx ð1  myz mzy Þ þ Ex ðmyx þ mzx myz Þ
3b
1
þ Ex ðmzx þ myx mzy Þg  edxx  fEx ðmyx þ mzx myz Þ þ Ey ð1  mxz mzx Þ
3b
1
þ Ey ðmzy þ mxy mzx Þg  edyy  fEx ðmzx þ myx mzy Þ þ Ey ðmzy þ mxy mzx Þ þ Ez ð1  mxy myx Þg  edzz ;
3b
ð5Þ
where b ¼ 1  mxy myx  mxz mzx  myz mzy  2myx mzy mxz .
From Eq. (5), it can be seen that in addition to the volumetric strain, the conventional
pressure also depends on the deviatoric strain tensor (Anderson et al., 1994). The defini-
tion of pressure Eq. (4) leads to the invariant quantity (the contraction of the stress tensor
and the unit tensor is divided by the norm of the unit tensor). This is not the case when the
pressure is defined as the portion of the mean stress that varies directly with the volumetric
144 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

strain, which is not an invariant quantity (i.e. this quantity cannot be expressed as contrac-
tion of the stress tensor and another second order tensor). However, the high-pressure
thermodynamic equation of state of anisotropic materials must be modified to account
correctly for the elastic behaviour at small volumetric strain.

2.2. Equation of state (isotropic material)

During shock loading, the material undergoes nonlinear behaviour; therefore an equa-
tion of state (EOS) is required to describe the material’s response under shock loading con-
ditions. Theoretically an equation of state can be determined from the thermodynamic
properties of the material, and ideally should not require dynamic data to construct the
relationship. However, in practice, the only practical way of obtaining data on the behav-
iour of the material at high strain rates is to carry out well-characterised dynamic exper-
iments (e.g. the planar shock wave experiment (Meyers, 1994)). It is convenient for use in
numerical codes to have an analytical form of the EOS. Such an analytic form is at best an
approximation to the true relationship. Further, the equation of state may be given in
extensive tabular form, and in that case the analytic form chosen can be considered as
an interpolation relationship. The EOS for isotropic materials typically defines the pres-
sure as a function of density q (or specific volume, m) and specific internal energy e. A very
popular form of equation of state that is used extensively for isotropic solid continua is the
Mie–Grüneisen EOS:

CðmÞ
p ¼ f ðq; eÞ ¼ P r ðmÞ þ ðe  er ðmÞÞ; ð6Þ
m

where m is the specific volume, CðmÞ is the Grüneisen gamma, and is defined as
 
op
CðmÞ ¼ m ð7Þ
oe m
Traditionally C is taken to be constant C ¼ C0 ; alternatively, it could have been assumed
that Cm00 ¼ Cm ¼ const and Eq. (6). Functions P r ðmÞ and er ðmÞ are assumed to be known func-
tions of v on some reference curve. Possible reference curves include: the shock Hugoniot
curve, a standard adiabatic curve (e.g. the adiabatic through the initial state ðp0 ; m0 Þ), the
0 °K isotherm, the isobar p ¼ 0, the curve e ¼ 0, or some composite curve of one or more
of the above curves to cover the complete range of interest of the parameter m. The most
commonly used form of the Mie–Grüneisen equation of state for solid materials which
uses the shock Hugoniot as the reference curve is given below:
 
C
p ¼ f ðq; eÞ ¼ P H  1  l þ qCe; ð8Þ
2
where PH is the Hugoniot pressure, l ¼ qq0  1 is the relative change of volume, C is the
Grüneisen parameter, q is the density, e is the specific internal energy, and m is the current
specific volume.
The Rankine–Hugoniot equations for the shock jump conditions can be regarded as
defining a relation between any pair of the q; p; e; up variables and U (Meyers, 1994). In
many dynamic experiments, up (the particle velocity directly behind the shock) and U
(the velocity at which the shock wave propagates through the medium) are measured. It
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 145

has been found that for most solids and many liquids over a wide range of pressures there
is an empirical linear relationship between these two variables:
U ¼ c þ Sup ; ð9Þ
where c is the intercept of the U–up curve, and S is the coefficient of the slope of the U–up
curve. Generally, the Hugoniot pressure and a shock velocity U is a non-linear function of
particle velocity up and it is given by the following relation (Steinberg, 1991):
u  u 2
p p
U ¼ c þ S 1 up þ S 2 up þ S 3 up ð10Þ
U U
and Grüneisen’s gamma is given by
c þ al
C¼ 0 : ð11Þ
1þl
Therefore, the Grüneisen equation of state with cubic shock velocity as a function of par-
ticle velocity defines pressure as
  
q0 c2 l 1 þ 1  C2 l  C2 l2
p¼h i2 þ ð1 þ lÞ  C  E; when l > 0; ð12Þ
l2 l3
1  ðS 1  1Þl  S 2 lþ1  S 3 ðlþ1Þ 2

p ¼ q0 c2 l þ ð1 þ lÞ  C  E; when l < 0; ð13Þ


where E is the internal energy per initial specific volume, c is the intercept of the U–up
curve, S 1 ; S 2 ; S 3 are the coefficients of the slope of the U–up curve Eq. (10), c0 is the Grün-
eisen gamma, and a is the first order volume correction to c0. Parameters c; S 1 ; S 2 ; S 3 ; c0 ; a
represent material properties which define its EOS. A similar form of EOS has been devel-
oped and validated at the Lawrence Livermore National Laboratory (Steinberg, 1991).
Parameters have been defined to cover a large number of materials (Steinberg, 1991).

2.3. Equation of state (Anderson et al., 1994)

In this section, the approach to describe an equation of state for orthotropic materials
proposed by Anderson et al. (1994) is discussed. Some attempts to use Anderson’s model
to simulate the behaviour of composite materials under shock loading have been made by
Chen et al. (1997) and Hayhurst et al. (1999).
The first important difference between Anderson’s and the conventional approach is in
the way the Hugoniot pressure is approximated. The conventional approximation of the
Hugoniot pressure PH as cubic least squares curve fit in l is:

A1 l þ A2 l2 þ A3 l3 ; l > 0;
PH ¼ ð14Þ
A1 l; l < 0;
where parameters Ai ; i ¼ 1; 2; 3 are determined by fit to experimental shock compression
(Hugoniot) data. Note that for an isotropic material, A1 is the bulk modulus. Anderson
et al. (1994) proposed that in order to have consistency and correct stresses in the elastic
regime for an orthotropic material, the Hugoniot pressure P AH should be approximated as
(
A01 l þ A2 l2 þ A3 l3 ; l > 0;
PH ¼ ð15Þ
A01 l; l < 0;
146 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

where A01 is defined by the following quantity (Anderson et al., 1994):


1
A01 ¼ ðEx ð1  myz mzy Þ þ Ey ð1  mxz mzx Þ þ Ez ð1  mxy myx Þ þ 2  ½Ex ðmyx þ mzx myz Þ
9b
þ Ex ðmzx þ myx mzy Þ þ Ey ðmzy þ mzx mxy ÞÞ; ð16Þ

where b ¼ 1  mxy myx  mxz mzx  myz mzy  2myx mzy mxz . Anderson et al. (1994) interpreted A01 as an
‘‘effective” or ‘‘average” bulk modulus K ef . The A2 ; A3 parameters are determined from the
fitting of the experimental data. This provides an appropriate description of material
behaviour at high pressures and reduces to the correct relations at small volumetric
strains. The final expression for the EOS for the general orthotropic case, based on
Eqs. (8), (15), (16), and proposed by Anderson et al. (1994), can be written in the following
form:
 
A C
p ¼ P H  1  l þ qCe
2
1
 Ex ð1  myz mzy Þ þ Ex ðmyx þ mzx myz Þ þ Ex ðmzx þ myx mzy Þ  edxx
3b
1
 Ex ðmyx þ mzx myz Þ þ Ey ð1  mxz mzx Þ þ Ey ðmzy þ mxy mzx Þ  edyy
3b
1
 Ex ðmzx þ myx mzy Þ þ Ey ðmzy þ mxy mzx Þ þ Ez ð1  mxy myx Þ  edzz ; ð17Þ
3b
where P AH is given by Eqs. (15) and (16).
Alternatively, A01 ; A2 ; A3 can be analytically determined through a Taylor’s series expan-
sion of the Hugoniot pressure PH. Assuming that the linear approximation between the
shock velocity U and particle velocity up Eq. (9) exists. Therefore, the Hugoniot curve
and Taylor’s series expansion of the Hugoniot pressure PH with respect to l can be written
in the form:
q0  c20  l  ð1 þ lÞ
PH ¼ ; ð18Þ
½1  l  ðS  1Þ2
2
A01 ¼ q0 c20 ; A2 ¼ A01  ½1 þ 2ðS  1Þ; A3 ¼ A01  ½2ðS  1Þ þ 3ðS  1Þ ; U ¼ c0 þ Sup ;
ð19Þ
where c0 is the speed of sound, which has the following definition:
sffiffiffiffiffiffiffi
K ef
c0 ¼ : ð20Þ
q0

It was mentioned above that in the case of an isotropic material, the hydrostatic stress
(or pressure) induces a change of scale, while the deviatoric stress only induces a change of
shape. It is obvious that in order to keep this property for anisotropic materials (e.g. for
orthotropic materials), the definition of ‘‘pressure” needs to be generalised, because a
hydrostatic pressure (isotropic state of stress) applied to an anisotropic material results
in an anisotropic state of strain. In other words, this loading will result not only in a
change of scale, but also in change of shape. This is inconsistent with the definition of
the ‘‘generalised pressure”.
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 147

3. Mathematically consistent model of anisotropic material

The definition of pressure in the case of an anisotropic body should be the result of stat-
ing that the ‘‘pressure” term should only produce a change of scale, i.e. isotropic state of
strain. This should allow the calculation of the direct stress ratios that will produce only a
change of scale. This defines the directions in stress space of the generalised ‘‘pressure”
vector (Lukyanov, 2006).

3.1. Decomposition of stress tensor

A new development of constitutive equations for anisotropic materials (e.g. description


of EOS) must be able to take into account isotropic materials as a special case. When the
isotropic material is considered as a special case of the anisotropic material, the aniso-
tropic equations must have the properties of isotropic equations. The direction of the gen-
eralised ‘‘pressure” vector is defined by the tensor aij such that the trace terms (diagonal
values) define the vector in this direction and the cross-terms of aij are zero. It is reasonable
to assume total ‘‘pressure” Pe as a constant times the tensor aij:
Pe ¼ p aij ; ð21Þ
where p* is the magnitude of total ‘‘pressure”. It will be seen later that total ‘‘pressure” is
defined by the contribution from the spherical and deviatoric parts of the strain tensor. A
similar effect was described by Anderson et al. (1994) in the definition of a constitutive
equation suitable for wave propagation. In this case, unlike in the approach proposed
by Anderson, the generalised ‘‘pressure” maintains all physical properties inherent to clas-
sical hydrostatic pressure for isotopic materials. This methodology can be used in the
problems of material behaviour under shock loading.
Expressing the requirement that the deviatoric stress is independent of the ‘‘pressure”
term, i.e. their contraction product is zero, gives

Pe : e
S ¼ 0; aij e
S ij ¼ 0; rij ¼ p aij þ e
S ij : ð22Þ

In this paper, everywhere contraction by repeating indexes is assumed. Then, following the
classical formulation, e
S ij is defined as the stress minus the ‘‘pressure”:
e
S ij ¼ rij þ p aij : ð23Þ
Using Eqs. (22) and (23), the following expression for ‘‘pressure” can be obtained:
rij aij 1
p ¼  ¼ rij aij ; ð24Þ
akl akl kak
where kak ¼ aij aij , and finally, the expression for the generalised deviatoric part of the
stress tensor can be rewritten in the following form:

e 1
S ij ¼ rij  aij  rkl akl : ð25Þ
kak
The methodology for calculation of components of the tensor aij has been defined (Lukya-
nov, 2006). Therefore, elements of tensor aij and parameter KC can be written in the fol-
lowing form:
148 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

aij ¼ C lk dkk dil dlj  3K C ; ð26Þ


rffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
h i
2 2 2 1
K C ¼ 1= 3  ðC 11 þ C 12 þ C 13 Þ þ ðC 12 þ C 22 þ C 23 Þ þ ðC 13 þ C 23 þ C 33 Þ ; K C ¼ ;
9K C
ð27Þ
aij aij ¼ kaij k ¼ 3: ð28Þ
A new definition of generalised ‘‘pressure” for anisotropic material must revert to conven-
tional pressure based on mean stress when applied to an isotropic material. The relation
shown in Eq. (28) describes the norm of the tensor aij which reduces to Kroncker delta
in the limit of isotropy, therefore the norm of aij is taken to be 3. The parameters aij
and KC describe fundamental properties of anisotropic materials.
In order to find a link between total generalised ‘‘pressure” p* and its components, the
generalised Hook’s law is considered. Substitution of the decomposed strain tensor (spher-
ical and deviatoric part) into the generalised Hook’s law gives:
 
1 e d 1 d
rij ¼ C ijkl dkl e þ ðe Þkl ¼ C ijkl dkl e þ C ijkl ðee Þkl
3 3
d
¼ K C e  aij þ C ijkl ðee Þkl ; epkk ¼ 0; ð29Þ
where Cijkl is the stiffness matrix, e is the volumetric deformation. The vector of pressure is
defined by the tensor
1
pij ¼  eC lk dkk dil dlj ¼ K C  e  aij ¼ p  aij ; p ¼ K C  e; ð30Þ
3
where p is a component of the total generalised ‘‘pressure” corresponding to the spherical
part of the strain tensor. Now, by multiplying Eq. (30) by the second order tensor bij (the
physical nature of which will be clarified later) the following scalar relation is obtained:
bij rij ¼ bij aij  K C e þ bij C ijkl ðee Þdkl ¼ bij aij  p þ bij C ijkl ðee Þdkl : ð31Þ
Assuming that
d
bij C ijkl ðee Þkl ¼ 0 ð32Þ
for any value of edkl . Substituting Eq. (32) into Eq. (31) yields:
bij rij
p¼ : ð33Þ
bij aij
The p parameter is part of the total generalised ‘‘pressure” p* and is directly linked to the
spherical part of the strain tensor. Besides, the value p is an invariant quantity. This solves
the problem (described by Anderson et al., 1994) when the pressure defined as the portion
of the mean stress that varies directly with the volumetric strain is not an invariant quan-
tity. In the problem of material behaviour under shock loading, the parameter p should be
substituted by the equation of state. Some attempts to develop the constitutive relation-
ships for an anisotropic material under shock loading have been suggested by Anderson
et al. (1994) based on the conventional definition of hydrostatic pressure. The methodol-
ogy developed in this paper provides a consistent procedure to interpret the high-pressure
material response (equation of state), so that the summation of the generalised deviatoric
stresses and the generalised ‘‘pressure” provides the correct stresses in the elastic regime.
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 149

The following relation describes the definition of the second order material tensor bij
(Lukyanov, 2006):

bij C ijkl ¼ 3K S dkl ; ð34Þ

where Ks represents the second generalised bulk modulus. Therefore Eq. (33) will take the
form:

bij d
p¼ ½aij K C  e þ C ijpr ðee Þpr  ¼ K C  e: ð35Þ
bkl akl

A linear system of equations (with respect to components of the tensor bij) represented by
relation Eq. (34), can be solved and the components of the tensor bij can be expressed in
terms of the compliance matrix (see Eq. (34)). Similar to tensor aij, the norm of the tensor
bij is assumed to be equal to 3, i.e. bij bij ¼ kbij k ¼ 3. Therefore, the following set of equa-
tions for the components of the tensor bij can be written as:
bij ¼ J lk dkk dil dlj  3K S ; ð36Þ
qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
2 2 2
K S ¼ 1= 3  ½ðJ 11 þ J 12 þ J 13 Þ þ ðJ 12 þ J 22 þ J 23 Þ þ ðJ 13 þ J 23 þ J 33 Þ ; ð37Þ
bij bij ¼ 3; ð38Þ
where Jij are elements of the compliance matrix (written in Voigt notation).
After defining the parameter p in terms of two second-order tensors aij ; bij and the stress
tensor, we can finally obtain the link between total generalised ‘‘pressure” p* and pressure
p:
bij aij  p ¼ bij rij ¼ bij  ðaij  p þ e
S ij Þ ð39Þ
and, finally,
bij e
S ij
p ¼ p þ : ð40Þ
aij bij
The methodology of decomposition of the stress tensor into the generalised ‘‘pressure”
and generalised deviatoric part of the stress tensor is fully defined. Using two fundamental
tensors aij and bij, the definitions of total ‘‘pressure” and pressure corresponding to the
volumetric deformation can be determined. A new definition of generalised ‘‘pressure”
for anisotropic material must be able to take into account the isotropic material as a spe-
cial case. In the limit of isotropy, tensors aij and bij have the following values
a11 ¼ a22 ¼ a33 ¼ 1; b11 ¼ b22 ¼ b33 ¼ 1 and the proposed generalisation reverts to the tra-
ditional classical case, where tensors aij ; bij equal dij and Eqs. (24) and (33) take the follow-
ing form:

rij dij 1 bij rij 1


p ¼  ¼  rkk ; p¼ ¼  rkk ; p ¼ p ð41Þ
dkl dkl 3 bij aij 3

which is the classical hydrostatic pressure. Besides, two parameters K C ; K S were considered
as the first and the second generalised bulk modulus. In the limit of isotropy they reduce to
the well-known expression for the conventional bulk modulus (Lukyanov, 2006).
150 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

3.2. Equation of state for anisotropic material

For anisotropic materials, the total hydrostatic ‘‘pressure” has been defined as
bij e
S ij
p ¼ p þ ; ð42Þ
aij bij
where p is the pressure related to the volumetric deformation, and e S ij is the generalised
deviatoric part of the stress tensor. Eq. (42) is the correct generalised ‘‘pressure” for the
elastic regime. To provide an appropriate description of behaviour for general anisotropic
materials at high pressures, an equation of state for p (pressure related to the volumetric
deformation) has to be defined:
8 q c2 l 1þ 1C lCl2
> 0 ½ ð 2Þ 2 
<h i2 þ ð1 þ lÞ  C  E; l > 0;
EOS l2 3
p ¼ 1ðS 1 1ÞlS 2 lþ1 S 3 l 2 ð43Þ
>
:
ðlþ1Þ

q0 c2 l þ ð1 þ lÞ  C  E; l < 0:
Therefore, an appropriate description of general hydrostatic ‘‘pressure” at high pressures
has the following form:
bij e
S ij
p ¼ pEOS þ ð44Þ
aij bij
which also describes correctly the material’s behaviour at small volumetric strains. To be
consistent with the definition of the speed of sound c, the following definition
sffiffiffiffiffiffiffi
KC
c¼ ¼ c0 ð45Þ
q0

is assumed, where the first generalised bulk modulus KC is defined according to Eq. (27).
Besides, in case of a linear relation between the shock and the particle velocity, the Hugon-
iot pressure PH reduces to Eq. (18) and expanding in Taylor’s series leads to:
2
A01 ¼ q0 c20 ; A2 ¼ A01  ½1 þ 2ðS  1Þ; A3 ¼ A01  ½2ðS  1Þ þ 3ðS  1Þ ; ð46Þ
where c0 is the speed of sound, which is defined by Eq. (45). Note that the methodology
described above can be applied for all anisotropic materials and represents a mathemati-
cally consistent generalisation of the conventional isotropic case.

3.3. Constitutive relations

For conventional isotropic materials, the constitutive relation or strength model defines
the deviatoric response of the material as a function of any combination of the deviatoric
strain tensor ~e, deviatoric strain rate tensor ~
d, temperature T, pressure P (for some models
e.g. geo-materials etc.), and, in general, the damage tensor x:

S ij ¼ hf iðeij ; dij ; T ; P ; xÞ; ð47Þ

where Sij is the deviatoric part of the stress tensor, hf i is the constitutive operator which
represents the link between Sij and other state variables. The operator hf i can have a com-
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 151

plex form and cannot be represented analytically. In the case of anisotropic materials the
following models can be applied.

3.4. Full decomposition of the stress tensor

In the case of full decomposition of the stress tensor, we assume that the stress tensor is
defined as:
rij aij bij e
S ij
rij ¼ p aij þ e
S ij ; p ¼  ; p ¼ p þ ; ð48Þ
akl akl aij bij
where p is the pressure related to the volumetric deformation and is described by Eq. (43),
and eS ij is the generalised deviatoric part of the stress tensor. For anisotropic materials, the
following set of assumptions for the anisotropic plasticity can be written (Lukyanov, 2006):
1: Fb ð e
S ij Þ ¼ Y ;
2: De ¼D ee þ De p;
b
oG ð49Þ
3: Dpij ¼ k_ ; Dpkk ¼ 0;
e ij
oS
4: aij  Dpij ¼ 0;
where Fb ð e b e
S ij Þ is the yield criterion, Y is the yield strength, and Gð S ij Þ is the plastic poten-
e
tial. Here, it is traditionally assumed that the strain rate D (the symmetric part of the
velocity gradient) is additively decomposed into the elastic strain rate D e e and the plastic
e p
strain rate D (plastic flow is incompressible), where
Se  ¼ hF iðC; De ; aij Þ; ð50Þ
ij

where hF i is the constitutive operator which represents the link between Se ij and other state
variables, the fourth-order tensor C is the elastic modulus (stiffness matrix), and S e  de-
notes the co-rotational stress rate (Bammann and Aifantis, 1987)
Se ¼ S e_  We S e þ SW
e e;
where We is the elastic spin tensor (Dafalias, 1985).
b Two pos-
In this study, we consider the associative flow plasticity model (i.e. Fb ¼ G).
sibilities are considered in this paper: Hill’s criteria is considered independent from the
generalised hydrostatic ‘‘pressure” but without property 4 of Eq. (49) (modified Hill crite-
ria (Hill-M)) and thermodynamically consistent anisotropic plasticity model which is
based on the all assumptions Eq. (49) (Hill-TCM). In the following sections, these two
models are described in details.

3.4.1. Modified hill criteria (Hill-M)


Modified Hill criteria (Hill-M) is defined by the following assumptions:
1: F ð e
S yy  e
S zz Þ þ Gð eS zz  e
S xx Þ þ H ð e
S xx  e
S yy Þ þ 2N e
S 2xy þ 2L e
S 2yz þ 2M e b e
2 2 2
S 2xz  Y 2 ¼ Gð S ij Þ;
e ¼D
2: D e þD
e e ; D ¼ 0;
p p
kk
o b
G
3: Dpij ¼ k_ ; Dpkk ¼ 0;
oe
S ij
ð51Þ
152 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

where Gðb eS ij Þ is the yield criterion and plastic potential; Y is the yield strength. The
assumption 1 of Eq. (51) describes the yield surface independent from the generalised pres-
sure which corresponds to the fact that volumetric deformation is elastic. In assumption 2
e is additively decomposed into
of Eq. (51), it is traditionally assumed that the strain rate D
e e e p
the elastic strain rate D and the plastic strain rate D (plastic flow is incompressible),
where F ; G; H ; N ; L and M are material constants which are fitted from experimental data.

3.4.2. Thermodynamically consistent anisotropic plasticity (Hill-TCM)


Thermodynamically consistent anisotropic plasticity model which based on the all
assumptions Eq. (49) (Hill-TCM) is defined by the following expressions:
1: Fb ð e
S yy  e b e
S zz Þ2 þ Gð S zz  e b ðe
S xx Þ2 þ H S xx  e be
S yy Þ2 þ 2 N Le
S 2xy þ 2 b be
S 2yz þ 2 M b e
S 2xz  Yb 2 ¼ Wð S ij Þ;
e e
2: D ¼ D þ D ;e e p

b
oW
3: Dpij ¼ k_ ; Dpkk ¼ 0;
oeS ij
4: aij  Dpij ¼ 0;
ð52Þ
where Wð b eS ij Þ is the yield criterion and plastic potential; Yb is the yield strength. Here, the
yield surface Wð b e b e
S ij Þ is different from the yield surface Gð S ij Þ of the Hill-M plasticity model
Eq. (51), because of difference of material parameters Fb ; G; b H b;N b; b b from the parame-
L; M
ters F ; G; H ; N ; L; M of the Hill-M plasticity model Eq. (51) which are fitted from experi-
mental data. Based on the assumption 4 of Eq. (52), the material parameters Fb ; G; b Hb are
defined analytically as follows (Lukyanov, 2006):
Fb ¼ A  F ðaij Þ; G b ¼ A  Gðaij Þ; H
b ¼ A  H ðaij Þ; ð53Þ
and F ; G; H purely depend on the material tensor aij:
½a22  a11  ½a11  a22 
F ðaij Þ ¼  ; Gðaij Þ ¼  ; H ðaij Þ ¼ 1: ð54Þ
½a22  a33  ½a11  a33 
Eq. (54) is obtained for the case when a11 6¼ a22 6¼ a33 ; all other special cases are addressed
by the fundamental relation Eq. (55) :
F  0 þ Gða11  a33 Þ þ H ða11  a22 Þ ¼ 0;
F ða22  a33 Þ þ G  0 þ H ða22  a11 Þ ¼ 0; ð55Þ
F ða33  a22 Þ þ Gða33  a11 Þ þ H  0 ¼ 0;
which is true in any case, including cases such as a11 ¼ a33 6¼ a22 , the solution obtained
from the Eq. (55) for other cases will be different to Eq. (54). Besides, in the case of isot-
ropy and the special case of anisotropy when aij  dij , it is reasonable to assume that
b ¼H
Fb ¼ G b ¼ A . Consequently, we can see that parameters A; N b; b b are material con-
L; M
stants which must be fitted from experimental data.

4. Numerical simulation

4.1. Description of experiment

The work discussed below concerns the shock response of anisotropic aluminium allow
aluminium alloy 7010-T6. This is done by the technique of plate impact, whereby a flat
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 153

plate of constant thickness and a known material (aluminium alloy 6082-T6 – dural) is
impacted onto a target plate made from the test material (AA 7010-T6). The flyer plates
are launched using a 75 mm bore, 1 mm long and 50 mm bore, and 5 m long single stage
gas gun the Royal Military College of Science (Bourne and Stevens, 2001; Vignjevic et al.,
2002). Thick plates of 7010-T6 (50 mm  50 mm  5 mm size) were cut from a single heat-
treated block, and ground flat and parallel to within 5 optical fringes over 50 mm. The hot
working process on this material gave a characteristic pancake grain structure, with the
long axis of the grains in the rolling direction. On impact, a planar shock front starts prop-
agating into the target. The shock propagation in the target is monitored using manganin
stress gauges, placed between target plate and 12 mm PMMA (poly methylmethacrylate)
plate within the target assembly. The target assembly is shown in Fig. 1. More detailed
description of the test can be found in paper Vignjevic et al. (2002).
The 2.5 mm thick flyer plates of 6082-T6 were impacted onto the targets over the veloc-
ity range 234 to 895 m/s, inducing stresses in the range 2.7–7.2 GPa. The aluminium alloy
6082-T6 was chosen as the flyer due to the close similarity in density and wave speeds, so
that the impact experiments were near symmetrical (Vignjevic et al., 2002). The elastic
material properties of 7010-T6 were taken from the paper Vignjevic et al. (2002) and pre-
sented in Table 1. Material properties of plates 6082-T6 and PMMA are presented in
Tables 2 and 3 (Steinberg, 1991).

4.2. Description of numerical experiment

The plane shock-wave technique provides a powerful tool for studying material prop-
erties at different strain rates (e.g. Steinberg, 1991; Meyers, 1994; Johnson et al., 1994;
Kanel et al., 1997; and Millett et al., 2002). Such characteristics as spall pressure, shock
velocity, particle velocity, Hugoniot elastic limit (HEL), thickness of the spall section, time
to spall, and free-surface velocity of the spall section can be measured and used for the
characterization of material dynamic response (e.g. Meyers, 1994; Kanel et al., 1997;
Kanel, 1998; Millett et al., 2002; Bourne and Gray, 2003; and Stoffel, 2005). In the earliest

Fig. 1. Schematic diagram of the experimental target assembly.


154 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

Table 1
Material properties of AA 7010-T6
Quantity Value
q0 (kg/m3) 2810
Ea (GPa) 70.6
Eb (GPa) 71.1
Ec (GPa) 70.6
mab 0.342
mca 0.342
mbc 0.342
Gab (GPa) 26.31
Gca (GPa) 26.48
Gbc (GPa) 26.48

Table 2
Material properties
Plate q0 (kg/m3) l0 (GPa) Y0 (MPa) Ep (MPa)
Flyer Al 2024-T4 2785 28.6 260 700
Flyer Al 6082-T6 2700 26.8 250 130
Target Cu OFHC 1/2 Hard 8930 47.7 120 0.0
PMMA 1180 2.68 200 300

Table 3
Mie–Grüneisen EOS
Plate c (m/s) S1 (–) S2 (–) S3 (–) c0 (–) a (–)
Flyer Al 2024-T4 5328 1.338 0 0 2.0 0.48
Flyer Al 6082-T6 5240 1.4 0 0 1.97 0.48
Target Cu OFHC 1/2 Hard 3940 1.489 0 0 2.02 0.47
PMMA 2600 1.52 0 0 1.0 0.0

plane wave experiments two parameters that could be determined were the Hugoniot elas-
tic limit (HEL) (or stress level associated with the elastic precursor wave) and the dynamic
compressibility (or bulk modulus) associated with the following plastic wave. More com-
plex theories of material behaviour have led to the extraction of much additional informa-
tion from such experiments definition of non-standard parameters (Kiselev and
Lukyanov, 2002). From an experimental point of view, it is clear that the production of
waves in a metal target and the measurement of their characteristics, such as speed and
intensity, provide one of the most convenient methods of investigating the physical prop-
erties of a material under high pressure. Besides, the theoretical prediction of wave prop-
agation in certain specified circumstances can provide simple basic test problems to assess
a suggested theoretical model by comparing its predictions with experimental data (e.g.
Meyers, 1994; Kanel et al., 1995; Kanel et al., 1997; Kiselev and Lukyanov, 2002). There-
fore, in this work, a simulation of the plate impact test is considered.
To describe the dynamic response of different materials under shock loading, the meth-
odology based for the plate impact test has been developed and successfully used to val-
idate the Mie–Grüneisen equation of state for isotropic materials by measuring shock
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 155

velocity U and particle velocity up (Steinberg, 1991; Meyers, 1994). Using U  up experi-
mental data, further approximation of this experimental curve using Eq. (9) or Eq. (10)
can be done. The same methodology can be used to validate the modified Mie–Grüneisen
equation of state for anisotropic materials; thus a numerical simulation of the anisotropic
plate impact test is considered.
A schematic of the plate impact test is depicted in Fig. 1, in which a flyer plate impacts a
target plate which is bonded to a PMMA plate. In this paper, the case is considered where
the diameters of the flyer and the target are much greater than their thicknesses and the
characteristic time of the process is the time of several runs of elastic waves across the
thickness of the target plate. In such a case, the problem may be solved using a uniaxial
strain state (one-dimensional mathematical formulation in strain space) and the adiabatic
approximation; therefore, planar impact generates two one-dimensional shock waves. One
propagates into the target and the other into the flyer plate. These shock waves reflect as
rarefaction waves from the free surfaces of the flyer and from the back of target plates con-
nected to the PMMA. With a thin flyer, these rarefaction waves interact inside the target,
producing a state of tension in some regions which leads to the spallation. A large set of
spall experiments have been carried out by Kanel et al. (1996a,b). In this work, the
dynamic anisotropic elasto-plastic material behaviour before spallation is considered.
Plate-impact numerical simulations were performed with a 5 mm thick target plate,
2.5 mm thick flyer plate and 5 mm PMMA plate. Based on the characteristics of this plate
impact problem, the plates (numerical domains), which are used in the numerical simula-
tion, are modeled as rectangular bars (shown in Fig. 2) of 5  5 elements in the z and y
directions (final geometries of the bars: flyer 2.5 mm  0.4 mm  0.4 mm, target  5.0
mm  0.4 mm  0.4 mm, PMMA  5.0 mm  0.4 mm  0.4 mm).
Symmetry planes were applied on all sides, and as a result, a one-dimensional wave will
travel along the length of the bar. A non-reflective boundary condition is applied at the back
of the PMMA-block (Meyers, 1994). The flyer is modeled with 30 elements along the axis of
impact, the test specimen is modeled with 110 elements along the axis of impact, and the
PMMA block is modeled with 75 elements. With this mesh resolution, all the characteristic
features of the stress pulse can be resolved. An increase in mesh resolution does not affect the
results. A surface-to-surface contact interface was specified between the flyer and the test
specimen (target). The stress time histories were recorded in the middle of the target plate
and at the back of the test specimen (the first element in the PMMA connected to test plate).
These mesh resolutions were sufficient to allow the resolution of all the relevant elastic

Fig. 2. Schematic diagram of the numerical target assembly. Numerical domain (FE Model).
156 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

and plastic waves in the target and flyer. The three impacts were simulated with velocities
200 m/s, 504 m/s, and 700 m/s.
The process of verification of the model and simulation of shock wave propagation in
anisotropic aluminum alloys is presented in the following sections.

4.3. Verification of the model

The process of verification is intended to provide, and quantify, confidence in numerical


modelling and the results from the corresponding simulations. Note that in the limit of
isotropy, the aforementioned models reduce to the conventional constitutive equations
suitable for the shock wave propagation in isotropic materials described by Steinberg
(1991) and Meyers (1994). Therefore, in order to be confident that the implementation
of the constitutive equation into the FE code has been done correctly, a special case of
anisotropy (i.e. isotropy of the materials) is considered. Plate impact tests of isotropic alu-
minum are considered; the data for the aluminum was taken from the report issued by
Lawrence Livermore National Laboratory (Steinberg, 1991).
The material properties (density q0, shear modulus l0, yield strength Y0, isotropic hard-
ening parameter) of the flyer (aluminum (Al 2024-T4)) and the target (copper (Cu OFHC
1/2 Hard)) are presented in Table 2.
The target parameters have been used in the proposed anisotropic elasto-plasticity
model (without kinematic or isotropic hardening), in order to represent an isotropic mate-
rial, which is a special case of anisotropy. Parameters for the Mie–Grüneisen equation of
state (Eqs. (11) and (43)) for the flyer Al 2024-T4 and
8 q c2 l 1þ 1C lCl2
> 0 ½ ð 2Þ 2 
<h i2 þ ð1 þ lÞ  C  E; l > 0
EOS l2 3
p ¼ 1ðS 1 1ÞlS 2 lþ1 S 3 l 2 ð56Þ
>
:
ðlþ1Þ

q0 c2 l þ ð1 þ lÞ  C  E; l < 0
for the target Cu OFHC 1/2 Hard are presented in Table 3. The back plate PMMA has
been modelled as an isotropic material with the properties presented in Tables 2 and 3.
A numerical simulation was performed using LS-DYNA transient dynamic non-
linear finite element analysis software (Livermore Software Technology Corporation.
LS-DYNA, 1998). In Fig. 3, the comparison between LS-DYNA results and the imple-
mentation of the new constitutive equations is presented. It shows the time history of lon-
gitudinal stress at the element which lies halfway through the target plate (Cu OFHC 1/2
Hard). The main characteristics, such as Hugoniot elastic limit (HEL) rHEL and maximum
of longitudinal deviatoric stress jS xx j obtained
 1m  from the numerical simulation, are equal to
the theoretical values rTheoryHEL ¼ 12m
Y 0 ¼ 253:8 MPa and max jS xx j ¼ 23 Y 0 ¼ 80 MPa
respectively. Good agreement can be observed in the stress level, thickness of the stress sig-
nal, and the shape of the stress signal. A slight difference in unloading path between LS-
DYNA results and the proposed model is explained by the numerical aspects (accumula-
tion of numerical error) of the new implementation of plasticity models Hill-M and Hill-
TCM. In both cases of plasticity Hill-M and Hill-TCM, the iterative correction algorithm
of the trial generalized deviatoric part of the stress tensor was used. The iterative algorithm
is performed till the convergence condition jdkiþ1  dki j < jdki j  error is realized, where
dkiþ1 ; dki are scale factors in Eq. (51) or Eq. (52) at i þ 1 and i iterative steps respectively,
error is numerical error of iterative algorithm. At the same time, the stresses are scaled
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 157

Fig. 3. Stress history at the middle of the target plate.

back to the yield surface in LS-DYNA using an analytical expression (see, e.g., LS-DYNA
theoretical manual, 1998). Therefore, the numerical results of the new models show
slightly later unloading path compared with LS-DYNA results. This is due to accumulated
numerical errors of iterative algorithm. The decreasing of the error parameter will lead to a
better comparison and an increasing of the simulation time.
Besides, the experimentally defined expression U ¼ c þ Sup allows us to estimate the
stress wave propagation velocity and the stress amplitude. The stress amplitude (for the
wave propagation problem, positive stress values indicate compression, the usual notation
in shock physics) in the direction of the wave propagation can be estimated by the follow-
ing analytical expression:
2
rxx ¼ P þ S xx ¼ q0 Uup þ Y 0 : ð57Þ
3
Numerically obtained values of rxx = 2.234 GPa, 5.733 GPa, 8.145 GPa, correspond
to the analytical
 values rxx = 2.239 GPa,  5.735 GPa,
 8.152 GPa for particle mvelocities

m m m
up ¼ 60:1
 s
impact velocity V ¼ 200:0
 s
, 152:8 s
impact velocity V ¼ 504:0 s and
213:9 ms impact velocity V ¼ 700:0 ms respectively. Therefore, we conclude that the imple-
mentation was performed correctly: a simplification of the model to the conventional con-
stitutive equations in the limit of isotropy can be achieved numerically.
In the plate impact test, stress gauges to measure the stress history are bonded to the
back of the target with the block of PMMA (Meyers, 1994; Millett and Bourne, 2001;
Bourne and Gray, 2003); therefore, the stress history at the first axial element of PMMA
can be used to compare the numerical simulation with experimental data directly. Fig. 4
shows the inverted stress history at the first axial element of PMMA resulting from the
158 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

numerical simulation using LS-DYNA versus the finite element simulation with the new
material model.
Similar to the previous stress history (i.e. Fig. 3), good agreement can be observed in the
stress level, the thickness of the stress signal, and the shape of the stress signal in the
PMMA.

4.4. Simulation of shock wave propagation in anisotropic materials

In this section the shock wave propagation in an anisotropic material is considered.


First, it is important to investigate only the contribution of the anisotropic equation of
state; hence the first series of computer simulations are related to shock wave propagation
in elastic anisotropic materials.

4.4.1. Shock wave propagation in anisotropic elastic materials


Numerical computations were made to simulate an anisotropic plate that impacts a
smooth rigid surface with impact velocity V ¼ 5:0 ms . At this velocity, the behaviour of
the selected anisotropic material can be considered as elastic (local deformations are elas-
tic), and therefore, using anisotropic elasticity with the anisotropic equation of state is rea-
sonable. The next two series of computer runs explored the effect of anisotropy. The
following compliance matrix parameters were considered (Anderson et al., 1994). For
the first example (Case 1) the following parameters are considered: Ex ¼ 2  E0 ; Ey ¼
Ez ¼ E0 ; mxy ¼ myz ¼ 0:33, and mzx ¼ 0:165; Gyx ¼ 1:5G0 ; Gzx ¼ 1:5G0 ; Gzy ¼ G0 . For the
second example (Case 2): Ex ¼ 3  E0 ; Ey ¼ Ez ¼ E0 ; mxy ¼ myz ¼ 0:33, and mzx ¼ 0:11;

Fig. 4. Stress signal at PMMA–target plate interface (inverted real value).


A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 159

Gyx ¼ 2G0 ; Gzx ¼ 2G0 ; Gzy ¼ G0 , where parameters E0 ; G0 are the reference material proper-
ties presented in Table 4.
The stress amplitude (in shock physics positive stress values indicate compression) in
the direction of wave propagation for a low particle velocity and a small deformation
can be approximated by the following analytical expression:
rxx ¼ q0 cup : ð58Þ
For anisotropic elastic materials (described by Hook’s law), the wave speed in the x-direc-
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
tion is given by c ¼ C 11 =q0 , where the C11 is the first element in the stiffness matrix. In
the case of shock wave propagation, the parameters presented in Table 5 were used to de-
scribe the anisotropic equation of state (EOS).
The results of the numerical simulation for the first and second cases plotted in Fig. 5
explore the effects of anisotropy in case 1 and case 2 presented above. The stress history rxx
of numerical simulation at two different locations (0.25 cm and 0.5 cm from the rigid wall
(RW)) is shown in Fig. 5.
The numerical values obtained during the simulation rxx ¼97.2 MPa, U ¼ 7123 ms (Case
1) and rxx ¼106 MPa, U ¼ 8470 ms (Case 2) correspond to the analytical (acoustic) approx-
imation rxx ¼109.1 MPa, U ¼ 7850 ms (Case 1) and rxx ¼129 MPa, U ¼ 9316 ms (Case 2).
The maximum difference between the numerical results and the analytical (acoustic)
approximation is 10%. Note that the analytical approximation does not represent the real
behaviour, but it can be used to understand the physics of shock wave propagation. From
the numerical simulation it is clear that the behaviour of the stress waves correspond to the
analytical solution. The error between the numerical simulation and analytical results
looks reasonable and corresponds to the simplifications which have been done to obtain
the analytical results.

4.4.2. Shock wave propagation in 7010-T6 anisotropic aluminum alloy


In this section the shock wave propagation in the anisotropic aluminium alloy 7010-T6
is considered. The two identification methods (Hill-M) and (Hill-TCM) were used for cal-
ibration of two possibilities: (1) Hill (1948,1950) criteria are considered independent from
the generalised hydrostatic ‘‘pressure” (Hill-M) (see Lukyanov, 2006); (2) a thermodynam-
ically consistent anisotropic plasticity model (Hill-TCM) is based on all assumptions Eq.
(49) presented by Lukyanov (2006). The parameters for (Hill-M) and (Hill-TCM) yield
surfaces are presented in Table 6. Basic quasistatic mechanical properties for 7010-T6

Table 4
Reference material parameters
q0 (kg/m3) E0 (GPa) G0 (GPa) m (–) Y0 (MPa)
2780 71.71 27.58 0.33 324.0

Table 5
Parameters for the EOS – anisotropic material
Case (No.) c (m/s) S (–) c0 (–) a (–)
1 4928 0.0 0.0 0.0
2 5210 0.0 0.0 0.0
160 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

Fig. 5. Stress waves for 1-D wave propagation (numerical simulation) at two different locations (0.25 cm and
0.5 cm from the rigid wall (RW)) – anisotropic material case 1 ðEx ¼ 2  E0 ; Ey ¼ Ez ¼ E0 ; mxy ¼ myz
¼ 0:33; mzx ¼ 0:165; Gyx ¼ 1:5G0 ; Gzx ¼ 1:5G0 ; Gzy ¼ G0 Þ and case 2 ðEx ¼ 3  E0 ; Ey ¼ Ez ¼ E0 ; mxy ¼ myz ¼
0:33; mzx ¼ 0:11; Gyx ¼ 2G0 ; Gzx ¼ 2G0 ; Gzy ¼ G0 Þ.

are presented in Tables 1 and 6, therefore they can act as a reference for the following
shock experiments. It can be seen that the longitudinal orientation has higher strengths
and ductility than the short transverse. The elastic level of the anisotropy can be defined
by using components of the fundamental tensor aij. In the case of isotropy and the special
case of anisotropy (weak anisotropy) we have aij  dij , where dij is the Kronecker delta
symbol. For the aluminium alloy 7010-T6, components of the tensor aij have the following
values: a11 0:9976; a22 1:0029; a33 0:9994. Based on the values of the aij tensor, we
can conclude that aluminium alloy 7010-T6 is a general anisotropic material
a11 6¼ a22 6¼ a33 and its level of anisotropy (difference from the isotropic or weak aniso-
tropic cases) is about maxðja11  d11 j; ja22  d22 j; ja33  d33 jÞ  100% 0:29%. From
another side, the experimental values, 0.39 GPa and 0.33 GPa for elastic response from
the longitudinal and short transverse directions have been reported by Millett and Bourne

Table 6
Parameters of yield criteria AA7010-T6
Yield criterion Material constants
Hill-M F ¼ 0:6898, G ¼ 0:2884, H ¼ 0:6821, Y ¼ 500:0 MPa
Hill-TCM F ¼ 0:7694, G ¼ 1:4843, H ¼ 0:5067, Y ¼ 500:0 MPa
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 161

Table 7
Anisotropic Mie–Grüneisen EOS
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Plate c ¼ K C =q0 ¼ c0 (m/s) KC (GPa) S1 (–) S2 (–) S3 (–) c0 (–) a (–)
Target AA 7010-T6 5154 74.65 1.4 0 0 2.0 0.48

and published by Vignjevic et al. (2002). Therefore, the relative difference in HEL for dif-
ferent directions is about 15.4%.
Plate impact experiments on aluminum alloy 7010-T6 have been performed to deter-
mine the effect of orientation of the loading axis to the longitudinal and transverse direc-
tions for the shock induced mechanical properties reported by Millett and Bourne and
published by Vignjevic et al. (2002). In this experiment, the flyer plate of 6082-T6 was
impacted onto the target plate of 7010-T6 over the velocity range 234–895 m/s. Material
parameters and parameters for the Mie–Grüneisen equation of state Eq. (56) of the flyer
Al 6082-T6 are presented in Tables 2 and 3 (Steinberg, 1991).
The set of parameters for the constitutive equation Eq. (50) and the EOS (equation of
state) Eqs. (43), (44) and (49) for anisotropic aluminum alloy 7010-T6 is presented in
Tables 1 and 7. The plasticity model is described by the set of parameters presented in
Table 6.
Some material data (e.g. S 1 ; S 2 ; S 3 ; c0 ; a) has been obtained from tabulated data for iso-
tropic aluminum alloys (Steinberg, 1991); hence they cannot be considered as final mate-
rial properties. This data can be defined from the reference plate impact test in different
directions (for orthotropic material: longitudinal and transverse directions respectively).
The longitudinal and transverse stress histories have been recorded at the first axial ele-
ment of PMMA. The experimental data for AA7010-T6 presented here correspond to the
plate impact test (Vignjevic et al., 2002) with impact velocities of 450 m/s and 895 m/s.
Figs. 6 and 7 show the comparison between experimental data and the numerical simula-
tion resulting from the new system of constitutive equations for the longitudinal and trans-
verse cases.
The experimental values, 0.39 GPa and 0.33 GPa for elastic response from the longitu-
dinal and short transverse directions, respectively Figs. 6 and 7, are in good correlation
with the modeled values of the HEL longitudinal – 0.424 GPa (for Hill-M model) and
short transverse – 0.333 GPa (for Hill-M model). The errors with respect to the experimen-
tal values are 1.4% and 0.9%, respectively to the longitudinal and short transverse direc-
tions. Besides, another important characteristic, the arrival time to the HEL and the
plastic wave velocity in the case of Hill-M plasticity model are in good correlation with
experimental data. It has been shown (Lukyanov, 2006) that the thermodynamically con-
sistent plasticity model (Hill-TCM) cannot adequately predict the yield surface; however
this model accurately describes the evolution of the plastic strain tensor. Further compar-
ison shows that the pulse width and the reloading trace are in good agreement with the
experimental data (see Figs. 6 and 7). The main conclusion obtained from these results
is that the model, as it stands, is suitable for simulating this type of wave propagation
in anisotropic solids. Different HELs are obtained when the material is impacted in differ-
ent directions; their excellent agreement with the experiment demonstrates that the aniso-
tropic plasticity model is adequate. Furthermore, the good agreement of the general pulse
shape and Hugoniot stress levels suggests that the EOS is performing satisfactorily. How-
ever, further work is required to combine the non-associative plasticity model to be able to
162 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

Fig. 6. Back-surface gauge stress traces from plate-impact experiments versus numerical simulation of stress
(PMMA) waves for plate impact test (impact velocity 450 m/s) – aluminum alloy AA7010-T6 (LD – longitudinal
direction, TD – transverse direction) with two different plasticity models Hill-M and Hill-TCM.
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 163

Fig. 7. Back-surface gauge stress traces from plate-impact experiments versus numerical simulation. Stress
(PMMA) waves for plate impact test (impact velocity 895 m/s) – aluminum alloy AA7010-T6 (LD – longitudinal
direction, TD – transverse direction) with two different plasticity models Hill-M and Hill-TCM.
164 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

accurately describe the yield surface and a thermodynamically consistent evolution of the
plastic strain tensor.

5. Conclusion

In this paper, thermodynamically and mathematically consistent constitutive equations


suitable for characterizing shock wave propagation in an anisotropic material are pre-
sented. Based on the methodology described herein, two fundamental tensors aij and bij
which represent anisotropic material properties and the equation of state for anisotropic
materials have been defined. The EOS represents a physical generalisation of the classical
Mie–Grüneisen equation of state for isotropic materials. Based on the generalised decom-
position of stress tensor, the modified Mie–Grüneisen equation of state, and the modified
Hill criteria, combined with associated flow rule, a system of constitutive equations suit-
able for shock wave propagation have been formulated. Numerical simulations based
on the new system of constitutive equations were performed and numerical verification
of the model has been done by comparison with LS-DYNA simulations in the special case
of anisotropy - i.e. isotropic materials. Good agreement was observed between LS-DYNA
simulations and the numerical simulation based on the new model. In this paper, the
behaviour of the aluminum alloy 7010-T6 under shock loading conditions was also inves-
tigated. Plate impact experiments on aluminum alloy 7010-T6 were carried out and pub-
lished by Vignjevic et al. (2002). A comparison of the experimental HELs with numerical
simulation shows an excellent agreement (maximum error is 1.4%). Furthermore, the good
agreement of the general pulse shape and Hugoniot stress level suggests that the EOS is
performing satisfactorily. Besides, there is an interesting possibility that the strain-rate sen-
sitivity is itself orientation dependent in 7010-T6 (see Vignjevic et al. (2002)). It has been
described by Vignjevic et al. (2002) that the spall strengths have been shown to be similar
in both orientations at lower impact stresses, while at higher levels, the spall strength is
higher in the longitudinal direction. This would seem to indicate a higher degree of
strain-rate sensitivity in the longitudinal orientation, and would seem to agree with the
observations made with the HELs (see Vignjevic et al. (2002)). Therefore, further develop-
ment of the constitutive equations taking into account strain rate sensitivity is required. In
order to be consistent with fundamental physics laws, and to be in a good agreement with
experimental data, it is important to describe the plastic behaviour using a non-associated
model. In this case, Hill’s shape function can be used as a yield criterion (e.g. Hill-M yield
surface) and flow potential (e.g. Hill-TCM surface) (see Lukyanov (2006)). In all these
cases it is clear that the formulation of the anisotropic plastic flow must be done indepen-
dently from the generalised hydrostatic ‘‘pressure”. To be able to simulate these signals
accurately, further work is required in the development of spall models, especially if one
is to simulate the complex spall behavior observed in Al7010-T6, where the spall strength
increases with strain rate, but reduces with stress levels. This will require further work both
on the experimental and constitutive modeling levels.

Acknowledgements

I would like to thank Cranfield University and all colleagues from CISM group for sup-
porting this work. Many useful suggestions made by anonymous referees are also greatly
appreciated.
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 165

References

Anderson, C.E., Cox, P.A., Johnson, G.R., Maudlin, P.J., 1994. A constitutive formulation for anisotropic
materials suitable for wave propagation computer program-II. Comput. Mech. 15, 201–223.
Asay, J.R., Shahinpoor, M., 1993. High-Pressure Shock Compression of Solids. Springer, New York.
Bammann, D.J., Aifantis, E.C., 1987. A model for finite-deformation plasticity. Acta Mech. 69, 97–117.
Barker, L.M., Hollenbach, R.E., 1972. Laser interferometer for measuring high velocities of any reflecting
surface. J. Appl. Phys. 43, 4669–4675.
Boidin, X., Chevrier, P., Klepaczko, J.R., Sabar, H., 2006. Identification of damage mechanism and validation of
a fracture model based on mesoscale approach in spalling of titanium alloy. Int. J. Solids Struct. 43, 4595–
4615.
Bourne, N.K., Stevens, G.S., 2001. A gas gun for plane and shear loading of inert and explosive targets. Rev. Sci.
Instrum. 72 (4), 2214–2218.
Bourne, N.K., 2003. A 50 mm bore gas gun for dynamic loading of materials and structures. Meas. Sci. Technol.
14, 273–278.
Bourne, N.K., Gray III, G.T., 2003. Equation of state of polytetrafluoroethylene. J. Appl. Phys. 93 (11), 8966–
8969.
Bronkhorst, C.A., Cerreta, E.K., Xue, Q., Maudlin, P.J., Mason, T.A., Gray III, G.T., 2006. An experimental
and numerical study of the localization behavior of tantalum and stainless steel. Int. J. Plast. 22 (7), 1304–
1335.
Butcher, B.M., 1968. Spallation of 6061-T6 Aluminum: Behavior of Dense Media under High Dynamic Pressure.
Gordon and Breach, New York.
Chen, J.K., Allahdadi, A., Carney, T., 1997. High-velocity Impact of Graphite/Epoxy Composite Laminates.
Comp. Sci. Techn. 57, 1268–1379.
Chidambaram, R., Sharma, S.M., 1991. Frontiers in high-pressure physics research. Current Sci. 60 (7), 397–
408.
Chijioke, A.D., Nellis, W.J., Silvera, I.F., 2005. High-pressure equations of state of Al, Cu, Ta, and W. J. Appl.
Phys. 98, 073526.
Dafalias, Y., 1985. The plastic spin. J. Appl. Mech. 52, 865–871.
Davison, L., Graham, R.A., 1979. Shock compression of solids. Phys. Rep. 55, 255–379.
Drumheller, D.S., 1998. Introduction to Wave Propagation in Nonlinear Fluids and Solids. Cambridge
University Press, Cambridge, UK.
Eliezer, S., Ghatak, A., Hora, H., Teller, E., 1986. An introduction to equations of state, theory and applications.
Cambridge University Press, Cambridge.
Gray III, G.T., Bourne, N.K., Zocher, M.A., Maudlin, P.J., Millett, J.C.F., 2000. Influence of Crystallographic
Anisotropy on the Hopkinson Fracture ‘‘Spallation” of Zirconium. In: Furnish, M.D., Chhabildas, L.C.,
Hixson, R.S. (Eds.), Shock Compression of Condensed Matter – 1999. AIP Press, Woodbury, NY, pp. 509–
512.
Gray III, G.T., Lopez, M.F., Bourne, N.K., Millett, J.C.F., Vecchio, K.S., 2002. Influence of Microstructural
Anisotropy on the Spallation of 1080 Eutectoid Steel. In: Furnish, M.D., Thadhani, N.N., Horie, Y. (Eds.),
Shock Compression of Condensed Matter-2001, Melville, NY. AIP Press, pp. 479–483.
Gebbeken, N., Greulich, S., Pietzsch, A., 2006. Hugoniot properties for concrete determined by full-scale
detonation experiments and flyer-plate-impact tests. Hugoniot properties for concrete determined by full scale
detonation experiments and flyer-plate-impact tests. Int J. Impact Engng. 32, 2017–2031.
Gu, Y.B., Ravichandran, G., 2006. Dynamic behavior of selected ceramic powders. Int. J. Impact Engng. 32,
1768–1785.
Hayhurst, C.J., Hiermaier, S.J., Clegg, R.A., Riedel, W., Lambert, M., 1999. Development of material models for
Nextel and Kevlar-epoxy for high pressures and strain rates. Int. J. Impact Engng. 23 (1), 365–376.
Johnson, J.N., Barker, L.M., 1969. Dislocation dynamics and steady plastic wave profiles in 6061-T6 aluminum.
J. Appl. Phys. 40, 4321–4334.
Johnson, J.N., Hixson, R.S., Gray III, G.T., 1994. Shock-wave compression and release of aluminum/ceramic
composites. J. Appl. Phys. 76, 5706–5718.
Kanel, G.I., Ivanov, M.F., Parshikov, A.N., 1995. Computer simulation of the heterogeneous materials response
to the impact loading. Int. J. Impact Engng. 17, 455–464.
166 A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167

Kanel, G.I., Razorenov, S.V., Bogatch, A., Utkin, A.V., Fortov, V.E., 1996a. Spall fracture properties of
aluminum and magnesium at high temperatures. J. Appl. Phys. 79 (11), 8310–8316.
Kanel, G.I., Razorenov, S.V., Utkin, A.V., 1996b. Spallation in solids under shock-wave loading: analysis of
dynamic flow, methodology of measurements and constitutive factors. In: Davison, L., Grady, D.E., Mohsen,
S. (Eds.), High Pressure Shock Compression of Solids, II. Springer, Berlin.
Kanel, G.I., Razorenov, S.V., Bogatch, A., Utkin, A.V., Grady, D.E., 1997. Simulation of spall fracture of
aluminum and magnesium over a wide range of load duration and temperature. Int. J. Impact Engng. 20
(6–10), 467–478.
Kanel, G.I., 1998. Some new data on deformation and fracture of solids under shock-wave loading. J. Mech.
Phys. Solids 46 (10), 1869–1886.
Kanel, G.I., Baumung, K., Bluhm, H., Fortov, V.E., 1998. Possible applications of the ion beams technique for
investigations in the field of equation of state. Nucl. Instr. and Meth. in Phys. Res. A 415 (3), 509–516.
Kiselev, A.B., Lukyanov, A.A., 2002. Mathematical Modeling of Dynamic Processes of Irreversible Deforming,
Micro- and Macrofracture of Solids and Structures. Int. J. Forming Process. 5, 359–362.
Krüger, L., Meyer, L.W., Razorenov, S.V., Kanel, G.I., 2003. Investigation of dynamic flow and strength
properties of Ti-6-22-22S at normal and elevated temperatures. Int. J. Impact Engng. 28 (8), 877–890.
Livermore Software Technology Corporation. LS-DYNA : Theoretical Manual, May 1998.
Lopatnikov, S.L., Gama, B.A., Haque, J.M., Krauthauser, C., Gillespie Jr., J.W., 2004. High-velocity plate
impact of metal foams. Int. J. Impact Engng. 30, 421–445.
Lukyanov, A.A., 2006. Thermodynamically consistent anisotropic plasticity model. Proceeding of IPC 2006,
ASME, ISBN 0-7918-3788-2.
Meyers, M.A., 1994. Dynamic Behavior of Materials. Wiley, Inc., New York.
Malvern, L.E., 1969. Introduction to the Mechanics of a Continuous Medium. Prentice-Hall, Englewood Cliffs,
NJ.
Millett, J.C.F., Bourne, N.K., Gray III, G.T., 2002. Behavior of the shape memory alloy NiTi during one-
dimensional shock loading. J. Appl. Phys. 92 (6), 3107–3110.
Millett, J.C.F., Bourne, N.K., 2001. Lateral stress measurements in a shock loaded alumina: Shear strength and
delayed failure. J. Mat. Sci. 36 (14), 3409–3414.
Nellis, W.J., Mitchell, A.C., Young, D.A., 2003. Equation-of-state measurements for aluminum, copper, and
tantalum in the pressure range 80-440 GPa (0.8-4.4 Mbar). J. Appl. Phys. 93 (1), 304–310.
Petit, J., Dequiedt, J.L., 2006. Constitutive relations for copper under shock wave loading: Twinning activation.
Mech. Mater. 38, 173–185.
Rosenberg, Z., Luttwak, G., Yesherun, Y., Partom, Y., 1983. Spall studies of differently treated 2024Al specimen.
J. Appl. Phys. 54, 2147–2152.
Rosenberg, Z., Yaziv, D., Partom, Y., 1980. Calibration of foil-like manganin gauges in planar shock wave
experiments. J. Appl. Phys. 51, 3702–3705.
Rubin, M.B., 1990. Analysis of weak shocks in 6061-T6 aluminum. In: Schmidt, S.C., Johnson, J.N., Davison,
L.W. (Eds.), Shock Waves in Condensed Matter-1989. American Institute of Physics, New York, pp. 321–328.
Sedov, L.I., 1972. A Course in Continuum Mechanics. Wolters-Noordhoff, Groningen.
Steinberg, D.J., Cochran, S.G., Guinan, M.W., 1980. A constitutive model for metals applicable at high-strain
rate. J. Appl. Phys. 51 (3), 1498–1504.
Steinberg, D.J., 1991. Equation of State and Strength Properties of Selected Materials, Report No. UCRL-
MA-106439, Lawrence Livermore National Laboratory, Livermore, CA.
Stevens, A.L., Tuler, F.R., 1971. Effect of shock precompression on the dynamic fracture strength of 1020 steel nd
6061-T6 Aluminum. J. Appl. Phys. 42 (13), 5665–5670.
Stoffel, M., 2005. An experimental method to validate viscoplastic constitutive equations in the dynamic response
of plates. Mech. Mater. 37 (12), 1210–1222.
Swegle, J.W., Grady, D.E., 1985. Shock viscosity and the prediction of shock wave rise times. J. Appl. Phys. 58
(2), 692–701.
Vignjevic, R., Bourne, N.K., Millett, J.C.F., De Vuyst, T., 2002. Effects of orientation on the strength of the
aluminum alloy 7010-T6 during shock loading: Experiment and simulation. J. Appl. Phys. 92 (8), 4342–4348.
Wallace, D.C., 1981. Nature of the process of overdriven shocks in metals. Phys. Rev. B 24 (10), 5607–5615.
Wackerle, J., 1962. Shock-wave compression of quartz. J. Appl. Phys. 33, 922–937.
Wilkins, M.L., 1969. Calculation of elastic–plastic flow. UCRL-7322. Rev. 1, Lawrence Livermore National
Laboratory.
A.A. Lukyanov / International Journal of Plasticity 24 (2008) 140–167 167

Zaretsky, E.B., Kanel, G.I., Razorenov, S.V., Baumung, K., 2005. Impact strength properties of nickel-based
refractory superalloys at normal and elevated temperatures. Int. J. Impact Engng. 31 (1), 41–54.
Zel’dovich, Y.B., Raizer, Y.P., 1966. In: Physics of Shock Waves and High-temperature Hydrodynamic
Phenomena, vols. 1 and 2. Academic Press, New York.

View publication stats

You might also like