An Anisotropic Elastoplastic Constitutive Formulation

Continuum Mech. Thermodyn.
(2018) 30:825–860
https://doi.org/10.1007/s00161-018-0645-7
O R I G I NA L A RT I C L E
M. K. Mohd Nor · N. Ma’at · C. S. Ho
An anisotropic elastoplastic constitutive formulation

generalised for orthotropic materials
Received: 26 October 2017 / Accepted: 11 March 2018 / Published online: 27 March 2018
© Springer-Verlag GmbH Germany, part of Springer Nature 2018
Abstract This paper presents a finite strain constitutive model to predict a complex elastoplastic deformation
behaviour that involves very high pressures and shockwaves in orthotropic materials using an anisotropic
Hill’s yield criterion by means of the evolving structural tensors. The yield surface of this hyperelastic–plastic
constitutive model is aligned uniquely within the principal stress space due to the combination of Mandel stress
tensor and a new generalised orthotropic pressure. The formulation is developed in the isoclinic configuration
and allows for a unique treatment for elastic and plastic orthotropy. An isotropic hardening is adopted to define
the evolution of plastic orthotropy. The important feature of the proposed hyperelastic–plastic constitutive
model is the introduction of anisotropic effect in the Mie–Gruneisen equation of state (EOS). The formulation
is further combined with Grady spall failure model to predict spall failure in the materials. The proposed
constitutive model is implemented as a new material model in the Lawrence Livermore National Laboratory
(LLNL)-DYNA3D code of UTHM’s version, named Material Type 92 (Mat92). The combination of the
proposed stress tensor decomposition and the Mie–Gruneisen EOS requires some modifications in the code
to reflect the formulation of the generalised orthotropic pressure. The validation approach is also presented
in this paper for guidance purpose. The ψ tensor used to define the alignment of the adopted yield surface is
first validated. This is continued with an internal validation related to elastic isotropic, elastic orthotropic and
elastic–plastic orthotropic of the proposed formulation before a comparison against range of plate impact test
data at 234, 450 and 895 ms−1 impact velocities is performed. A good agreement is obtained in each test.
Keywords Elastoplastic deformation · Shockwave propagation · Spall failure · Orthotropic materials
1 Introduction
Most of the engineering materials exhibit orthotropic behaviour while undergoing large elastoplastic defor-
mation at unit-cell level due to the preferred orientation. Sheet of aluminium alloys formed by squeezing thick
sections of metal between heavy rollers is an example of orthotropic materials produced in manufacturing
industry. There are many other examples such as advanced high-strength steel (AHSS) and fibre-reinforced
elastomers. The prediction of aluminium alloys behaviour undergoing finite strain deformation including shock
wave propagation and spall failure has attracted attention due to its broad engineering applications. The formu-
lation to describe complex material responses at high strain rate and pressures should be one of the integrations
between elasticity, plasticity and equation of state (EOS) including spall failure. It is difficult for engineer
and the user of metal structures to ignore the realm of this topics, regardless of his particular field of interest.
Communicated by Andreas Öchsner.
M. K. Mohd Nor (B) · N. Ma’at · C. S. Ho
Crashworthiness and Collisions Research Group, Mechanical Failure Prevention and Reliability Research Center,
Faculty of Mechanical and Manufacturing Engineering, Universiti Tun Hussein Onn Malaysia, Parit Raja, Malaysia
E-mail: khir@uthm.edu.my
Content courtesy of Springer Nature, terms of use apply. Rights reserved.

826 M. K. Mohd Nor et al.
Lack of this knowledge can impose limitations on engineering design. The prime motivation in this work,
therefore, is to propose a new constitutive model that is capable of modelling of the deformation behaviour of
such materials undergoing finite strain deformation including spall failure.
Earlier work by Butcher [11] predicted the spall strength of 6061-T6 should vary in accordance with the one-
dimensional stress yield strength according to orientation. Further it was concluded that there is no significant
effect on crack formation. Other works have also conducted to investigate the behaviour of spall response
of aluminium alloys. Stevens and Tuler [51], for instance, concluded that the degree of precompression,
which is the shock amplitude, had no effect on the spall strength of 6061-T6. In addition, it is shown that
the spall strength of 2024-T86 decreased with increasing temperature [45]. However, both works concerned
on geometrical effects. During shock loading, the compressive input stress is details on the rising part of the
shock defined by the HEL of the material. It is observed that the shock is reflected back as a release wave when
it reaches a free surface, which consequently takes the material back to ambient stress conditions. Releases
from the rear of flyer and target can be arranged by controlling the thickness of the specimen and flyer plate
to meet in the middle of the target itself, and where they do so, a zone of net tension will result. If that tension
exceeds the tensile strength of the material, the net result is failure (spall), which can be detected by appropriate
measurement techniques [53].
In many hydrocodes, the common approach is to model spall failure using pressure cut-off and maximum
principal stress. Pressure cut-off and maximum principal stress can be regarded the simplest models. Generally
speaking, the pressure cut-off model compares the pressure with a user-defined pressure cut-off value. The
material is assumed to have spall when the pressure is less than this pressure cut-off value. The deviatoric
stress tensor and the pressure are set to zero, and no hydrostatic tension is subsequently allowed. The pressure
is set to zero again with no hydrostatic tension being subsequently allowed and the maximum principal stress
criterion checks whether the detected the deviatoric stress tensor.
Grady spall model is an energy-based failure model. It assumes the material spall when the strain energy
reaches a certain level. The model provides a rigorous mathematical modelling of fragmentation. Grady reports
on results from dynamic compression and dynamic tension (spall) tests [20]. Grady failure model was first
developed to predict spall in ductile metals in 1997 as performed in Grady and Kipp [21]. This model has
been used to analyse fracture and fragmentation of naturally fragmenting munitions of different materials
and geometries [57]. The model has shown great ability to predict the test results in terms of fragment mass
distributions and proved to be excellent in modelling the breakup of 4140 steel, 70% tungsten and thick-walled
Aermet 100, and reasonable in matching the data from 8-in. Aermet 100 tests. Fountzoulas et al. [19] have
proposed a constitutive model to predict impact and penetration of tungsten carbide sphere into high-strength
low-alloy (hsla)-100 steel targets. Generally, the Grady failure model produced a good agreement with respect
to the experimental data of the crater diameter and dept. Further, De Vuyst [14] proposed a constitutive model
for impact on water using spall model pressure cut-off, maximum principal stress and Grady spall criterion.
The results obtained from the pressure-based spall criterion and the maximum principal stress spall criterion
are almost identical. It is observed that these spall models are inadequate to model spall behaviour qualitatively,
while a better agreement is obtained using Grady failure model.
With respect to the behaviour of engineering metals at quasi-static strain rates, there are much efforts have
been organised as shown in Sinha and Gosh [39,40,48]. For a comprehensive discussion on the elastic solution
of the transversely isotropic materials under various loading conditions, the reader is directed to [1,2,18].
Further, the topic related to the anisotropic influence on material behaviour undergoing finite strain deformation,
including shock wave propagation, has been extensively studied in this community; see, for example, [22,29–
31,38,42,49,50,53]. In addition, the investigation on the shock wave propagation in anisotropic materials has
shown momentous progress in recent decades, specifically in the isotropic solid-state physics and mechanics
(Zel’dovich and Raizer, 1966; Davison and Graham, 1979; Eliezer et al., 1986; Asay and Shahinpoor, 1993;
Meyers, 1994; Drumheller, 1998).
With respect to yield function, the theory proposed by Hill [24] can be regarded as the first anisotropic
homogeneous yield function of degree two to describe an orthotropic plastic response of rolled sheet. This is
generally accepted as a solid foundation for the case of metals. The ability to represent the full behaviour of
orthotropic materials is one of the advantages of Hill’s yield criterion. For instance, Hill’s formulation in a
plane stress case which refers to the principle orthotropic axes and consist of a shear stress component. When
the shear parts disappeared, the yield function is limited to planar isotropy and this is important as a yield
function as emphasised in Barlat [7] and Barlat and Lian [8].
The benefit is that the formulation of Hill’s effective stress conserves this yield criterion homogeneous
characteristic. Thus, by using Hill’s yield criterion the convexity of the yield surface is upheld. The expectations

An anisotropic elastoplastic constitutive formulation generalised 827
of the Hill’s yield criterion are easy to understand in terms of formulation. Furthermore, the parameters consist
in the yield functions expression and have a direct physical meaning and this is the main cause for the criterion’s
wide use in practice. Moreover, this model provides a simple formulation in a three-dimensional case [6]. The
yield function meanwhile can be characterised from a low number of mechanical parameters. One consideration
particularly which motivates the need for modification of the Hill yield criterion using an alternative stress
tensor is considered in this work.
2 New constitutive formulation
It can be observed that much research has been conducted to improve the prediction capability of many finite
element codes, leading to results in various technologies involving analytical and computational methods.
However, it is generally agreed that there is a need for improved constitutive model to deal with the material
behaviour under impulsive loading due to high-velocity impact.
The important feature of the new hyperelastic–plastic constitutive formulation is the multiplicative decom-
position of the deformation gradient F. A new definition of Mandel stress tensor is defined by a combination
with the new generalised orthotropic pressure. The Hill’s yield criterion is adopted to characterise plastic
orthotropy by means of the evolving structural tensors that is defined in the isoclinic configuration. Plastic
anisotropy is considered through yield surface and an isotropic hardening defined in the unique alignment
of deviatoric plane within the stress space. To predict a complex deformation behaviour involves very high
pressures and shockwaves, a newly shock equation of state (EOS) of the generalised orthotropic pressure and
spall failure are adopted.
2.1 Kinematics for finite strain deformations
Anisotropic materials are classified as orthotropic if they have three mutually orthogonal symmetry planes.
In many engineering components, the orthotropy is induced by a number of manufacturing processes such
as rolling and stamping. Composites are also examples of orthotropic materials. In fact, most elastoplastic
materials exhibit anisotropic behaviour due to their structure orientation and evolution [16].
The constitutive equation based on additive decomposition of generalised strain measures is basically not
suitable for the modelling of orthotropic materials’ behaviour. As demonstrated by Itskov [27], this kind of
constitutive model leads to spurious shear stresses which are independent of the elastic material properties for
orthotropic materials.
Conversely, the multiplicative decomposition of the deformation gradient-based model provides the true
behaviour of a constant shear stress [27]. In addition, the evolution of material symmetry in orthotropic
materials due to large deformations could not be tracked by the additive strain decomposition-based model [28].
Therefore, it can be deduced that this approach is only valid in the case of small strain elastoplasticity. Strictly
speaking, it is restricted for the case where loadings are colinear with the axis of material orthotropy. Using
this motivation, the construction of the new hyperelastic–plastic constitutive model for orthotropic materials
in this manuscript is formulated based on the multiplicative decomposition of the deformation gradient F:
F = Fe F p (1)
where Fe and F p represent thermo-elastic part of the deformation and plastic part of the deformation (dislocation
mechanics), respectively. The intermediate (generally non-Euclidean) configuration corresponds to elastically
unloaded material and is known as the elastic reference configuration which can be physically obtained by
elastic unloading of material (unstressed condition).
Using the definition in Eq. (1), the elastic right Cauchy–Green tensor Ce and the elastic Green–Lagrange
strain tensor Ee can be expressed as
1 1 T
Ce = FeT · Fe , Ee = (Ce −I) = Fe · Fe −I (2)
2 2
The experimental work in Man (1995) shows a strong correlation between elastic and plastic material sym-
metries. Therefore, the constitutive model is developed and integrated in the isoclinic configuration. In other
words, the hyperelastic part of the constitutive model is based on the assumption that the principal directions of
material elastic and plastic orthotropy coincide and are not influenced by inelastic deformation. This definition

Fig. 1 Definition of isoclinic configuration
simplifies the numerical implementation since the explicit use of any corotational rate can be avoided [4,55].
To avoid confusion, () is used in this manuscript upon each of kinematic and kinetic variables defined with
respect to the isoclinic configuration.
The structural tensors Mii | i = 1, 2, 3 [10] are introduced to construct the orthotropy symmetric group
ϑ. These tensors can be defined as a tensor product of unit vectors as M1 = n1 ⊗ n1 , M2 = n2 ⊗ n2
and M3 = n3 ⊗ n3 where n1 , n2 and n3 are unit vectors represent an orthonormal frame of the material.
Figure 1 shows the schematic diagram of the configurations typically used to describe motion of continuum.
A triad of unit vectors which represent the substructural spin (material symmetries) is schematically depicted
by two orthogonal axes with arrows. Further, the orthogonal dash lines without arrows denote the principal

directions of the continuum. By using the structural tensors M̂ii i = 1, 2, 3 in the isoclinic configuration,
the elastic free energy function for orthotropic materials considers the principal directions of material elastic
orthotropy (material symmetry) in this configuration to be aligned with the unit (director) vectors n̂1 , n̂2
and n̂3 . Push forward transformations of the structural tensors from an initial configuration o to elastically
unloaded configuration ¯ p can be defined as Mii = F p Mii F−1 p . As depicted in Fig. 1, the orientation of the
ˆ
orthogonal triad vectors in the isoclinic configuration i is the same as the unit vectors in 0 . To be precise,
the substructures’ orientation in ˆ i and 0 is identical.
The structural tensors Mii are pulled back from the elastically unloaded configuration (having an arbitrary
orientation) ¯ p to the isoclinic configuration ˆ i by rotating back for plastically induced rigid body rotation
due to plastic-related deformations using the following orthonormal transformation.
M̂ii = QTp Mii Q p (3)
where Q p is an orthonormal tensor that defines the rigid rotation due to plastic-related deformation and is
T
updated as Qn+1 p Qn+1
p = Q p · QTp = I. Both elastic stretching and rotation are contained in the elastic
part Fe of the deformation gradient F. Plastic part of F with respect to ¯ p and
ˆ i is represented by F p and F̂ p ,
respectively. The plastic rotation R p is assigned to F p to ensure there is no rotation to the material principal
axes of orthotropy (remains fixed or unaltered by plastic deformation) in the isoclinic configuration ˆ i.
As examined in Schröder and Hackl [47], the rotation and distortion during elastoplastic deformation are
contained in the elastic part of F as a result of isoclinic configuration. Despite general incompatibility tensor
fields between elastic and plastic deformations, the elastic deformation may incidentally compatible to plastic
deformation when plastic or elastic deformation is homogeneous [47]. Alternatively, the plastic rotation R p
can be assigned to the deformation due to damage Fd to define F̂e = Fe Fd R p if one considers the elastic

material parameters evolve due to damage since the changes in material compliance due to damage [55]. The
formulation-related damage is, however, not considered in this work.
The elastic and plastic parts of the deformation gradient F of the proposed constitutive model then can be
defined explicitly in the isoclinic configuration as:
F̂e = Fe , F̂ p = R Tp F p = R Tp R p U p = U p (4)
where plastic rotation R p is an orthogonal tensor obtained by the polar decomposition of plastic deformation
gradient (rotation induced by plastic deformation). The update of this tensor is similar to Q p , and U p refers
to plastic right stretch tensor. Subsequently, the total velocity gradient L̂ can be decomposed additively as
follows:
˙ ˙ ˙ ˙
L̂ = F̂ · F̂−1 = F̂e · F̂e−1 +F̂e · F̂ p · F̂−1 −1
p · F̂e = L̂e +L̂ p (5)

An incompressibility constraint is assumed to hold for plastic deformation which therefore gives det F̂ p = 1.
2.2 New stress tensor decomposition
The shock response of an orthotropic material cannot be accurately predicted using the conventional decompo-
sition of the stress tensor into isotropic and deviatoric parts [3]. Constitutive models developed for modelling
of shock wave propagation in solids comprise two parts, an equation of state (EOS) and a strength model which
define the response of the material to uniform compression (change in volume) and the response of the material
to shear deformation (change in shape), respectively. This separation of material response into volumetric and
deviatoric strain components is matched for isotropic materials which have an isotropicelastic stiffness tensor
Ci jkl . As a consequence, the spherical part of the stress tensor −Pδi j = Ci jkl δkl ε pp 3, being a product of
two isotropic tensors, is isotropic itself. Furthermore, the isotropy of Ci jkl results in the colinearity of the
principal axis of the stress and strain tensors. In other words, components of stress and strain are proportional
to each other and orthogonality between the volumetric and deviatoric components of strain is reflected in
orthogonality between the volumetric and deviatoric components of stress.
It can be shown in the following equations that this orthogonality reflects to the definition of pressure as
an average normal stress:
σi j = −Pδi j + S i j
σi j δi j = −Pδkl δkl + S i j δi j
σi j δi j = −Pδkl δkl (6)
where Si j δi j = 0. Using the final expression in the above equation, it is simple to show that
σi j δi j σii
P=− =− (7)
δkl δkl 3
Further, using the above definition, Hooke’s law for isotropic materials can be expressed as

2
σi j = λεkk δi j +2μεi j = λ + μ δi j εkk +2μεidj = −Pδi j + Si j (8)
3
where εidj and Si j refer to the deviatoric parts of strain and stress tensor, respectively, while λ and μ are Lame
parameters. This conventional decomposition is only suitable for isotropic materials since both parts of the
stress tensor induce volumetric and deviatoric strain.
However, the decomposition does not hold for orthotropic materials since the colinearity between stress
and strain is no longer in place specifically when the materials undergoing shockwave propagation. It should
be noted that the simplicity in the modelling of elastic behaviour in isotropic materials is a consequence of
the fact that the material can be fully characterised by two elastic parameters only. In the case of elastic
orthotropic materials, nine parameters are required to define the stress–strain relationship using the stiffness
and compliance matrices as follows
σi j = Ci jkl εkl (9)

εi j = Ci−1
jkl σkl = Bi jkl σkl (10)
where Ci jkl and Bi jkl are the stiffness and compliance matrices, respectively. The stiffness matrix can be
written in Voigt notation as
⎛ ⎞ ⎛ ⎞
C1111 C1122 C1133 0 0 0 C11 C12 C13 0 0 0
⎜ C1122 C2222 C2233 0 0 0 ⎟ ⎜ C12 C22 C23 0 0 0 ⎟
⎜ ⎟ ⎜ ⎟
⎜ C1133 C2233 C3333 0 0 0 ⎟ ⎜ C13 C23 C33 0 0 0 ⎟
⎜ 0 =⎜
0 ⎟ 0 ⎟
(11)
⎜ 0 0 C1212 0 ⎟ ⎜ 0 0 0 C44 0 ⎟
⎝ 0 0 0 0 C1313 0 ⎠ ⎝ 0 0 0 0 C55 0 ⎠
0 0 0 0 0 C2323 0 0 0 0 0 C66
The above expression of stiffness matrix is equal to
⎛ 1−ν23 ν32 ν21 +ν23 ν31 ν31 +ν21 ν32 ⎞
E2 E3 E2 E3 E2 E3 0 0 0
⎜ ν12 +ν32 ν13 1−ν13 ν31 ν23 +ν13 ν21 ⎟
⎜ E1 E3 E1 E3 E1 E2 0 0 0 ⎟
1 ⎜
⎜ ν13 +ν12 ν23 ν32 +ν12 ν31 1−ν12 ν21
0 0 0
⎟
⎟ (12)
⎜ ⎟
E1 E2 E1 E3 E1 E2
⎜ 0 0 0 G 23 0 0 ⎟
⎝ 0 0 0 0 G 31 0 ⎠
0 0 0 0 0 G 12
where ν, E and G are Poisson’s ratio, Young’s modulus and shear modulus, respectively. In addition, is set
equals to 1−ν12 ν21 −ν23 νE321−ν31 ν13 −2ν21 ν32 ν13
E2 E3 . The material compliance tensor can be expressed using the similar
Voigt notation. As has been mentioned, the isotropic states of stress (pressure) are induced by the isotropic
states of strain for isotropic materials, which are related to bulk modulus. Again, it must be emphasised that
the equivalent relationship cannot be adopted for orthotropic materials.
If one maintains the assumption that pressure is the state of stress induced by an isotropic state of strain
(uniform compression or expansion), then more general definition of pressure is required [54]. This leads to a
number of possible definitions of pressure as a vector in the principal stress space which does not colinear with
the conventional hydrostatic alignment for orthotropic materials. To explore this statement further, Vignjevic
has proposed a new expression for generalised pressure or stress related to uniform compression. The ability
to describe shock propagation in orthotropic materials is investigated with experimental plate impact data and
showed a good agreement with the physical behaviour of the considered material (carbon fibre-reinforced
epoxy).
To derive the formulation of this generalised pressure, let first write the stress due to the isotropic component
of strain (isotropic strain pressure) as

− P̃ψi j = Ci jkl δkl εss 3 = Ci jkk εv (13)
where Ci jkl δkl = Ci j Ckl δkl = Ci jkk , ψi j = 0 ∀ i = j, ψi j = 0 ∀ i = j and εv = εss /3. In the above
equation, P̃ and ψi j can be defined as

1
− P̃ = εv Ci jkk Ci jll (14)
ψst ψst
and
1
ψi j = −Ci jkk εv / P̃ = Ci jkk / C pr kk C prll (15)
ψst ψst
The double contraction tensor ψst ψst must be defined to uniquely define P̃ and tensor ψi j . In this work, the
double contraction is to set ψst ψst = 3. Note that the tensor ψi j is fully defined by the material elastic stiffness
properties. Further, Eqs. (13)–(15) can be expressed in Voigt notation as shown in Eqs. (16)–(18), respectively:
⎧ ⎫ ⎡ ⎤⎧ ⎫
⎪
⎪ ψ1 ⎪⎪ C11 C12 C13 0 0 0 ⎪ εv ⎪
⎪ ⎪
⎪
⎪ ψ2 ⎪⎪ ⎢ C12 C22 C23 0 ⎥⎪ ⎪ εv ⎪
⎪
⎪
⎨ ⎬ ⎪ ⎢
0 0
⎥⎪ ⎨ ⎪ ⎬
ψ3 ⎢ C13 C23 C33 0 0 0 ⎥ εv
P̃ = −⎢ ⎥ (16)
⎪
⎪ 0 ⎪⎪ ⎢ 0 0 0 C44 0 0 ⎥⎪ ⎪ 0⎪⎪
⎪
⎪ 0 ⎪⎪ ⎣ ⎦ ⎪
⎪ 0⎪⎪
⎩ ⎪
⎪ ⎭
0 0 0 0 C55 0
⎩ ⎪
⎪ ⎭
0 0 0 0 0 0 C66 0

The scalar P̃ which is used to defines the magnitude of pressure can be expressed as

(C11 +C12 +C13 )2 + (C12 +C22 +C23 )2 + (C13 +C23 +C33 )2
P̃ = − εv = −3K ψ εv (17)
3
The parameter K ψ = E/3 (1 − 2v) reduces to the conventional bulk modulus in the limit of material isotropy.
Tensor ψi j which is used defines direction of the new volumetric axis in stress space (defines points in stress
space) then can be defined as
Ci1 +Ci2 +Ci3
ψ(ii) = (18)
(C11 +C12 +C13 )2 +(C 12 +C 22 +C 23 ) +(C 13 +C 23 +C 33 )
2 2
3
Repeated indices in brackets in the above equation indicate no summation. The alternative formulation of
generalised pressure for orthotropic materials finally can be expressed as follows:
σkl ψkl
P̃ = (19)
ψsr ψsr
It should be noted that ψi j becomes δi j when dealing with isotropic materials. Hence, the new decompo-
sition reduces to the conventional decomposition developed for isotropic materials. This new stress tensor
decomposition has been used to develop a new yield criterion for orthotropic sheet metals under plane stress
conditions by assuming the yield surface to be circular in the new deviatoric plane [39]. The predictions of the
new effective stress expression showed good agreement with respect to the experimental data for 6000 series
aluminium alloy sheet (A6XXX-T4) and Al-killed cold-rolled steel sheet SPCE.
Bear in mind that any arbitrary stress state in stress space can be decomposed into hydrostatic and deviatoric
parts. Figure 2 describes the representation of this decomposition for isotropic materials which is best presented
in the principal stress space by a blue line with their directions perpendicular to each other. Recalling the new
stress tensor decomposition, the representations of this decomposition ψi j in the principal stress space are
represented by a purple line. This figure represents the graphical interpretation of the generalised pressure axes
of the new decomposition ψi j in three-dimensional space of the principal stress. It can be observed that this
decomposition leads to a shift of the pressure vector away from the common alignment (equal angle with the
principal stress directions). Based on the definition of the new stress tensor decomposition, any considered
orthotropic materials will uniquely define their own deviatoric plane (yield surface) within the principal stress
space.
2.3 Mandel stress tensor
In general, the Mandel stress tensor can be defined as follows [35]:
=C· S (20)
where C and S refer to right Cauchy–Green tensor and second Piola–Kirchhoff stress tensor, respectively.
¯ p as
These tensors can be expressed in the elastically unloaded intermediate configuration
C̄e = FeT · Fe (21)

S̄ = Fe−1 ·τ · Fe−T (22)
The Kirchhoff stress tensor τ in Eq. (22) is a symmetric tensor defined in the current configuration t as
τ = J · σ = det (F) ·σ (23)
where J is a volume ratio. Substituting Eqs. (21) and (22) into Eq. (20), the Mandel stress tensor in the
¯ p can be defined as
intermediate configuration
¯ = FeT · τ · Fe−T = det (F) · FeT ·σ·Fe−T

(24)

Fig. 2 ψ and δ as a vector in a principal stress space
The Mandel stress tensor defined in Eq. (24) is frequently used to describe the behaviour of plastic materials
ˆ i as a point for integration, the Mandel stress tensor can be rewritten
[26]. Choosing the isoclinic configuration
as
ˆ = F̂eT · τ · F̂e−T
(25)
Further, Eq. (24) also can be expressed as follows:
ˆ = F̂eT · τ · F̂e−T = F̂eT · det (F) ·σ·F̂e−T

= det (F) · F̂eT ·σ·F̂e−T (26)
Using σi j = Pδi j +Si j , the above equation can be defined as
ˆ = det (F) · F̂eT ·σ·F̂e−T = det (F) · F̂eT · (Pδ+S) · F̂e−T

(27)
The formulation of the new generalised pressure for orthotropic metals is introduced in Eq. (27):

ˆ = det (F) · F̂eT · S+ σψ · ψ · F̂e−T
(28)
ψψ
Equation (28) can be further extended as
ˆ = det (F) · F̂eT · S · F̂e−T + det (F) · F̂eT · σ ψ · ψ · F̂e−T

(29)
! ψψ
ˆ
!
=deviatoric
ˆp

where ˆ p denotes the volumetric part (pressure) of the new Mandel stress tensor. Focusing on the deviatoric
part instead of a full stress tensor, the deviatoric part of the new Mandel stress tensor is given by

ˆ σψ
= det (F) · F̂e σ−
T
· ψ F̂e−T (30)
ψψ
¯ i can be written as
Finally the new deviatoric Mandel stress tensor defined in the isoclinic configuration

ˆ = det (F) · F̂eT · σ · F̂e−T − det (F) · F̂eT · σψ ψ · F̂e−T

! ψψ
ˆ
!

ˆp

= det (F) · F̂eT · S · F̂e−T (31)
Only the symmetric part of the Mandel stress is considered in this work [44,56]. As asserted in Mont’ans
and Bathe [41], the existing experimental evidence shows that it is hard to deduce sound data about a continuum
elastic domain for the skew part of the Mandel stress tensor a and also for the plastic spin W p evolution.
Moreover, there is still much effort required in terms of experimental work to establish the elastic domain and
yield functions for the skew part of the Mandel stress tensor.
2.4 Coupling of the proposed stress decomposition with equation of state EOS
Appropriate constitutive equations to describe the strength effect and the equation of state must be investigated
to describe the anisotropic material response under shock loading. Therefore, the proposed formulation is
combined with an equation of state (EOS) in addition to the conservation laws to mathematically describe the
material’s nonlinear behaviour and propagation of strong shock waves in solids due to shock loading.
The EOS is a closure equation which allows for the solution of the balance equations across the shockwave
which usually defines the pressure as a function of internal energy (temperature) and density (volumetric
deformation) for a material undergoing deformation. Theoretically, the relationship described by EOS can
be determined from the thermodynamic properties of the material, and required no dynamic data. However,
practically, extensive dynamic experiment such as the planar shock wave experiment is required to characterise
data on the material’s behaviour at high strain rates. In contemporary hydrocodes, available EOSs are either
of an analytical or a tabulated type. In this manuscript, a very popular EOS that is extensively used for solid
continua the Mie–Gruneisen EOS [23,52], implemented in DYNA3D, is used. This an analytic EOS frequently
used with solid materials. The combination of the proposed stress tensor decomposition and the Mie–Gruneisen
EOS requires few modifications in the code to reflect the formulation of the generalised orthotropic pressure
and is discussed in Sect. 3.
It defines the pressure as a function of density ρ (referential density) or specific volume and specific internal
energy e as shown below
(v)
P = f (ρ, e) = pr (v) + (e − er (v)) (32)
v
where v is the specific volume, (v) is the Gruneisen gamma which defined as

∂P
(v) = v (33)
∂e v
Generally is defined as constant = 0 , or assumed that v00 = v = const alternatively. Functions pr and
er are considered known functions of v on some reference curve. There are few possible reference curves to
name such as the shock Hugoniot curve and the 0◦ K isotherm. However, the most widely adopted form of the
Mie–Gruneisen equation of state for solid materials which uses the shock Hugoniot as the reference curve is
defined below

P = f (ρ, e) = p H · + 1 − μ + ρe (34)
2
where p H refers to Hugoniot pressure, μ = ρρ0 − 1 is relative change in volume, is Gruneisen parameter, ρ
is density and e defined as the specific internal energy. The Rankine–Hugoniot equations for the shock jump
conditions can be regarded by defining a relation between any pair of the ρ, P, e, u p (the velocity of the particle
directly behind the shock) and U (the velocity of shockwave that propagates through the medium). There is
an empirical linear relationship between U and u p for many liquids and most solids:
U = c + Su p (35)

where c is the intercept of the U − u p curve (U shock velocity vs. u p particle velocity curve), and S is the
coefficient of the U − u p curve slope. The Hugoniot pressure and a shock velocity normally U are normally
defined as a nonlinear function of particle velocity u p as follows [52]:
u u 2
p p
U = c + S1 u p + S2 u p + S3 up (36)
U U
The Gruneisen’s gamma for the undeformed materials can be expressed by
γ0 + au
= (37)
1+u
Hence, the pressure as a function of Gruneisen equation of state with cubic shock velocity can be defined as
" #
ρ0 c2 μ 1+ 1− 2 μ− 2 μ2
PEOS = " 2 3
#2 + (1 + μ) · · E when μ > 0
μ
1−(S1 −1)μ−S2 μ+1 −S3 μ 2 (38)
(μ+1)
PEOS = ρ0 c2 μ + (1 + μ) · · E when μ < 0
where
E is the internal energy per initial specific volume,
S1 , S2 , S3 are the coefficients of the slope of the U − u p curve,
γ0 is the Gruneisen gamma for the undeformed material,
a is the first-order volume correction to γ0 ,
c, S1 , S2 , S3 , γ0 , a, ρ0 represent the material properties supplied by the user to characterise this EOS.
The combination of the proposed stress tensor decomposition and the Mie–Gruneisen EOS requires some
modifications to reflect the formulation of the generalised orthotropic pressure.
2.5 Elastic free energy function
The behaviour of orthotropic metals within elastic and plastic regimes is formulated as a free strain energy
function and a plastic level set function (orthotropic yield criterion), respectively. As mentioned previously, an
orthotropic symmetry group ϑ is used in this work and assumed to be unchanged during plastic deformation.
The Helmholtz free energy is used to formulate the elastic orthotropy as a function of evolving structural
tensors. The Helmholtz free energy is additively decomposed into elastic and plastic parts in the isoclinic
configuration ˆ i as

Ψ̂ = Ψ̂e Êe +Ψ̂ p(isot) (α) (39)

In the above expression, Ψ̂e Êe represents the energy stored due to elastic deformations defined in terms of
elastic Green–Lagrange strain tensor Êe . In addition, Ψ̂ p(isot) (α) represents energy resulting from isotropic
plastic hardening, where α is the isotropic hardening variable (i.e. accumulated plastic deformation). The
response of elastic material in the isoclinic configuration must be invariant under transformations of the
material symmetry group ϑ. This is necessary based on the definition of isotropic functions.

Ψ̂e QT Êe Q = Ψ̂e Êe , ∀Q ∈ ϑ, Êe (40)
where Q is orthogonal rotation tensor. Ψ̂e is then known as a ϑ-invariant function. Using the structural tensors
M̂ii , the elastic component of free energy function for orthotropic materials can be expressed in terms of
isotropic function in the isoclinic configuration as follows:

Ψ̂e = Ψ̂e Êe , M̂11 ,M̂22 = Ψ̂e QÊe QT , QM̂11 QT , QM̂22 QT (41)
Ψ̂e can be defined in terms of a set of invariants of the elastic Green–Lagrange strain tensor Êe . As mentioned
in the preceding section, push forward transformations of the structural tensors from an initial configuration
o to elastically unloaded configuration ¯ p can be defined as Mii = F p Mii F−1
p , and are pulled back from

the elastically unloaded configuration (having an arbitrary orientation) ¯ p to the isoclinic configuration
ˆ i by
rotating back for plastically induced rigid body rotation due to plastic-related deformations. It is simple to show
that using structural tensors that reflect the material symmetry of orthotropic materials, the irreducible invariants
and the additional pseudo-invariants in the isoclinic configuration ˆ i also can be established [16,46,59].
2.6 Orthotropic yield criterion
The yield function used to model the dependence on plastic anisotropy in this formulation is defined using the

structural tensors M̂ii i = 1, 2, 3 of the material symmetry group ϑ. The Hill’s yield criterion is adopted to
define the orthotropic yield function [24]. A yield surface of Hill’s yield criterion aligned uniquely within the
principal stress space is adopted to characterise plastic orthotropy by means of the evolving structural tensors.
Using this hypothesis, the hardening is modelled as an isotropic hardening in this constitutive model. Using
the new definition of Mandel stress tensor, the yield surface of the proposed formulation will always maintain
its initial shape (only change in size, no change in shape) and defined in a unique deviatoric plane within the
stress space. Using Eq. (31), the yield function can be written in terms of the symmetric Mandel stress tensor
as
fˆ = fˆ ˆ , α (42)
As mentioned above, α refers to isotropic hardening variable. The structural tensors M̂ii is introduced to
describe the properties of symmetric orthotropy as follows:

fˆ = fˆ ˆ ,M̂ii , α (43)
Accordingly, the plastic anisotropy of the new constitutive model is characterised by Hill’s anisotropy yield
function as follows:
fˆ = ˆ − fˆ (α) = 0
ˆ :ĥ :
(44)
where ĥ is a fourth-order tensor defined in the isoclinic configuration ˆ i . The dependence of the above
yield function on Hill’s yield criterion and structural tensors is represented by this tensor. fˆ (α) in the above
equation defines the evolving flow stress that is controlled by isotropic hardening. The Hill’s effective stress
can be expressed in terms of the deviatoric Mandel stress in the isoclinic configuration ˆ i as follows:
$ ⎡ ⎤
% 2 2 2
% ˆ z + G
ˆ y − ˆ z −
ˆ x + H ˆ x − ˆ 2yz + 2M
ˆ y + 2L ˆ 2zx + 2N
ˆ 2x y
%3 ⎢ F ⎥
ˆ¯ =%
&2 ⎣ + G + H ⎦ (45)
F
2.7 The evolution equations
The evolution of the proposed constitutive model is defined using the second law of thermodynamic framework.
The formulation can be expressed using the Clausius–Plank inequality as
D = S : Ė − Ψ̇ ≥ 0 (46)
where Ψ̇ is a rate of Helmholtz free energy function, and Ė can be defined as
1
Ė = Ċ (47)
2
Using Eqs. (47) and (46) can be rewritten as follows:
1
D=S: Ċ − Ψ̇ ≥ 0 (48)
2

Considering the free strain energy (Helmholtz free energy) function is represented by the elastic Cauchy–Green
' (
strain tensor Ce , and the strain-like internal variable describing isotropic plastic hardening ξ h , Ψ = Ψ Ce , ξ h ,
this function can be easily divided into elastic and plastic as
' (
Ψ = Ψe (Ce ) + Ψ p ξ h (49)
The differentiating of the above equation with respect to time leads to
∂Ψe ∂Ψ p ˙
Ψ̇ = : Ċe + · ξh (50)
∂Ce ∂ξ h
Substituting Eq. (50) into Eq. (49) gives

1 ∂Ψe ∂Ψ p ˙
D = S : Ċ − : Ċe + · ξh ≥0 (51)
2 ∂Ce ∂ξ h
Using Ce = F−T −1 −T −1 −T −1 −T −1
p · C·F p and its material time derivative Ċe = Ḟ p · C·F p + F p · Ċ · F p + F p · C·Ḟ p ,
the Mandel stress tensor can be expressed as:
∂Ψe

∂Ψe
= C·S = FTp · Ce · F p · 2·F−T p · · F −1
p = 2·Ce · (52)
∂Ce ∂Ce
Using Eq. (52), the non-negative of the internal dissipation can finally be defined as follows:
∂Ψ p ˙
D = :L p − · ξh ≥ 0 (53)
∂ξ h
The evolution equations for the plastic strain tensors are derived based on the principle of maximum plastic
dissipation. In addition, the normality rules give function of L p and ξ˙h as
∂f ∂f
L p = λ̇ , ξ˙h = λ̇ (54)
∂ ∂α
These equations satisfy the associative flow rule and the expression for evolution equation, respectively. Equally
ξ˙h is a work conjugate of the stress-like internal variable describing isotropic hardening α. Finally the local
dissipation inequality can be expressed with respect to the isoclinic configuration ˆ i as
ˆ :L̂ p −α :ξ˙h ≥ 0
D̂ = (55)
where
∂ fˆ ∂ fˆ
L̂ p = λ̇ , ξ˙h = λ̇ (56)
ˆ
∂ ∂α
Since the Mandel stress adopted in the formulation is symmetry and due to the fact that this stress measure is
thermodynamically conjugate to the plastic velocity gradient, only the symmetric part of the plastic velocity
gradient is adopted in the formulation. However, it should be noted that this assumption holds rigorously for
plastic isotropic but not necessarily holds for plastic anisotropic case. In this work, the assumption is made that
the plastic spin vanishes. In other words, the difference between the spin of continuum and its substructure is
set to zero; W p = 0.
Using symmetry of Mandel stress and assuming that the plastic spin is equal to zero in the isoclinic
configuration ˆ i , Eq. (55) can be rewritten as
ˆ : symL̂ p −α :ξ˙h ≥ 0
D̂ = (57)
The above expression is also equal to

ˆ :D̂ p −α :ξ˙h ≥ 0
D̂ = (58)

By using the yield function Eq. (44) in the first part of Eq. (56), the evolution of the plastic deformation D̂ p
can be expressed as follows:
T T
ˆ + ĥ ·
ĥ · ˆ +ĥT · ˆ + ĥT ·
ˆ
D̂ p = λ̇ ) (59)
ˆ : ĥ :
4 ˆ
Therefore, the inequality of dissipation energy in Eq. (58) can be expressed as follows:
⎡ T T ⎤
ˆ + ĥ ·
ĥ · ˆ +ĥT · ˆ + ĥT ·
ˆ
⎢ ⎥
D̂ = ˆ · ⎢λ̇ ) ⎥ −α : ξ˙h ≥ 0 (60)
⎣ ˆ :ĥ:
ˆ ⎦
4
Using the above identities, eventually the Clausius–Plank inequality of the second law of thermodynamics for
ˆ i as follows:
the new constitutive model can be expressed with respect to the isoclinic configuration
)
D̂ = λ̇ ˆ −α : ∂ f ≥ 0
ˆ :ĥ: (61)
∂α
The evolution equation which determines the relation between the stress and strain increments of the constitutive
relations is further defined from the consistency condition. Let first write the plastic velocity gradient in the
isoclinic configuration ˆ i as

˙ ˆ
L̂ p = F̂ p · F̂−1
p = λ̇r̂ , α (62)
ˆ
The plastic flow direction is marked by r̂ and set equals to ∂ ˆf . Further, the evolution of isotropic hardening
∂
law is given by
α̇ = λ̇Ĥ ˆ ,α (63)
where Ĥ represents the plastic modulus. By using the yield function Eq. (43), the loading–unloading condition
˙
and considering the structural tensors remain constant in the isoclinic configuration, M̂ii = 0, the consistency
condition of the new constitutive model can be defined as
f˙ˆ = fˆˆ , :
˙ˆ + fˆ · α̇ = 0
α (64)
where
∂ fˆ ˆ ∂ fˆ
fˆˆ = fα = (65)
ˆ
∂ ∂α
* T
¯ˆ =
Using 3 ˆ
ĥ ˆ to define the orthotropic yield criterion, fˆ (α) in Eq. (43) can be re written as

2
*
2 ¯
fˆ (α) = ˆ z
(66)
3
The rate of this expression α̇ can be expressed as
*
2 ˙¯
α̇ = ˆ (67)
3
˙¯ˆ
where is given by
˙¯ˆ ¯
= ĤD̂ p (68)
Combining Eqs. (66) and (67) results in *
2 ¯
α̇ = ĤD̂ p (69)
3


¯
Subsequently, simplifying 2
3 Ĥ as K̂ f and defining D̂ p = 3 λ̇,
2
the above equation can be rewritten as
*
2
α̇ = λ̇ K̂ f (70)
3
Using this equation, the consistency condition of the new constitutive model Eq. (64) can be re-expressed as
*
˙
ˆ ˆ ˙
ˆ ˆ
f = f ˆ : + f α · λ̇
2
K̂ f = 0 (71)
3
˙ ˙ˆ in the above equation and knowing D̂ = D̂−D̂ as a division between
Subsequently, using D̂e = Êe = e p
elastic and plastic rate of deformation, the consistency condition of the proposed formulation in the isoclinic
configuration ˆ i can be further derived as follows:
*
˙
ˆ ˆ ˆD
ˆ
f = f ˆ : Cel : D̂−D̂ p + f α · λ̇
2
K̂ f (72)
3
Introducing D̂ p which equals to symL̂ p = λ̇symr̂ in the above equation [9], the plastic rate parameter λ̇ can
be expressed as
ˆD
fˆˆ : C
el : D̂
λ̇ = (73)
ˆD
− fˆα · 23 K̂ f + fˆˆ : C
el : symr̂
The plastic rate parameter λ̇ can be used to explicitly define the relation between the stress and strain increments
of the new constitutive model in the isoclinic configuration ˆ i within elastic and plastic regimes using the
following equation:
˙ˆ = Cˆ D : D̂ − D̂ = Cˆ D : D̂ − λ̇symr̂ (74)
el p el
Using Eq. (73), this equation can be further written as follows:

⎛ ⎞
ˆD
˙ˆ =Cˆ D : ⎝D̂− fˆˆ : C : D̂
el el
symr̂⎠ (75)
ˆ
− fˆα · 2
K̂ f + fˆˆ : C D : symr̂
3 el
Finally, with some arrangements the relationship between the stress and strain increments within elastic and
ˆ i is concluded by the elastoplastic tangent modulus C ˆ ;
plastic regimes in
ˆ
C D
: symr̂ ⊗ fˆˆ : Cˆ D
ˆ el el
Cˆ = C
el − (76)
ˆ
− fˆα · 23 K̂ f + fˆˆ : C el
D
: symr̂
2.8 Grady failure model
The prediction of damage initiation and evolution in materials undergoing high-velocity impacts is focusing
on materials behaviour under high pressures and high strain rates such as ballistic impact, blast loading, space
debris impact of space vehicles and satellites, automobile crash, geological events. Spallation is a mode related
to material failure during high-velocity impact. One definition of this phenomenon refers to material failure
due to relations between two or more rarefaction waves [14]. It is observed that an impact on the outer surface
which is not powerful enough to cause penetration, nevertheless, could cause the formation of spalls at the
inner surface.
Spall can be defined as the ejection of fragments of a structural element from the opposite side from
which the structural element is impacted and/or impulsively loaded. Spall failure is the main failure mode
in shock-loaded metallic materials and is heavily influenced by material microstructure. Three-dimensional
characterisation can be used to determine where spall damage nucleates, grows and coalesces within the
microstructure of shock-loaded materials. It can be observed, the common approach to model spall failure

in many hydrocodes are using pressure cut-off and maximum principal stress models. Pressure cut-off and
maximum principal stress can be regarded the simplest models. Therefore, in this work, a Grady spall model
is adopted in the proposed constitutive model. This is an energy-based failure model. The material spall is
assumed when the strain energy reaches a certain level.
Generally, Grady model is used to check how much energy is required. In this Grady model, two mechanisms
are investigated which are brittle and ductile fracture. In the case of brittle fractured, the energy required is
based on the critical fracture toughness. In contrast, for the case of ductile fracture is based on work required
to reach failure strain. This procedure then results in the following spall strengths:

ps(ductile) = 2ρc02 σY εfail ductile failure (77)

ps(brittle) = 3 3ρc0 K c2 ε̇ brittle failure (78)
where
ρ = density
c0 = bulk sound speed
σY = yield stress
εfail = critical strain failure
K c = fracture toughness
ε̇ = rate of volumetric dilatation
This spall stress is calculated for each cell at each cycle, thus including the local conditions in the cell. For a
certain strain rate, there is a transition point between ductile and brittle spall. This critical strain rate, εcrit , can
be calculated from:
8B02 (σY εfail )3
ε̇crit = (79)
9ρ K c4
where
ρ = density
σY = yield stress
εfail = critical strain failure
K c = fracture toughness
B = isotropic/kinematic hardening
3 New EOS of the generalised orthotropic pressure
The proposed formulations in the preceding sections are implemented into the LLNL-DYNA3D of UTHM’s
version and named Material Type 92 (Mat92). As been mentioned, a new generalised pressure is adopted in
the proposed formulation. Therefore, few modifications required to uniquely characterise the code with the
new orthotropic pressure which are discussed in this section. In other words, a new definition of EOS had
to be introduced to the code by decoupling the volumetric parts from the full stress tensor since subroutine
f3dm92 calculates only the deviatoric part. Subroutine f3dmxx (xx refers to the material type number) of
such a material model only calculates the deviatoric part of the constitutive relation. It should be noted that
any constitutive model that uses an EOS formulation calculates the pressure component in a separate set of
subroutines.
Anyone who needs to modify the code must first learn the structure of the LLNL-DYNA3D code. Briefly,
the structure of the DYNA3D code can be divided into five main sections: input, initialise, restart, solution
and output as depicted in Fig. 3. Particularly, it is very useful to understand of the COMMON block, as well
as the way that FORTRAN stores data in arrays when learning the DYNA3D code. A good overview of the
DYNA3D code can be found in Campbell [12], De Vuyst [14] and [37]. The other available documentation
for DYNA3D is the User Manual [32] and the Theoretical Manual [25].
The first step of implementation discussed in this paper is to ensure the proposed Material Type 92 is
defined with a new EOS of the generalised pressure analysis. This is performed in subroutine matin in
the code (this subroutine is used to read in the data from the input file) at lines 80 and 90 where additional
arguments are added to include Material Type 92 and restrict the material model to work only with EOS type

Fig. 3 Basic structure of DYNA3D code
1 (Linear Polynomial) and 4 (Gruneisen), respectively. The EOS parameters are initialised in the subroutine
inse92 and stored as auxiliary variables in the subroutine blkdat. These variables are then passed to the
subroutine f3dm92 via common/aux14/, arranged after six components of the Cauchy stress tensor and the
equivalent plastic strain. This approach (passing the EOS parameters using the common block /aux14/) allows
these values to be used in other subroutines related to the proposed constitutive model (Material Type 92).
In subroutine solde (subroutine that processes all the solid elements), a specific section is added to
process solid elements for all UTHM’s material model (at line 481) as depicted in Fig. 29 in “Appendix”.
Additional subroutines called for EOS implementation of the Material Type 92 are shown in this figure. These
are the additional subroutines involved if an EOS is used by any constitutive model in the DYNA3D code. The
subroutine f3dm92 is also modified to provide the EOS formulation data. For instance, the old stress tensors
are calculated and stored in the second to seventh entry of the common block /aux11/ to pass the values to
the subroutine hieupd92. The algorithm modification in this subroutine f3dm92 is critically performed to
ensure the deviatoric stress tensor is always updated. The new generalised pressure for orthotropic materials
is completely updated at the end of this implementation. Subsequently, the new stress tensor and the auxiliary
variables of the proposed constitutive model are stored in the main database of the DYNA3D code. Similar to
the other finite element codes, the conventional DYNA3D considers only the conventional volumetric part of
the stress tensor for both isotropic and anisotropic materials. Therefore, a series of modifications are required
to reflect the formulation of the new orthotropic pressure, hence giving the DYNA3D of UTHM’s version a
unique characteristic.
The implementation of the new orthotropic pressure requires modifications on subroutines f3dm92,
hieupd92 and eqos92. It should be borne in mind that the formulation of EOS is performed in the current
configuration t . At this point, it is assumed that all of the required data have been successfully passed and
calculated in subroutine f3dm92. The first step is to calculate the old stresses that required in subroutine
hieupd92. The following deviatoric stress tensor equation Si j is used to ensure that subroutine f3dm92
calculates and updates a deviatoric stress tensor.
Si j = σi j + P̃ψi j (80)
The ψi j tensor is completely defined by the elastic stiffness properties ci j as stated in Eq. (18). The common
block /aux14/ is used to store the value of the tensor in this equation. Hence, it can be easily passed between
the Material Type 92 subroutines. To reflect the new orthotropic pressure of the proposed constitutive model,
the volumetric part of Eq. (80) is replaced by the pressure P̃EOS that is calculated from EOS formulation.
Equation (80) now can be rewritten as
Si j = σi j + P̃EOS ψi j (81)
The above equation calculates old stresses of the proposed constitutive model. They are stored in the common
block /aux11/ for usage in the subroutine hieupd92. The old pressure P0 meanwhile is stored as the first

Fig. 4 Proposed validation method
entry of this common block /aux11/. The subroutine hieupd92 is used to calculate the trial value of the
internal energy e∗ , such that
1 n n+1/2 n+1/2 n+1/2 n+1 2
e∗ = en − P V Dψ t + V n+1/2 Sn+1/2 D / t n+1/2 (82)
2
where P, V , S and D are used to define pressure, velocity, deviatoric stress and rate of deformation tensor,
respectively. Using the definition of the new orthotropic pressure, Dψ is calculated as follows
Dψ = ψ11 D11 + ψ22 D22 + ψ33 D33 (83)
The trial value of the internal energy is named epx2. It is stored in the common block /aux14/ which is
passed to the subroutine eqos92 to calculate the final internal energy. This value is used to finally update the
orthotropic pressure P̃EOS . The updated orthotropic pressure is used to update the full stress tensor at the end
of the subroutine eqos92. Eventually, this full stress tensor is transformed to the global coordinate system. It
should be noted that the proposed procedure is developed for an incremental strain formulation within explicit
time integration scheme.
4 Validation and results
Figure 4 shows the adopted validation process. The work is structured to deeply examine and validate each part
of the proposed formulation including the bookkeeping of the algorithm to deal with multiple elements analysis
(complex structure). The adopted process can be a good guidance for other researchers in this particular area.
It can be observed the validation process is divided into three main stages: analysis of a new yield surface
alignment, internal validation against relevant material models available in the code and ended up with a
validation against published experimental data of plate impact test. For the sake of brevity, only the selected
results of internal validation are included in this manuscript. To be specific, using a single element analysis,
only the results in x direction are included for each formulation. In addition, only the analysis of elastic isotropy
formulation is selected to show a good bookkeeping of the implemented algorithm using multiple elements
analysis of plate impact test. At the end of this process, a new elastoplastic spall failure constitutive model
for orthotropic materials is considered successfully implemented and validated. A cm–g–μs unit system is
adopted in the involved numerical tests.

Table 1 Elastic properties of AA7010
Elastic properties Value

Young’s modulus
Ea 70.6 GPa
Eb 71.1 GPa
Ec 70.6 GPa
Poisson’s ratio
vba 0.342
vca 0.342
vcb 0.342
Shear modulus
G bc 26.31 GPa
G ab 26.48 GPa
G ac 26.48 GPa
4.1 New yield surface alignment
The alignment of a new yield surface (ψ tensor) is defined by alpha1 for ψ11 , alpha2 for ψ22 and alpha3
for ψ33 , calculated in subroutine f3dm92 in the code. This subroutine calculates the deviatoric part of the
full stress tensor. It is very crucial to get these parameters correct. Therefore, an excel file named checking
spreadsheet.xls is developed to confirm the compliance and stiffness matrices including the value of ψ tensor
calculated by the proposed algorithm in f3dm92. The elastic properties of AA7010 published in Panov [43] that
used in this analysis is shown in Table 1. Figure 30 given in “Appendix” shows the value of ψ tensor calculated
at the nested loop 20 in subroutine f3dm92, while the value calculated using the checking spreadsheet.xls is
shown in Fig. 31. It can be clearly observed that the same value calculated for each parameter. A negligible
difference (1.49 × 10−8 ) is obtained, hence confirming the implemented algorithm.
Recalling the previous discussion, ψ tensor defines a yield surface alignment of the materials under consid-
eration. Therefore, a unique alignment of a material yield surface within the principal stress space is obtained.
As can be seen, the values of this tensor are no longer equal to identity as generalised for all materials—adopted
conventionally in this research area.
4.2 Internal validation
The second stage of the proposed validation method is a validation of elastic isotropy, elastic orthotropy and
elastoplastic orthotropy parts of the proposed constitutive model against relevant material models available
in the DYNA3D code. To be specific, a uniaxial stress and uniaxial strain test are conducted using a single
element approach. In addition, a series of multiple element analysis of plate impact test is performed at two
impact velocities to ensure a clean and efficient bookkeeping of the implemented algorithm, to get ready for a
complex analysis.
4.2.1 Finite element (FE) model for a single element analysis
Figure 5 shows the FE model used in this analysis. It should be noted that the principal directions of material
orthotropy are defined with respect to x, y and z directions of the global coordinate system.
Only x direction analysis is presented and discussed in this section since the identical setting is applied to
the analysis in y and z directions. Table 2 presents the displacement boundary conditions adopted in a uniaxial
stress and uniaxial strain tests.
Displacement load curves are prescribed for nodes 1, 2, 3 and 4, to assign tension and compression loads.
Two identical solid elements adopted in each analysis are shown in Fig. 6.
The above configuration is used to compare results of the new constitutive model with the reference material
model.
4.2.2 Finite element (FE) model for multiple element analysis
Plate impact test is used to ensure an efficient bookkeeping of the algorithm to deal with multiple elements
analysis and to check the capability to capture shockwave propagation including spall failure in orthotropic

Fig. 5 FE model for single element analysis
Table 2 Boundary conditions for a single element analysis in x direction
Node number Displacement boundary condition

Uniaxial stress Uniaxial strain
1 No constraints Constrained y and z displacements
5 Constrained x, y and z displacements Constrained x, y and z displacements
6 Constrained x displacement Constrained x, y and z displacements
Fig. 6 Solid elements used in the single element analysis
materials. Figure 7 shows the configuration used to simulate each plate impact test analysis in this work. The
model is divided into three parts of rectangular bars with 4 × 4 solid elements for its cross section (symmetric
XY plane): the PMMA block, test specimen and flyer which are modelled with 100 solid elements (12 mm in
length), 75 solid elements (10 mm in length) and 25 solid elements (2.5 mm in length), respectively. As can be
observed, the model orientation is parallel to Z axis (impact axis).
The mesh applied to the model is set to allow a 1D wave propagation along bars during the impact. A
non-reflecting boundary condition is used to define the back of the poly(methyl methacrylate) (PMMA) block.
There is a contact interface defined between the test specimen and the flyer (impactor). A time history block

Fig. 7 Configuration of the plate impact test simulation
Table 3 Aluminium alloy (AA7010) material properties
Young’s modulus Poisson’s ratio Shear modulus Bulk modulus

71.00 GPa 0.30 26.75 GPa 69.75 GPa
is prescribed to the elements at the top of PMMA bar to ensure the stress time histories is recorded during the
impact.
4.2.3 Analysis of elastic isotropy formulation
Before dealing with a complex behaviour of material orthotropy, the isotropic behaviour is first analysed using
the uniaxial strain and uniaxial stress analysis of a single element approach. The isotropic–elastic–plastic–
hydrodynamic material model named Material Type 10 (Mat10) available in the DYNA3D code is used as a
reference [32].
The same elastic properties are assigned to both Mat92 and Mat10. Yielding must be prevented to ensure
materials remain in the elastic region for the range of deformations applied. Therefore, a high value of yield
is used. Table 3 shows the elastic material properties used in this analysis. The density of aluminium is
2.81 gcm−3 . To validate the elastic isotropy formulation of the proposed constitutive model, the trial value
of the new deviatoric Mandel stress tensor is examined. Hence, the plasticity and hardening formulations are
avoided.
Uniaxial stress analysis In this analysis, the true stress versus true strain curves obtained from the uniaxial
stress test in x, y and z direction tests are plotted for both Material Type 10 and Material Type 92. The value of
Young’s modulus calculated from the stress–strain curves should be identical to the given value in the input file.
The result obtained in x direction is presented in Fig. 8, while the same trend is obtained in y and z directions.
Referring to this figure, the proposed constitutive model is capable of predicting the state of uniaxial stress for
elastic isotropic materials since the stress–strain curves for both constitutive models are identical.
Uniaxial strain analysis The material is subjected to axial stress and pressure that develop an axial strain,
without strain in the radial direction in this analysis. The confining pressure is increased to force the radial
strain back to zero in the case of the radial strain starting to increase (Thomas and M. A., 2008). In addition, the
constraints applied in the uniaxial strain test generally can avoid the stress state of the material from reaching
the strength limit to prevent failure.
In this analysis, the uniaxial strain data from both Material Type 10 and Material Type 92 are analysed
by plotting the true stress vs. true strain curves in x, y, z directions. Figure 9 shows the result obtained in x

8.0
7.0 Mat10 = 71.017x
6.0
X-Stress (GPa)
5.0
4.0 Mat92 = 71.015x

Mat10
3.0 Mat92
Mat93
2.0
1.0
0.0
0 0.02 0.04 0.06 0.08 0.1
Strain
Fig. 8 Stress–strain curves in x direction of uniaxial stress, Mat92 versus Mat10
12.0
10.0 Mat10 = 102.21x

X-Stress (GPa)
8.0
6.0 Mat10
Mat92 = 102.19x
Mat92
4.0
2.0
0.0
0 0.02 0.04 0.06 0.08 0.1
Strain
Fig. 9 Stress–strain curves in x direction of uniaxial strain, Mat92 versus Mat10
direction. The identical trend is obtained in y and z directions. Based on the results, it can be concluded the
proposed constitutive model is capable of predicting the state of uniaxial strain for elastic isotropic materials
since the stress–strain curves for both constitutive models are identical.
Multiple elements analysis As mentioned previously, only the analysis of elastic isotropy formulation is
included in this manuscript to check the bookkeeping of the implemented algorithm. The flyer and the PMMA
bars are defined with a Material Type 3 of the DYNA3D code. 1 and 10 ms−1 impact velocities are used in this
analysis. Again, the material properties given in Table 3 are used. In this analysis, the stress parallel to Z axis
in the elements at the top of the PMMA block obtained for Material Type 10 is compared to Material Type 92.
The comparison is shown in Figs. 10, 11, 12 and 13.
Generally, both constitutive models produce the identical behaviour in every aspect and an appropriate
behaviour of elasticity. It can be seen no plastic slope is generated since the increment of elastic curve is
immediately followed by a Hugoniot stress curve without producing a Hugoniot elastic limit (HEL) slope. In
addition, no plastic unloading is produced along the unloading curve due to the low-impact velocities involved.
Further, the Hugoniot stress levels produced by both constitutive models at 10 ms−1 impact velocity are
consistently higher than 1 ms−1 impact velocity. As expected, the longitudinal stresses start to increase earlier
for the case of higher-impact velocity (10 ms−1 ). Therefore, it can be concluded the proposed constitutive
model is capable of predicting the correct behaviour of the elastic isotropy formulation and produce an efficient
bookkeeping.
4.2.4 Analysis of elastic orthotropy formulation
The identical two-element model (Fig. 6) is adopted in this analysis, using the orthotropic properties given in
Table 4. Again, Material Type 92 is expected to provide the correct values of Young’s modulus in different
material directions (AOPT) as produced by Material Type 22—fibre composite with damage.
There are several material axes definitions used in the LLNL-DYNA3D code; however, only material axes
type 2 (globally orthotropic) is considered in this analysis.

3.0E-05
2.5E-05
Z-stress (x 102GPa)
2.0E-05
1.5E-05
1.0E-05
5.0E-06
0.0E+00
10 15 20 25
-5.0E-06
Time (
Fig. 10 Z-stress versus time of material type 92 at 1 ms−1 impact velocity
3.0E-05
2.5E-05
Z-stress (x 102GPa)
2.0E-05
1.5E-05
1.0E-05
5.0E-06
1.1E-19
10 15 20 25
-5.0E-06
Time (
3.0E-04
2.5E-04
Z-stress (x 102GPa)
2.0E-04
1.5E-04
1.0E-04
5.0E-05
0.0E+00
2.5 7.5 12.5 17.5
-5.0E-05
Time (
3.0E-04
2.5E-04
Z-stress (x 102GPa)
2.0E-04
1.5E-04
1.0E-04
5.0E-05
1.0E-18
2.5 7.5 12.5 17.5
-5.0E-05
Time (

Table 4 Elastic orthotropy properties for aluminium alloy
Material properties Value

Young’s modulus
Ea 70.6 GPa
Eb 71.1 GPa
Ec 70.6 GPa
Poisson’s ratio
vba 0.342
vca 0.342
vcb 0.342
Shear modulus
G bc 26.3 GPa
G ab 26.5 GPa
G ac 26.5 GPa
Fig. 14 Definition of orthotropic material axes type 2 (AOPT 2) [32]
Figure 14 represents the configuration of the orthotropic material axes type 2 (AOPT 2). The orthotropic
material axes of this option are defined by vectors (c = a × d) and (b = c × a). In this figure, vectors a, b, c
refer to axes x, z and y, respectively. In the LLNL-DYNA3D code, the user is required to define the x, y and
z components of vectors a and d in the input file:
a = ax + a y + az and b = d = dx + d y + dz (84)
For instance, if the vectors are defined as a = 0ax + 0a y + 1az and d = 0dx + 1d y + 0dz then we get the
following vectors’ orientation representing the orthotropic material axes of the orthotropic material model:
a = 0ax + 0a y + 1az (z axis of DYNA model)
b = d = 0dx + 1d y + 0dz (y axis of DYNA model)
c = 1c x + 0c y + 0cz (x axis of DYNA model) (85)
For orthotropic materials, the structural tensors can be purely orientational, such as unit vectors n1 , n2 and
n3 [33]. Strictly speaking, the principal directions of material orthotropy in the current configuration t can
be considered coincident with the directions n1 , n2 and n3 . In fact n1 , n2 and n3 represent an orthonormal
privileged frame of the material. Bearing in mind the definition of the chosen material axes type 2, various
material directions are tested in this work to show the capability of the proposed formulation to track the
evolution of the principal directions of material orthotropy—directional changes in the structural axes.
Uniaxial stress analysis As performed previously, the true stress versus true strain curves obtained from the
uniaxial stress test in x, y and z direction tests are plotted for both Material Type 22 and Material Type 92.
The results obtained in x direction (using different orientation of AOPT 2) are presented in Figs. 15, 16 and
17 and the summary of analyses performed at various orientations of AOPT 2 are given in Table 5. Referring
to the results, it can be concluded the proposed constitutive model is capable of predicting the state of uniaxial
stress for elastic orthotropic materials.

8.0
7.0 Mat22 = 70.6x
6.0
X-Stress (GPa)
5.0
4.0 Mat22
Mat92 = 70.6x
3.0 Mat92
2.0
1.0
0.0
0.00 0.02 0.04 0.06 0.08 0.10
Strain
Fig. 15 Stress–strain curves of uniaxial stress in x direction, Mat92 versus Mat22, AOPT2; a = 0ax + 0ay + 1az , d = 0dx +
1dy + 0dz
8.0
7.0 Mat22 = 71.10x
6.0
Y-Stress (GPa)
5.0
Mat92 = 71.17x
4.0 Mat22
3.0 Mat92
2.0
1.0
0.0
0.00 0.02 0.04 0.06 0.08 0.10
Strain
Fig. 16 Stress–strain curves of uniaxial stress in x direction, Mat92 versus Mat22, AOPT2;a = 1ax +0ay +0az , d = 0dx +1dy +
0dz
8.0
7.0 Mat22 = 70.60x
6.0
X-Stress (GPa)
5.0
Mat92 = 70.59x
4.0 Mat22
3.0 Mat92
2.0
1.0
0.0
0.00 0.02 0.04 0.06 0.08 0.10
Strain
Fig. 17 Stress–strain curves of uniaxial stress in z direction, Mat92 versus Mat22, AOPT2; a = 1ax + 0ay + 0az ,d = 0dx +
1dy + 0dz
Uniaxial strain analysis The analysis on elastic orthotropic of the proposed constitutive model is further
extended with the uniaxial strain analysis. The results obtained in x direction (using different orientation of
AOPT 2) are presented in Fig. 18 until Fig. 20. It can be observed the proposed constitutive model is capable
of predicting the state of uniaxial strain for elastic orthotropic materials. The summary of analyses performed
at various orientations of AOPT 2 are given in Table 6.
4.2.5 Analysis of elastic–plastic orthotropy formulation
The plasticity formulation of the proposed constitutive model is first examined at this stage. The work consists
of validations of the isotropic elastic–perfectly plastic and the orthotropic elastic–plastic with hardening. The

Table 5 Comparison results of elastic orthotropy in uniaxial stress analyses
Tested directions/AOPT 2 Young’s modulus

Material type 22 (GPa) Material type 92 (GPa)
x direction
a = 0ax + 0a y + 1az , d = 0dx + 1d y + 0dz 70.60 70.60
a = 0ax + 1a y + 0az , d = −1dx + 0d y + 0dz 71.10 71.17
a = 0ax + 1a y + 0az , d = 1dx + 0d y + 0dz 71.10 71.10
y direction
a = 0ax + 0a y + 1az , d = 0dx + 1d y + 0dz 71.10 71.17
a = 0ax + 1a y + 0az , d = 1dx + 0d y + 0dz 70.60 70.60
a = 1ax + 0a y + 0az , d = dx + 1d y + 0dz 71.10 71.17
z direction
a = 0ax + 0a y + 1az , d = 0dx + 1d y + 0dz 70.60 70.59
a = 0ax + 1a y + 0az , d = 1dx + 0d y + 0dz 70.60 70.59
a = 1ax + 0a y + 0az , d = 0dx + 1d y + 0dz 70.60 70.59
12.0
Mat22 = 110.32x
10.0
8.0
X-Stress (GPa)
6.0 Mat92 = 110.32x Mat22

Mat92
4.0
2.0
0.0
0.00 0.02 0.04 0.06 0.08 0.10
Strain
Fig. 18 Stress–strain curves of uniaxial strain in x direction, Mat92 versus Mat22, AOPT2; a = 0ax +0ay +1az , d = −1dx +0dy
+0dz
12.0
Mat22 = 109.44x
10.0
8.0
Y-Stress (GPa)
6.0 Mat92 = 109.43x Mat22

Mat92
4.0
2.0
0.0
0.00 0.02 0.04 0.06 0.08 0.10
Strain
Fig. 19 Stress–strain curves of uniaxial strain in y direction, Mat92 versus Mat22, AOPT2; a = 0ax +1ay +0az , d = −1dx +0dy
+0dz
first and second analyses are performed with mild anisotropic material (aluminium alloy) and a pronounced
anisotropic material (tantalum), respectively. A single element of uniaxial stress in x, y, z directions is adopted
in this analysis with slight modifications. First, the material is brought to yield in tension by prescribing a
displacement load curve, followed by a reversed loading to yield the material in compression.
Isotropic elastic–perfectly plastic analysis The elastic properties including Hill’s parameters are set to the
isotropic case in this analysis. Strictly speaking, the Hill’s parameters are set as F = G = H = 0.5 and
L = M = N = 1.5 to analyse reduce the Hill’s to the von Mises yield criterion. It should be noted that Hill’s

12.0
Mat22 = 110.32x
10.0
8.0
Z-Stress (GPa)
Mat92 = 110.32x
6.0 Mat22
Mat92
4.0
2.0
0.0
0.00 0.02 0.04 0.06 0.08 0.10
Strain
Fig. 20 Stress–strain curves of uniaxial strain in z direction, Mat92 versus Mat22, AOPT2;a = 1ax +0ay +0az , d = 0dx +0dy
−1dz
Table 6 Comparison results of elastic orthotropy in uniaxial strain analyses
Tested directions/orientation of AOPT 2 Young’s modulus

Material type 22 (GPa) Material type 92 (GPa)
x direction
a = 0ax + 0a y + 1az , d = −1dx + 0d y + 0dz 110.32 110.32
a = 0ax + 0a y + 1az , d = 0dx + 1d y + 0dz 109.65 109.65
a = 1ax + 0a y + 0az , d = 0dx + 0d y − 1dz 109.44 109.43
y direction
a = 0ax + 1a y + 0az , d = −1dx + 0d y + 0dz 109.44 109.43
a = 0ax + 1a y + 0az , d = 1dx + 0d y + 0dz 109.44 109.43
a = 0ax + 0a y + 1az , d = 0dx + 1d y + 0dz 110.32 110.32
z direction
a = 1ax + 0a y + 0az , d = 0dx + 0d y −1dz 110.32 110.32
a = 1ax + 0a y + 0az , d = 0dx + 1d y + 0dz 109.65 109.65
a = 0ax + 0a y + 1az , d = 0dx + 1d y + 0dz 109.44 109.43
Table 7 Aluminium alloy properties for isotropic elastic–perfectly plastic analysis

Young’s modulus Ea = Eb = Ec 70.6 GPa
Poisson’s ratio, vba = vca = vcb 0.342
Shear modulus, G bc = G ab = G ac 26.48 GPa
Yield stress in a direction σ y σy 564 MPa
Hill’s parameters R 0.5
P 1.0
Q bc 1.0
Q ba 1.0
Q ca 0.5
parameters is changed into Langford parameters R, P, Q bc , Q ba , Q ca , i.e. the input parameters used in the
DYNA3D code. The parameters are given in Table 7.
The result of this analysis in x direction is shown in Fig. 21. The results at each direction are summarised
in Table 8.
Referring to the results in Table 8, it can be seen the yield stress in x, y, z directions is identical (within the
bounds of numerical roundoff). Furthermore, the correct Young’s modulus is obtained under both tension and
compression loadings. Therefore, it can be concluded that the behaviour of the elastic–plastic isotropy (Von
Mises yield surface) can be appropriately captured by the proposed constitutive model. The identical yield
stress obtained in tension and compression tests is due to the absence of Bauschinger effect in the proposed
formulation.

8.0E-03
6.0E-03
4.0E-03
X-Stress (GPa)
2.0E-03
0.0E+00
0 0.025 0.05 0.075 0.1
-2.0E-03
-4.0E-03
-6.0E-03
-8.0E-03
Strain
Fig. 21 Stress versus strain curve, elastic–perfectly plastic in x direction of Mat92
Table 8 Summary of elastic–perfectly plastic analysis
Parameters Value
In tension In compression
x direction
Yield stress (Mpa) 563.9 564.0
Young’s modulus (GPa) 70.6 70.6
y direction
Yield stress (MPa) 564.1 564.0
z direction
Yield stress (MPa) 563.8 564.0
4.2.6 Orthotropic elastic–plastic with linear hardening analysis
The analysis related to the elastic–plastic with hardening formulation is performed at the end of the internal
validation process. The results obtained from both the proposed constitutive model and the Material Type 33
(Mat33) are compared in this analysis using the tantalum properties given in Table 9, with density 16.64 gcm−3 .
Only Grady spall failure is excluded in this analysis.
Hardening is included in addition to the setting used in the preceding analysis; again, the proposed consti-
tutive model must be able to produce the identical behaviour in tension and compression with respect to Mat33
to be validated at this stage. The material axes AOPT 2 are set to a = 0ax +0a y +1az and d = 0dx +1d y +0dz .
The result in x direction is shown in Fig. 22:
It can be seen in the above figure that Mat92 is capable of predicting the elastic–plastic with hardening
behaviours of materials with pronounced orthotropy (Tantalum), in x direction. A similar behaviour is obtained
in y and z directions where the identical value is achieved for the Young’s modulus, yield stress and the
hardening slopes.

Table 9 Tantalum material properties

Young’s modulus
Ea 191.35 GPa
Eb 195.75 GPa
Ec 208.38 GPa
Poisson’s ratio
vba 0.371
vca 0.331
vcb 0.306
Shear modulus
G bc 70.86 GPa
G ab 62.31 GPa
G ac 64.63 GPa
Yield stress in a direction
σy 172 MPa
Tangent plastic modulus in a direction
H 2.2 GPa
Hill’s parameters
R 1.5760
P 1.5760
Q bc 0.4125
Q ba 0.3553
Q ca 0.4049
0.30
0.20
0.10 Mat33
Mat92
X-Stress (GPa)
0.00
0 0.005 0.01 0.015 0.02 0.025 0.03
-0.10
-0.20
-0.30
-0.40
Strain
Fig. 22 Stress–strain curve in x direction, Mat92 versus Mat33
4.3 Validation against experimental data of plate impact test
The capability of the proposed constitutive model to predict the deformation behaviour of orthotropic materials
under high pressure and shockwave is finally examined using the experimental data of plate impact test
published in De Vuyst [14].
4.3.1 Plate impact test analysis
The Z stress in the top elements of the PMMA bar is compared against the experimental data in short transverse
and the longitudinal (rolling) directions of the specimen, performed at three different impact velocities: 234,
450 and 895 ms−1 . Table 10 shows the material properties of each parts including the Gruneisen EOS and the
Grady failure model adopted in this analysis. The flyer is prescribed as Aluminium 6082-T6. Mat10 is used to
describe the behaviour of the flyer and the PMMA blocks.
The material axes AOPT 2 is set to a = 0ax + 0a y + 1az and d = 0dx + 1d y + 0dz in these analyses. The
results obtained for each impact velocity with respect to both longitudinal and transverse directions are shown
in Figs. 23 until 28.
Referring to the results in the above figures, it can be observed that the proposed Mat92 is capable of
describing the elastic–plastic loading–unloading behaviours of the Al7010 since a close prediction is obtained
in each test. The comparison between Mat92 and the experimental data are summarised in Table 11.

Table 10 Material properties used in the plate impact test analysis
Parameters Materials
Al7010 Al6082 PMMA
Young’s modulus
Ea 70.6 GPa − −
Eb 71.1 GPa – −
Ec 70.6 GPa – –
Poisson’s ratio
vba 0.342 – –
vca 0.342 – –
vcb 0.342 – –
Shear modulus
G bc 26.31 GPa 26.8 GPa 2.3 GPa
G ab 26.48 GPa 26.8 GPa 2.3 GPa
G ac 26.48 GPa 26.8 GPa 2.3 GPa
Yield stress in a direction
σy 564 MPa 250 MPa 70 MPa
Tangent plastic modulus in a direction
H 0.13 GPa 130 MPa 300 MPa
Pressure cut-off density Hill’s parameters
pcut – 2.5 GPa –
ρ 2.81 gcm−3 2.7 gcm−3 1.18 gcm−3
R 1 – –
P 0.719 – –
Q bc 1 – –
Q ba 1 – −
Q ca 1 – –
Gruneisen parameters
C 5200 ms−1 5240 ms−1 2180 ms−1
s1 1.36 1.4 2.088
s2 0.00 0.00 −1.124
s3 0.00 0.00 0.00
2.2 1.97 0.85
a 0.48 0.48 0.00
Grady parameters
ρ 2.81 gcm−3 – –
c0 0.52 – –
εfail 0.5 – –
Kc 0.0025 – –
B ρ/c02 – –
0.80
0.70
exp
0.60
mat92
0.50
Stress (GPa)
0.40
0.30
0.20
0.10
0.00
2.5 3.0 3.5 4.0 4.5 5.0
-0.10
-0.20
Time (µs)
Fig. 23 Longitudinal stress at 234 ms−1 impact in longitudinal direction

0.8
0.7
0.6 exp
mat92
0.5
0.4
Stress (GPa)
0.3
0.2
0.1
0
2.5 3 3.5 4 4.5 5
-0.1
-0.2
Time (µs)
Fig. 24 Longitudinal stress at 234 ms−1 impact in transverse direction
1.60
1.40
1.20 exp
mat92
1.00
Stress (GPa)
0.80
0.60
0.40
0.20
0.00
1.5 2.0 2.5 3.0 3.5 4.0
-0.20
Time (µs)
1.60
1.40
1.20 exp
mat92
1.00
Stress (GPa)
0.80
0.60
0.40
0.20
0.00
1.5 2.0 2.5 3.0 3.5 4.0
-0.20
Time (µs)
In general, the shape of generated pulse shows a good agreement compared to the experimental data. The
Hugoniot elastic limit (HEL) is described by the initial slope of the longitudinal stress increment. An adequate
level of anisotropy of the material is reflected by a different HEL value in each direction. In addition, the
generated pulse and the Hugoniot stress levels are closely agreed with the experimental data to confirm the
elastic–plastic formulation including the newly implemented orthotropic pressure of the proposed constitutive
model is capable of describing the behaviour of orthotropic material undergoing finite strain deformation.

3.50
3.00
2.50 exp
mat92
2.00
Stress (GPa)
1.50
1.00
0.50
0.00
1.0 1.5 2.0 2.5 3.0 3.5
-0.50
Time (µs)
3.50
3.00
2.50
exp
2.00 mat92
Stress (GPa)
1.50
1.00
0.50
0.00
1.0 1.5 2.0 2.5 3.0
-0.50
Time (µs)
The simulated longitudinal stress at the interface between the target material and PMMA block (recorded
in the elements at the back of the specimen) for plate impact at 234 ms−1 is presented in Figs. 23 and 24. It can
be clearly seen that the tensile wave failure or spall (demonstrated by the reloading of' the longitudinal
( stress
after the first loading–unloading pulse) is not generated with a lower-impact velocity 234ms−1 . However, a
' (
clear spall criterion can be observed when higher-impact velocities 450 and 895 ms−1 are applied.
At 450 ms−1 impact velocity, a clear break in slope in the start rising part of trace (HEL) can be seen in
both traces as shown in Figs. 25 and 26. The value of the HEL shows a slight difference in each direction.
Specifically, the values in the longitudinal and transverse directions are 0.41 and 0.38 GPa, respectively. The
associated values obtained in the experiment are 0.39 and 0.33 GPa, respectively. A very close agreement is
observed with respect to Hugoniot stress levels. This confirmed the capability of the implemented orthotropic
pressure of Mat92 to capture shockwaves in orthotropic materials.
In addition, clear pullback signals (spall) are developed in both traces, measured 0.31 GPa in the longitudinal
and 0.42 GPa in the transverse direction. The values obtained using the Grady’s spall criterion are 0.21 and
0.10 GPa in the longitudinal and transverse directions, respectively. It can be observed the values are smaller
than the values in experiment. The constitutive model, however, shows sensitivity to the direction of impact.
This behaviour is depicted by the point where the spall starts. The similar behaviour is obtained experimentally.
Figures 27 and 28 show the results for 895 ms−1 impact velocity. It is clear that the shape of the pulse
including the Hugoniot stresses shows a good agreement with the experimental data in both impact directions
as summarised in Table 11. As observed in 450 ms−1 impact velocity, again the proposed constitutive model
predicts smaller pullback signals in both impact directions.
Referring to the results, the model as it stands is capable of simulating elastic–plastic, shockwave prop-
agation and spall failure in orthotropic materials of aluminium alloy AA7010. The general pulse shape, the
Hugoniot stress level and the EOS are predicted satisfactorily. A higher HEL is observed in the longitudinal

Table 11 Mat92 versus plate impact test data
Impact velocity/direction Analysis criteria

HEL (GPa) Hugoniot stress level (GPa) Pulse (ms)
234 ms−1 (longitudinal) :

Simulation 0.40 0.64 1.15
Experiment 0.39 0.65 1.65
234 ms−1 (transverse) :
direction compared to the short transverse. The Hugoniot stress levels show a very good prediction capability.
The pulse shape including the width is in line with the experimental data. The adopted Grady’s spall fail-
ure model is obviously capable of predicting spall in the materials. However, it must be emphasised that the
criterion is not capable of providing a very good accuracy to deal with a complex spall strength evolution devel-
oped at different impact velocity. Further works, therefore, are required in both experimental and constitutive
modelling aspects.
At the end of this chapter, it can be concluded that the validation of Material Type 92 is completed since the
capability to predict the deformation behaviour of orthotropic materials undergoing finite strain deformation
including shockwave propagation and spall failure has been fully examined.
5 Conclusion
A new hyperelastic–plastic constitutive model for orthotropic materials undergoing finite strain deformation
including shockwave propagation and spall failure is proposed in this manuscript. A new Mandel stress tensor
that is combined with the new stress tensor decomposition of the generalised pressure is adopted. The formula-
tion is developed in the isoclinic configuration using a multiplicative decomposition of the deformation gradient
framework and further combined with equation of states (EOSs). The expansion of orthotropic yield surface is
performed in a unique alignment of deviatoric plane within the principal stress space that is defined uniquely
by the elastic properties of orthotropic materials. The newly implemented constitutive model is systematically
validated using few reference constitutive models available in the code before finally validated against plate
impact test data. The results are satisfactory with respect to the experimental data. This achievement is a good
indication for more appropriate orthotropic constitutive models in future in order to help towards a better
understanding of the complexity of material orthotropy undergoing finite deformation.
Acknowledgements The authors wish to convey a sincere gratitude to Universiti Tun Hussein Onn Malaysia (UTHM) and the
Ministry of Higher Education (MOHE) for providing the financial means during the preparation to complete this work under
Incentive Grant Scheme for Publication (IGSP), Vot U674, and the Fundamental Research Grant Scheme (FRGS), Vot 1547,
respectively.
6 Appendix
See Figs. 29, 30 and 31.

Fig. 29 Section for UTHM material models in subroutine solde
Fig. 30 Values of ψ tensor calculated in subroutine f3dm92

Fig. 31 Values of ψ tensor calculated in checking spreadsheet.xls
References
1. Ahmadi, S.F., Eskandari, M.: Vibration analysis of a rigid circular disk embedded in a transversely isotropic solid. J. Eng.
Mech. 140(7), 04014048 (2013)
2. Ahmadi, S.F., Eskandari, M.: Rocking rotation of a rigid disk embedded in a transversely isotropic half-space. Civil Eng.
Infrastruct. J. 47(1), 125–128 (2014)
3. Anderson, C.E., Cox, P.A., Johnson, G.R., Maudlin, P.J.: A constitutive model for anisotropic materials suitable for wave
propagation computer program-II. Comput. Mech. 15, 201–223 (1994)
4. Aravas, N.: Finite-strain anisotropic plasticity and the plastic spin. Model. Simul. Mater. Sci. 2, 483–504 (1994)
5. Asay, J.R., Shahinpoor, M.: High-Pressure shock compression of solids. Springer, New York (1993)
6. Banabic, D.: Sheet Metal Forming Processes. Springer, Heidelberg (2010)
7. Barlat, F.: Crystallographic texture, anisotropic yield surface and forming limits of sheet metals. Mater. Sci. Eng. 91, 55
(1987)
8. Barlat, F., Lian, J.: Plastic behaviour and stretchability of sheet metals. Part I: a yield function for orthotropic sheets under
plane stress conditions. Int. J. Plast. 5, 51–66 (1989)
9. Belytschko, T., Liu, W.K., Moran, B.: Nonlinear Finite Elements for Continua and Structures. Wiley, Chichester (2000)
10. Boehler, J.P.: On irreducible representations for isotropic scalar functions. ZAMM 57, 323–327 (1977)
11. Butcher, B.M.: Behaviour of Dense Media Under High Dynamic Pressure, p. 245. Gordon and Breach, New York (1968)
12. Campbell, J.: Lagrangian hydrocode modeling of hypervelocity impact on spacecraft. Ph.D. thesis, Cranfield University,
Cranfield, UK (1998)
13. Davison, L., Graham, R.A.: Shock compression of solids. Phys. Rep. 55, 255–379 (1979)
14. De Vuyst, T.A.: Hydrocode Modelling of Water Impact. Ph.D. thesis, Cranfield University, Cranfield, UK (2003)
15. Drumheller, D.S.: Introduction to wave propagation in nonlinear fluids and solids. Cambridge University Press, Cambridge,
UK (1998)

16. Eidel, B., Gruttmann, F.: Elastoplastic orthotropy at finite strains: multiplicative formulation and numerical implementation.
Comput. Mater. Sci. 28, 732–742 (2003)
17. Eliezer, S., Ghatak, A., Hora, H., Teller, E.: An introduction to equations of state, theory and applications. Cambridge
University Press, Cambridge (1986)
18. Eskandari, M., Shodja, H.M., Ahmadi, S.F.: Lateral translation of an inextensible circular membrane embedded in a trans-
versely isotropic half-space. Eur. J. Mech. A Solids 39, 134–143 (2013)
19. Fountzoulas, C.G., Gazonas, G.A., Cheeseman, B.A.: Computational modeling of tungsten carbide sphere impact and pen-
etration into high-strength-low-alloy (HSLA)-100 steel targets. J. Mech. Mater. Struct. 2(10), 1965 (2007)
20. Grady, D.E.: The spall strength of condensed matter. J. Mech. Phys. Solid 36, 353–384 (1988)
21. Grady, D.E., Kipp, M.E.: Fragmentation properties of metals. Int. J. Impact Eng. 20(1–5), 293–308 (1997)
22. Gray, G.T., Bourne, N.K., Millett, J.C.F.: Shock response of tantalum: lateral stress and shear strength through the front. J.
Appl. Phys. 94(10), 6430–6436 (2003)
23. Gruneisen, E.: The State of Solid Body, NASA R19542 (1959)
24. Hill, R.: A theory of the yielding and plastic flow of anisotropic metals. Proc. R. Soc. Ser. A 193, 281–297 (1948)
25. Hallquist, J.: Theoretical manual for DYNA3D. Technical report, Lawrence Livermore National Laboratory (1983)
26. Holzapfel, G.A.: Nonlinear Solid Mechanics, A Continuum Approach for Engineering. Wiley, Chichester (2007)
27. Itskov, M.: On the application of the additive decomposition of generalized strain measures in large strain plasticity. Mech.
Res. Commun. 31, 507–517 (2004)
28. Itskov, M., Aksel, N.: A constitutive model for orthotropic elasto-plasticity at large strains. Arch. Appl. Mech. 74, 75–91
(2004)
29. Khan, A.S., Kazmi, R., Farrokh, B.: Multiaxial and non-proportional loading responses, anisotropy and modeling of Ti–
6Al–4V titanium alloy over wide ranges of strain rates and temperatures. Int. J. Plast. 23(6), 931–950 (2007a)
30. Khan, A.S., Kazmi, R., Farrokh, B., Zupan, M.: Effect of oxygen content and microstructure on the thermo-mechanical
response of three Ti–6Al–4V alloys: experiments and modeling over a wide range of strain-rates and temperatures. Int. J.
Plast. 23(7), 1105–1125 (2007b)
31. Khan, A.S., Kazmi, R., Pandey, A., Stoughton, T.: Evolution of subsequent yield surfaces and elastic constants with finite
plastic deformation. Part-I: a very low work hardening aluminum alloy (Al6061-T6511). Int. J. Plast. 25(9), 1611–1625
(2009)
32. Lin, J.I.: DYNA3D: A Nonlinear, Explicit, Three-Dimensional Finite Element Code for Solid and Structural Mechanics User
Manual. Lawrence Livermore National Laboratory, Livermore (2004)
33. Lubarda, V.A., Krajcinovic, D.: Some fundamental issues in the rate theory of damage-elastoplasticity. Int. J. Plast. 11,
763–797 (1995)
34. Man, C.: On the correlation of elastic and plastic anisotropy in sheet metals. J Elast 39(2), 165–173 (1995)
35. Mandel, J.: Plasticité Classiqueet Viscoplastié, CISM Lecture Notes. Springer, Wien (1972)
36. Meyers, M.A.: Dynamic Behaviour of Materials. Wiley, Inc., New York (1994)
37. Mohd Nor M.K.: Modelling Rate Dependent Behaviour of Orthotropic Metals. Ph.D. thesis, Cranfield University, Cranfield,
UK (2012)
38. Mohd Nor, M.K., Vignjevic, R., Campbell, J.: Modelling of shockwave propagation in orthotropic materials. Appl. Mech.
Mater. 315, 557–561 (2013a)
39. Mohd Nor, M.K., Vignjevic, R., Campbell, J.: Plane-stress analysis of the new stress tensor decomposition. Appl. Mech.
Mater. 315, 635–639 (2013b)
40. Mohd Nor, M.K., Mohamad Suhaimi, I.: Effects of temperature and strain rate on commercial aluminum alloy AA5083.
Appl. Mech. Mater. 660, 332–336 (2014)
41. Mont’ans, F.J., Bathe, K.J.: Towards a model for large strain anisotropic elasto-plasticity. In: Onate, E., Owen, R. (eds.)
Computational Plasticity, pp. 13–36. Springer, Berlin (2007)
42. Nakamachi, E., Tam, N.N., Morimoto, H.: Multi-scale finite element analyses of sheet metals by using SEM-EBSD measured
crystallographic RVE models. Int. J. Plast. 23(3), 450–489 (2007)
43. Panov, V.: Modelling of behaviour of metals at high strain rates. Ph.D. dissertation, Cranfield University, Cranfield (2006)
44. Reese, S., Vladimirov, I.N.: Anisotropic modelling of metals in forming processes. IUTAM Symposium on Theoretical
Computational and Modelling Aspects of Inelastic Media, vol. 11, pp. 175–184 (2008)
45. Schmidt, R.M., Davies, F.W., Lempriere, B.M.: Temperature dependent spall threshold of four metal alloys. J. Phys. Chem.
Solids 39(4), 375–385 (1978)
46. Schröder, J., Gruttmann, F., Löblein, J.: A simple orthotropic finite elasto-plasticity model based on generalized stress-strain
measures. Comput. Mech. 30, 48–64 (2002)
47. Schröder, J., Hackl, K.: Plasticity and Beyond: Microstructures, Crystal-Plasticity and Phase Transitions. Springer, Berlin
(2014)
48. Sinha, S., Ghosh, S.: Modeling cyclic ratcheting based fatigue life of HSLA steels using crystal plasticity FEM simulations
and experiments. Int. J. Fatigue 28(12), 1690–1704 (2006)
49. Sitko, M., Skoczeń, B., Wróblewski, A.: FCC-BCC phase transformation in rectangular beams subjected to plastic straining
at cryogenic temperatures. Int. J. Mech. Sci. 52(7), 993–1007 (2010)
50. Smallman, R.E.: Modern Physical Metallurgy, 4th edn. Butterworths, London (1985)
51. Stevens, A.L., Tuler, F.R.: Effect of shock precompression on the dynamic fracture strength of 1020 steel and 6061-T6
aluminum. J. Appl. Phys. 42(13), 5665 (1971)
52. Steinberg D.J.: Equation of State and Strength Properties of Selected Materials, Report No. UCRL-MA-106439, Lawrence
Livermore National Laboratory, Livermore, CA (1991)
53. Vignjevic, R., Bourne, N.K., Millett, J.C.F., De Vuyst, T.: 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 (2002)
54. Vignjevic, R., Campbell, J., Bourne, N.K., Djordjevic, N.: Modelling shock waves in orthotropic elastic materials. Conference
on Shock Compression of Condensed Matter, Hawaii, June (2007)

55. Vignjevic, R., Djordjevic, N., Panov, V.: Modelling of dynamic behaviour of orthotropic metals including damage and failure.
Int. J. Plast. 38, 47–85 (2012)
56. Vladimirov, I.N., Pietryga, M.P., Reese, S.: On the modelling of non-linear kinematic hardening at finite strains with appli-
cation to springback—comparison of time integration algorithms. Int. J. Numer. Methods Eng. 75(1), 1–28 (2008)
57. Wilson, L.T., Reedal, D.R., Kuhns, L.D., Grady, D.E., Kipp, M.E.: Using a numerical fragmentation model to understand the
fracture and fragmentation of naturally fragmenting munitions of differing materials and geometries. In: 19th International
Symposium of Ballistics, pp. 7–11, Interlaken, Switzerland (2001)
58. Zel’dovich, Y.B., Raizer, Y.P.: Physics of Shock Waves and High temperature Hydrodynamic Phenomena, vols. 1 and 2.
Academic Press, New York (1966)
59. Zheng, Q.S.: Theory of representations for tensor functions-A unified invariant approach to constitutive equations. Appl.
Mech. Rev. 47(11), 545 (1994)

Terms and Conditions
Springer Nature journal content, brought to you courtesy of Springer Nature Customer Service Center GmbH (“Springer Nature”).
Springer Nature supports a reasonable amount of sharing of research papers by authors, subscribers and authorised users (“Users”), for small-
scale personal, non-commercial use provided that all copyright, trade and service marks and other proprietary notices are maintained. By
accessing, sharing, receiving or otherwise using the Springer Nature journal content you agree to these terms of use (“Terms”). For these
purposes, Springer Nature considers academic use (by researchers and students) to be non-commercial.
These Terms are supplementary and will apply in addition to any applicable website terms and conditions, a relevant site licence or a personal
subscription. These Terms will prevail over any conflict or ambiguity with regards to the relevant terms, a site licence or a personal subscription
(to the extent of the conflict or ambiguity only). For Creative Commons-licensed articles, the terms of the Creative Commons license used will
apply.
We collect and use personal data to provide access to the Springer Nature journal content. We may also use these personal data internally within
ResearchGate and Springer Nature and as agreed share it, in an anonymised way, for purposes of tracking, analysis and reporting. We will not
otherwise disclose your personal data outside the ResearchGate or the Springer Nature group of companies unless we have your permission as
detailed in the Privacy Policy.
While Users may use the Springer Nature journal content for small scale, personal non-commercial use, it is important to note that Users may
not:
1. use such content for the purpose of providing other users with access on a regular or large scale basis or as a means to circumvent access
control;
2. use such content where to do so would be considered a criminal or statutory offence in any jurisdiction, or gives rise to civil liability, or is
otherwise unlawful;
3. falsely or misleadingly imply or suggest endorsement, approval , sponsorship, or association unless explicitly agreed to by Springer Nature in
writing;
4. use bots or other automated methods to access the content or redirect messages
5. override any security feature or exclusionary protocol; or
6. share the content in order to create substitute for Springer Nature products or services or a systematic database of Springer Nature journal
content.
In line with the restriction against commercial use, Springer Nature does not permit the creation of a product or service that creates revenue,
royalties, rent or income from our content or its inclusion as part of a paid for service or for other commercial gain. Springer Nature journal
content cannot be used for inter-library loans and librarians may not upload Springer Nature journal content on a large scale into their, or any
other, institutional repository.
These terms of use are reviewed regularly and may be amended at any time. Springer Nature is not obligated to publish any information or
content on this website and may remove it or features or functionality at our sole discretion, at any time with or without notice. Springer Nature
may revoke this licence to you at any time and remove access to any copies of the Springer Nature journal content which have been saved.
To the fullest extent permitted by law, Springer Nature makes no warranties, representations or guarantees to Users, either express or implied
with respect to the Springer nature journal content and all parties disclaim and waive any implied warranties or warranties imposed by law,
including merchantability or fitness for any particular purpose.
Please note that these rights do not automatically extend to content, data or other material published by Springer Nature that may be licensed
from third parties.
If you would like to use or distribute our Springer Nature journal content to a wider audience or on a regular basis or in any other manner not
expressly permitted by these Terms, please contact Springer Nature at
onlineservice@springernature.com

An Anisotropic Elastoplastic Constitutive Formulation

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

An Anisotropic Elastoplastic Constitutive Formulation

Uploaded by

Copyright:

Available Formats

Continuum Mech. Thermodyn.

M. K. Mohd Nor · N. Ma’at · C. S. Ho

An anisotropic elastoplastic constitutive formulation

Keywords Elastoplastic deformation · Shockwave propagation · Spall failure · Orthotropic materials

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

2 New constitutive formulation

2.1 Kinematics for finite strain deformations

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Fig. 1 Definition of isoclinic configuration

M̂ii = QTp Mii Q p (3)

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

2.2 New stress tensor decomposition

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

2.3 Mandel stress tensor

In general, the Mandel stress tensor can be defined as follows [35]:

C̄e = FeT · Fe (21)

τ = J · σ = det (F) ·σ (23)

¯ = FeT · τ · Fe−T = det (F) · FeT ·σ·Fe−T

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Fig. 2 ψ and δ as a vector in a principal stress space

ˆ = F̂eT · τ · F̂e−T = F̂eT · det (F) ·σ·F̂e−T

Using σi j = Pδi j +Si j , the above equation can be defined as

ˆ = det (F) · F̂eT ·σ·F̂e−T = det (F) · F̂eT · (Pδ+S) · F̂e−T

Equation (28) can be further extended as

ˆ = det (F) · F̂eT · S · F̂e−T + det (F) · F̂eT · σ ψ · ψ · F̂e−T

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

ˆ = det (F) · F̂eT · σ · F̂e−T − det (F) · F̂eT · σψ ψ · F̂e−T

= det (F) · F̂eT · S · F̂e−T (31)

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

2.5 Elastic free energy function

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

2.6 Orthotropic yield criterion

2.7 The evolution equations

where Ψ̇ is a rate of Helmholtz free energy function, and Ė can be defined as

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

The differentiating of the above equation with respect to time leads to

Substituting Eq. (50) into Eq. (49) gives

The above expression is also equal to

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Using Eq. (73), this equation can be further written as follows:

2.8 Grady failure model

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

3 New EOS of the generalised orthotropic pressure

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Fig. 3 Basic structure of DYNA3D code

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Fig. 4 Proposed validation method

Dψ = ψ11 D11 + ψ22 D22 + ψ33 D33 (83)

4 Validation and results

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Table 1 Elastic properties of AA7010

Elastic properties Value

4.1 New yield surface alignment

4.2 Internal validation

4.2.1 Finite element (FE) model for a single element analysis

4.2.2 Finite element (FE) model for multiple element analysis

Content courtesy of Springer Nature, terms of use apply. Rights reserved.

Fig. 5 FE model for single element analysis

Table 2 Boundary conditions for a single element analysis in x direction

Node number Displacement boundary condition

Fig. 6 Solid elements used in the single element analysis