Introduction To The Mechanics and Design of Undamageable Materials
Introduction To The Mechanics and Design of Undamageable Materials
Introduction To The Mechanics and Design of Undamageable Materials
net/publication/258140543
CITATIONS READS
16 1,253
2 authors:
Some of the authors of this publication are also working on these related projects:
All content following this page was uploaded by Peter I. Kattan on 21 July 2015.
Published by:
http://www.sagepublications.com
Additional services and information for International Journal of Damage Mechanics can be found at:
Subscriptions: http://ijd.sagepub.com/subscriptions
Reprints: http://www.sagepub.com/journalsReprints.nav
Permissions: http://www.sagepub.com/journalsPermissions.nav
What is This?
Article
International Journal of Damage
Mechanics
0(0) 1–13
Introduction to the mechanics ! The Author(s) 2012
Reprints and permissions:
and design of undamageable sagepub.co.uk/journalsPermissions.nav
DOI: 10.1177/1056789512446518
materials ijd.sagepub.com
Abstract
In this work proposed damage variables that are examined in the framework of continuum damage
mechanics are used to illustrate the new concept of an undamageable material. This type of material is
derived theoretically in this work with exact nonlinear stress–strain relationships. Such a material may
undergo deformation with zero damage, i.e. maintain a zero value for the damage variable in the context
of continuum damage mechanics. The theoretical formulation is presented in the hope that in the future,
the manufacturing of such a material may be realized. The proposed undamageable material is a type of
hyperelastic material similar to the Mooney–Rivlin material. It is hoped that this work will open the way
for new areas of research in materials science.
Keywords
Damage, undamageable, energy equivalence, higher-order energy, scalar variable
Introduction
In order to introduce the proposed undamageable material, it is necessary first to review some basic
issues of damage mechanics. The concept of effective stress for uniaxial tension was first introduced
by Kachanov (1958) and Rabotnov (1969). It has been argued (Lemaitre, 1984) that the assumption
of isotropic damage is sufficient to give good predictions of the load-carrying capacity, the number
of cycles or the time to local failure in structural components. However, the development of aniso-
tropic damage has been confirmed experimentally (Chow and Wang, 1987; Lee et al., 1985) even if
the virgin material is isotropic. For the case of isotropic damage mechanics, the damage variable is
scalar and the evolution equations are easy to handle (Kattan and Voyiadjis, 2001a, 2001b;
Lee et al., 1985; Voyiadjis and Kattan, 2006a).
Department of Civil and Environmental Engineering, Louisiana State University, Baton Rouge, LA, USA
Corresponding author:
George Z Voyiadjis, Department of Civil and Environmental Engineering, Louisiana State University, Baton Rouge, LA 70803, USA
Email: voyiadjis@eng.lsu.edu
(a) T (b) T
Remove Both
Voids and Cracks
A φ A
σ σ
T
T
Assume is the second-rank Cauchy stress tensor and the corresponding effective stress tensor
applied to a fictitious state of the material which is totally undamaged, i.e. all damage in this state
has been removed (Figure 1). This state is assumed to be mechanically equivalent to the actual
damaged state of the material. In this regard, the hypothesis of elastic energy equivalence is usually
used. The authors have shown the microstructural link of damage mechanics by characterizing
micro-cracks through the use of fabric tensors (Voyiadjis and Kattan, 2006). Damage in micro-
and mesomechanics of composite materials has also been extensively investigated (Lubineau, 2010;
Lubineau and Ladeveze, 2008).
In continuum damage mechanics, usually a phenomenological approach is adopted. In this
approach, the most important concept is that of the representative volume element (RVE). The
discontinuous and discrete elements of damage are not considered within the RVE; rather their
combined effects are lumped together through the use of a macroscopic internal variable. In this
way, the formulation may be derived consistently using sound mechanical and thermodynamic
principles (Doghri, 2000; Hansen and Shreyer, 1994; Luccioni and Oller, 2003).
In this work various new proposed damage variables are examined and used to introduce the new
concept of undamageable material. The proposed damage variables are scalar in nature to account for
the process of isotropic damage in materials. A forthcoming work by the authors examines other
proposed damage variables of tensorial nature to account for anisotropic damage. The undamageable
material may be theoretically envisaged to undergo deformation while maintaining a zero state of
damage, i.e. a zero value for the damage variable in the context of continuum damage mechanics. The
stress–strain relationships governing these hypothetical undamageable materials are derived in detail.
This theoretical formulation is presented in the hope that some time in the future, the manufacturing
technology will be developed to produce such types of materials and make their realization possible.
It is hoped that this work will open the way for new areas of research in materials science.
A A
1 ¼ ð1Þ
A
where A is the cross-sectional area in the damaged configuration while A is the cross-sectional area
in the fictitious configuration with A 4 A. It is clear that when a body is undamaged, i.e. when
A ¼ A, then 1 ¼ 0.
The value
The stress in the fictitious configuration is called the effective stress and is denoted by .
of the effective stress may be obtained using the relation A ¼ A where is the stress in the
damaged configuration. Therefore, using this relation along with the definition in equation (1), one
obtains
¼ ð2Þ
1 1
It should be mentioned that the equilibrium condition in the paragraph above reflects a mean-
field type of assumption on the stress redistribution (uniform over the resistive section) and therefore
appears to be appropriate only in the dilute damage regime, away from the stress–strain peak where
cooperate effects dominate and damage localization takes place. This holds true especially for the
tensorial case. However, the tensorial case is beyond the scope of this work.
In order to compute the effective elastic modulus E in this case, we may use the hypothesis of
elastic energy equivalence where the elastic strain energy is assumed to be equal in both configur-
ations (Sidoroff, 1981). This will be illustrated shortly.
The second scalar damage variable ‘1 is defined in terms of the reduction in the elastic modulus
as follows
E E
‘1 ¼ ð3Þ
E
where E is the elastic modulus in the damaged state while E is the effective elastic modulus (in the
fictitious state) with E 4 E (Figure 2). This damage variable was used recently by Celentano et al.
(2004), Nichols and Abell (2003) and Nichols and Totoev (1999). Voyiadjis (1988) used a similar
relation but in the context of elasto-plastic deformation. The reader is also referred to Voyiadjis and
Kattan (2009) for more details. The definition of the alternative damage variable of equation (3) may
be re-written in the following more appropriate form
E ¼ Eð1 þ ‘1 Þ ð4Þ
It is clear from the definition in equation (3) that ‘1 ¼ 0 when the body is undamaged, i.e. when
E ¼ E.
Using the hypothesis of elastic energy equivalence we assume the complementary elastic strain
2
energy (2E ) to be equal in both configurations, i.e.
2 2
¼ ð11Þ
2E 2E
Substituting for from equations (2) into equation (11) and simplifying, we obtain
E
E ¼ ð12Þ
ð1 1 Þ2
Next, in order to derive the relation between the two damage variables 1 and ‘1 , we equate
equations (4) and (12) to obtain
1
‘1 ¼ 1 ð13Þ
ð1 1 Þ2
The above expression defines the exact relation between the two damage variables ‘1 and 1 in the
case of the hypothesis of elastic energy equivalence. The form given in equation (13) above is much
simpler and easier to evaluate than the form that appeared in equation (13) in the paper by Voyiadjis
and Kattan (2009). The relation in equation (13) may be re-written in the following form
1
1 ¼ 1 pffiffiffiffiffiffiffiffiffiffiffiffiffi ð14Þ
1 þ ‘1
Based on equations (13) and (14), Table 1 can be generated which shows the minimum and
maximum values of each damage variable.
1 2 1 1 2 1 2
" ¼ 2 " or " ¼ " ð15aÞ
2 2 2 2
In this case, we can easily show that using ‘1 ¼ EE
E will yield the relation
p
3
ffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 1 þ ‘1 ð15cÞ
If we still postulate another new hypothesis of higher order energy equivalence in the form
1 2 2 1 2 2
" ¼ " ð16aÞ
2 2
In this case, we can easily show that using ‘1 ¼ EE
E will yield the relation
p
4
ffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 1 þ ‘1 ð16cÞ
In this case, we can easily show that using ‘1 ¼ EE
E will yield the general relation
p
n
ffiffiffiffiffiffiffiffiffiffiffiffiffi
¼ 1 þ ‘1 ð17bÞ
We now plot several curves on the same graph paper to show the relations between the ratio of
the stresses and ‘1 using equations (15c), (16c) and (17b) (Figure 3). It is clear that for the limiting
case when n ! 1, the curve has a constant value at 1. Note that the lower curves appearing in
Figure 3 are for larger values of n. What does this limiting case signify?
We pwill explain
ffiffiffiffiffiffiffiffiffiffiffiffi
ffi the above results using the formulas derived for ‘1 . We start with the formula
¼ n 1 þ ‘1 of equation (17b) which was derived in the previous paragraphs. We will study the
case when n ! 1. In this case, we get the following
p ffiffiffiffiffiffiffiffiffiffiffiffiffi 1 1
¼ n
1 ‘1 ¼ ð1 ‘1 Þn ¼ ð1 ‘1 Þ1 ¼ ð1 ‘1 Þ0 ¼ 1 ¼ ð18Þ
Therefore we get ¼ irrespective of the value of the damage variable ‘1 . This means that in this
limiting case, the material remains totally undamaged irrespective of the value of the damage vari-
able. Of course, this is a hypothetical case as it cannot be reached physically. However, it gives rise to
the following new issue. Does a material exist or can be manufactured which remains totally undam-
aged during the deformation process no matter what the load may be. That will be a seminal
achievement of damage mechanics and materials science in general – to design a new type of material
that cannot be damaged at all. The above equation will provide some guidelines in this respect.
As the limiting case when n ! 1 cannot be reached physically, then we can approach the con-
cept of an undamageable material by utilizing a very high value for exponent n. Since the value of n
depends on the type of energy form used, then we need to investigate these proposed types of higher
order energy forms in great detail. In fact these proposed forms were investigated in detail by
Voyiadjis and Kattan (2012). Table 2 provides a summary of the valid forms and resulting stress–
strain relationships that were derived by Voyiajdis and Kattan (2012). The detailed derivation of
these relations is illustrated in ‘Higher order strain energy forms’ section.
The valid results are shown in a graph in Figure 4 for various values of n.
Let us investigate in more detail the higher order strain energy forms with powers of " only
as shown in Table 2. The general form of these types of energy forms are of the form U ¼ 12 "n .
The corresponding nonlinear stress–strain relation for this general form is given by
1 ¼ E"
U ¼ "
2
1 1 2="
U ¼ "2 ¼E e
2 "2
1 1 1="2
U ¼ "3 ¼E e
2 "3
1 1 2=½ðn1Þ"ðn1Þ
U ¼ "n ; ¼E e
2 "n
n ¼ 1, 2, 3, . . . :
ðn1Þ
¼ E "1n e2=½ðn1Þ" as seen from Table 2. It is seen that this general form satisfies the initial
conditions ¼ 0 when " ¼ 0 because the limit of the expression for the stress approaches zero as
the strain approaches zero. Thus, the higher order strain energy forms with powers of strain only are
valid and can be used in our discussion of undamageable materials. Note also that the special valid
cases appearing at the top of Table 2 can be deduced directly from the general solution.
The stress–strain curves appearing in Figure 4 are similar to those of rubber materials. Arruda
and Boyce (1993) conducted extensive investigation of the constitutive equations of elastic rubber
materials. Figures 5, 6 and 7 show the stress–strain curves obtained for different types of rubber
materials based on the work of Arruda and Boyce (1993).
Thus, in order to design an undamageable material, the stress–strain relation of this hypothet-
ical material will have to follow the highly nonlinear stress–strain relation given in the last row of
Table 2 when we take the limit as the value of n goes to infinity. It is hoped that sometime in the
future, the manufacturing technology will advance to such a stage that such a material can be
realized.
It is clear from Figure 4 that the derived stress–strain relationship for undamageable materials is
distinctly different from that for ordinary materials. Clearly, there is no degradation of the elastic
modulus. Also, it is seen that the values of the stress remain identically zero until the strain reaches a
certain critical value. As we approach the hypothetical undamageable material (for higher values of
n in Figure 4), it is clearly seen that the stress–strain curve remains almost horizontal indicating
absolutely no stress or damage in the material. The value of the stress becomes nonzero only after a
considerable amount of strain has accumulated. At these high values of strain, it is also seen that the
modulus of elasticity actually strengthens instead of degrading like in ordinary materials. This is the
essence of the proposed undamageable materials.
The relation of the presented concept of undamageable materials to the Ideal Locking Material
should be mentioned. Prager (1957) proposed the concept of an ideal locking material that was later
formalized by Noble and Sewell (1972). It is clear that the ideal locking material may be obtained as
the limit in Figure 4 as n tends to infinity.
Substituting the above general stress–strain relation (given in the previous paragraph) in the
expression for U, we obtain
Z
U ¼ E f ð"Þ d" ð20Þ
Taking the derivative of both sides, we obtain the following formula for fð"Þ
dU
f ð"Þ ¼ ð21Þ
E
Thus, the above formula can be used to obtain the specific function of the sought nonlinear
stress–strain relation for the material. We will illustrate this with one example using the higher order
strain energy form U ¼ 12 "2 .
Let us start with the higher order strain energy form U ¼ 12 "2 . Substitute ¼ E fð"Þ in the above
formula for U to obtain
1
U ¼ Ef ð"Þ "2 ð22Þ
2
We now take the derivative of the above formula to obtain
1
dU ¼ E f 0 ð"Þ "2 þ E fð"Þ " ð23Þ
2
Substitute the above formula into equation (21) to obtain
1
fð"Þ ¼ f 0 ð"Þ "2 þ fð"Þ " ð24Þ
2
1 0
f ð"Þ "2 þ fð"Þ " fð"Þ ¼ 0 ð25Þ
2
Next, we solve the above differential equation to obtain the desired nonlinear relation between
stress and strain as follows. The solution to the above ordinary differential equation is obtained as
follows
1 2="
f ð " Þ ¼ eC e ð26Þ
"2
1 2="
¼ E eC e ð27Þ
"2
We can assume the constant C ¼ 0 to obtain the following nonlinear stress–strain relation
1 2="
¼E e ð28Þ
"2
Looking at the above stress–strain relation, it may appear that the stress approaches infinity at
the initial condition when the strain is zero but this is not the case. It is seen that the limit of the
above expression for the stress approaches zero as the strain approaches zero. Thus, the initial
conditions of zero strain and zero stress are satisfied.
Thus, it is seen that for the higher order strain energy form 12 "2 , the corresponding stress–strain
relation is nonlinear and is given by the expression ¼ E "12 e2=" .
The above procedure can be repeated for the other proposed higher order strain energy forms
to obtain their corresponding nonlinear stress–strain relations. These results are summarized
in Table 2.
Conclusion
The concept of an undamageable material is presented in this work. The stress–strain relationships
for such materials are shown to be based on proposed higher order strain energy forms. The for-
mulation is presented within the framework of continuum damage mechanics. It is hoped that the
equations presented along with the theoretical development will aid in the design of such hypothe-
sized materials.
It was proposed that a generalization of the hypothesis of elastic energy equivalence would lead to
the relation ¼ , irrespective of the value of the damage variable. This means that in this limiting
case, the material remains totally undamaged irrespective of the value of the damage variable.
Of course, this is a hypothetical case as it cannot be reached physically. However, it gives rise to
the following new issue. Does a material exist or can be manufactured which remains totally undam-
aged during the deformation process no matter what the load may be. That will be a seminal
achievement of damage mechanics – to design a new type of material that cannot be damaged
at all. In this work, we provide a glimpse or just a hint on how such a material may be designed.
The equations in ‘The analysis and design of an undamageable material’ section will provide some
guidelines in this respect. The proposed undamageable material is a type of hyperelastic material
similar to the Mooney–Rivlin material. It is hoped that this work will open the way for new areas of
research in damage mechanics.
The formulation in this work is carried out for the simple uniaxial case. Attempting to generalize
this formulation to the three-dimensional state of deformation and damage give rise to enormous
difficulties. The reason is that the higher order energy forms are not easily written mathematically
for the three-dimensional case. However, the authors are attempting to resolve this issue in a forth-
coming paper and illustrate that the concept of undamageable materials is also valid for general
states of deformation and damage.
Funding
This research received no specific grant from any funding agency in the public, commercial, or not-for-profit
sectors.
References
Arruda EM and Boyce MC (1993) A three-dimensional constitutive model for the large stretch behavior of
rubber elastic materials. Journal of the Mechanics and Physics of Solids 41(2): 389–412.
Celentano DJ, Tapia PE and Chaboche J-L (2004) Experimental and numerical characterization of damage
evolution in steels. In: Buscaglia G, Dari E and Zamonsky O (eds) Mecanica Computacional. Vol. XXIII,
Bariloche, Argentina: ONERA-TP-04-229.
Doghri I (2000) Mechanics of Deformable Solids: Linear and Nonlinear, Analytical and Computational Aspects.
Germany: Springer-Verlag.
Hansen NR and Schreyer HL (1994) A thermodynamically consistent framework for theories of elastoplasticity
coupled with damage. International Journal of Solids and Strucutres 31(3): 359–389.
Kachanov L (1958) On the creep fracture time. Izv Akad, Nauk USSR Otd Tech 8: 26–31 (in Russian).
Kattan PI and Voyiadjis GZ (1990) A coupled theory of damage mechanics and finite strain elasto-plasticity –
Part I: Damage and elastic deformations. International Journal of Engineering Science 28(5): 421–435.
Kattan PI and Voyiadjis GZ (1993) A plasticity-damage theory for large deformation of solids – Part II:
Applications to finite simple shear. International Journal of Engineering Science 31(1): 183–199.
Kattan PI and Voyiadjis GZ (2001a) Decomposition of damage tensor in continuum damage mechanics.
Journal of Engineering Mechanics, ASCE 127(9): 940–944.
Kattan PI and Voyiadjis GZ (2001b) Damage Mechanics with Finite Elements: Practical Applications with
Computer Tools. Germany: Springer-Verlag.
Ladeveze P, Poss M and Proslier L (1982) Damage and fracture of tridirectional composites. In: Progress in
Science and Engineering of Composites: Proceedings of the Fourth International Conference on Composite
Materials, Tokyo, Japan, Vol. 1, pp. 649–658.
Ladeveze P and Lemaitre J (1984) Damage effective stress in quasi-unilateral conditions. In: The 16th
International Congress of Theoretical and Applied Mechanics, Lyngby, Denmark.
Lubineau G (2010) A pyramidal modeling scheme for laminates – identification of transverse cracking.
International Journal of Damage Mechanics 19(4): 499–518.
Lubineau G and Ladeveze P (2008) Construction of a micromechanics-based intralaminar mesomodel, and
illustrations in ABAQUS/standard. Computational Materials Science 43(1): 137–145.
Lee H, Peng K and Wang J (1985) An anisotropic damage criterion for deformation instability and its appli-
cation to forming limit analysis of metal plates. Engineering Fracture Mechanics 21: 1031–1054.
Luccioni B and Oller S (2003) A directional damage model. Computer Methods in Applied Mechanics and
Engineering 192: 1119–1145.
Nichols JM and Abell AB (2003) Implementing the degrading effective stiffness of masonry in a finite element
model. In: North American Masonry Conference, Clemson, South Carolina, USA.