Composite Structures 142 (2016) 1–6

Contents lists available at ScienceDirect

Composite Structures
journal homepage: www.elsevier.com/locate/compstruct

On the closed form expression of the Mori–Tanaka theory prediction for

the engineering constants of a unidirectional fiber-reinforced ply
Sergey G. Abaimov a,⇑, Anastasia A. Khudyakova a, Stepan V. Lomov a,b
a
Center for Design, Manufacturing, and Materials, Skolkovo Institute of Science and Technology, Skolkovo, Russia
b
Department of Materials Engineering, KU Leuven, Kasteelpark Arenberg 44, B-3001 Leuven, Belgium

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

Article history: The Mori–Tanaka theory is well-known as one of the most accurate approximations of mechanical prop-
Available online 6 February 2016 erties of composite materials in structural analysis. However, while the closed form expressions of its
predictions for elastic stiffness constants are available, so far it has lacked similar expressions for the
Keywords: engineering constants typically required in applied engineering structural analysis. In this study, we pro-
Mori–Tanaka vide a closed form expression of the Mori–Tanaka theory prediction for the engineering constants of a
Continuous unidirectional ply unidirectional fiber-reinforced ply including the expression for the transverse modulus.
Engineering constants
Ó 2016 Elsevier Ltd. All rights reserved.
Transverse modulus

1. Introduction matrix operations and knowledge of Eshelby tensor. The summary

Many studies have investigated the prediction of the unidirec-
tional fiber-reinforced ply stiffness properties on the base of
empirical or statistical micromechanical models [1,2]. In particular,
predictions of engineering constants are of special importance due
to the demand of them in applied engineering structural analysis. A
comparison of different available formulae can be found in [3–7].
While the theories in general agree on the prediction of the
longitudinal modulus of the material, the transverse modulus
presents a difficult question. The most widely applied approaches
include the Chamis hypothesis [8] and the modified rule of
mixtures [9]. These two models are unique in the sense that they
provide closed form, compact expressions suitable for engineering
purposes. On the contrary, other theories including Halpin–Tsai
semiempirical expressions [10], the composite sphere and cylinder
assemblage models [11–14], the three-phase model [14], the
Mori–Tanaka theory [15,16], the self-consistent method [17–20],
and the differential scheme [21–28] are more accurate but either
not able to provide closed form expressions for all moduli, or these
expressions are lengthy and, therefore, have limited applicability.
In particular, for the Mori–Tanaka theory it is often said that
although the calculations are direct, the analytical results for the
transverse modulus are too lengthy to be published or employed
in applied structural analysis. Expressions for Mori–Tanaka predic-
tions are given in [29,30], but these formulae involve complex
⇑ Corresponding author.
E-mail address: s.abaimov@skoltech.ru (S.G. Abaimov).
of the available so far closed form expressions for the Mori–Tanaka
theory can be found in [31] but these expressions do not include
engineering constants.
The authors strongly believe in the usefulness of closed form
expressions for the engineering constants of a material. Although
in science the necessity for such expressions (when Hill’s constants
are available) may sometimes be limited, in industry such expres-
sions are often in great demand due to their simplicity and
convenience.
In this study, we overcome this drawback publishing the
engineering constants of the unidirectional fiber-reinforced ply in
closed form. The presented results are compact but exact (as much
as the Mori–Tanaka approximation is valid). For the detailed com-
parisons of the values of effective engineering constants obtained
by means of the Mori–Tanaka method with the results of other
theories, we refer the Reader to reviews [3–5,7], to Section 5 of
our paper, and to study [6] which presents the comparison of the
finite-element homogenization for a unit cell of unidirectional
composite with regular and random placement of fibers.
In Section 2, we discuss the properties of the composite con-
stituents used to illustrate the formulae applications. In Section 3,
we list the tensors used in homogenization as well as the expres-
sions leading to the effective engineering constants of the material.
In Section 4, we apply the Mori–Tanaka approximation to the uni-
directional fiber-reinforced ply and find the effective engineering
constants of the homogenized material. In Section 5, we present
numerical comparison of the results for different composite
constituents, considered earlier in Section 2.

http://dx.doi.org/10.1016/j.compstruct.2016.02.001
0263-8223/Ó 2016 Elsevier Ltd. All rights reserved.
2 S.G. Abaimov et al. / Composite Structures 142 (2016) 1–6


2. Materials considered Ef ;2 mf ;23 þ EEff ;2;1 m2f ;12
Lf ;2233 ¼ n o; ð10Þ
ð1 þ mf ;23 Þ 1  mf ;23  2 Ef ;2 m2f ;12
E
For illustrative purposes, we discuss the homogenization proce- f ;1
dure applied to several particular composites. Mechanical proper-
ties (engineering constants) of the considered fibers and matrix as Lf ;2323 ¼ Gf ;23 ; ð11Þ
composite phases are given in Table 1, where Ef,1, Ef,2, mf,12, mf,23,
Gf,12, and Gf,23 are the engineering constants (Young’s moduli, Pois- Lf ;3131 ¼ Lf ;1212 ¼ Gf ;12 : ð12Þ
son’s ratios, and shear moduli) of the fiber in its coordinate system
Again, all other elements of the tensor, not listed above or not
(axis 1 is the fiber axis) and Em, mm, and Gm are the engineering con-
obtained from the listed above by the implied symmetry relations
stants of the isotropic matrix.
Lf,ijkl = Lf,jikl = Lf,ijlk = Lf,klij, are zero.
Thereby, we consider the matrix and glass fibers as being isotro-
Here we explicitly formulated that considered fibers are not
pic. We approximate the carbon and flax fibers as being trans-
required to be transversely isotropic in the sense that
versely isotropic.
E
Gf ;23 – 2ð1þfm;2 . Therefore, the results obtained later will correspond
f ;23 Þ

3. Utilized tensors to this general case of fiber structure and material.

For the isotropic matrix, the compliance and stiffness tensors
Let us consider a fiber. In its coordinate system (xf,1, xf,2, xf,3), we are much simpler:
assume the stiffness properties of this fiber to be orthotropic with 1
the compliance tensor Mm;1111 ¼ M m;2222 ¼ M m;3333 ¼ ; ð13Þ
Em
1
M f ;1111 ¼ ; ð1Þ mm
Ef ;1 Mm;1122 ¼ M m;1133 ¼ M m;2233 ¼  ; ð14Þ
Em
1
M f ;2222 ¼ M f ;3333 ¼ ; ð2Þ 1 þ mm
Ef ;2 Mm;2323 ¼ M m;3131 ¼ M m;1212 ¼ ; ð15Þ
2Em
mf ;12
M f ;1122 ¼ M f ;1133 ¼  ; ð3Þ Em ð1  mm Þ
Ef ;1 Lm;1111 ¼ Lm;2222 ¼ Lm;3333 ¼ ; ð16Þ
1  mm  2m2m
mf ;23
M f ;2233 ¼  ; ð4Þ Em m m
Ef ;2 Lm;1122 ¼ Lm;1133 ¼ Lm;2233 ¼ ; ð17Þ
1  mm  2m2m
1
M f ;2323 ¼ ; ð5Þ Em
4Gf ;23 Lm;2323 ¼ Lm;3131 ¼ Lm;1212 ¼ : ð18Þ
2ð1 þ mm Þ
1 If we homogenize the unidirectional fiber-reinforced ply, its
M f ;3131 ¼ M f ;1212 ¼ : ð6Þ
4Gf ;12 properties are described by the definitions similar to (1–12) in
All other elements of the compliance tensor, not listed above or the material coordinate system (x1, x2, x3); only instead of fiber
not obtained from the listed above by implied symmetry relations engineering constants Ef,1, Ef,2, mf,12, mf,23, Gf,12, and Gf,23, we should
Mf,ijkl = Mf,jikl = Mf,ijlk = Mf,klij, are zero. substitute the effective engineering constants of the material:
The inverse of the compliance tensor constitutes the stiffness D E 1
tensor
M ply
1111 ¼ eff ; ð19Þ
E1
Ef ;1 ð1  mf ;23 Þ
Lf ;1111 ¼ ; ð7Þ D E D E 1
1  mf ;23  2 Ef ;2 m2f ;12
E
f ;1
M ply ply
2222 ¼ M 3333 ¼ eff ; ð20Þ
E2


Ef ;2 1  Ef ;2 m2f ;12
E
D E D E meff
Lf ;2222 ¼ Lf ;3333 ¼ n f ;1
o; ð8Þ M ply ply
1122 ¼ M 1133 ¼  eff ; ð21Þ
12
ð1 þ mf ;23 Þ 1  mf ;23  2 Ef ;2 m2f ;12
E
f ;1
E1

Ef ;2 mf ;12 D E meff
Lf ;1122 ¼ Lf ;1133 ¼ ; ð9Þ M ply
2233 ¼  eff ;
23
ð22Þ
1  mf ;23  m2f ;12
E
2 Ef ;2 E2
f ;1

Table 1
Engineering constants of the considered composite phases.

Material Young’s modulus E, GPa Poisson ratio Shear modulus G, GPa

HM Carbon P-100 fiber [32,33] Ef,1 = 775 Ef,2 = 6.8 mf,12 = 0.22 mf,23 = 0.28 Gf,12 = 20.6 E
Gf ;23 ¼ 2ð1þfm;2f ;23 Þ ¼ 2:7
HS Carbon T650 fiber [4,32,33] Ef,1 = 243 Ef,2 = 13.8 mf,12 = 0.29 mf,23 = 0.28 Gf,12 = 23.1 E
Gf ;23 ¼ 2ð1þfm;2f ;23 Þ ¼ 5:4
E-Glass fiber [34,35] Ef = 72 mf = 0.22 E
Gf ¼ 2ð1þf mf Þ ¼ 29:5
Flax fiber [36,37] Ef,1 = 60 Ef,2 = 10 mf,12 = 0.25 mf,23 = 0.25 Gf,12 = 3.2 E
Gf ;23 ¼ 2ð1þfm;2f ;23 Þ ¼ 4 .0
Epoxy PMR-15 [4,38] Em = 3.3 mm ¼ Em
2Gm  1 ¼ 0:375 Gm = 1.2
 S.G. Abaimov et al. / Composite Structures 142 (2016) 1–6 3

D E 1 4mm  1
Mply
2323 ¼ ; ð23Þ S2233 ¼ S3322 ¼ ; ð38Þ
4Geff
23
8ð1  mm Þ

D E D E 1 mm
Mply ply S3311 ¼ S2211 ¼ ; ð39Þ
3131 ¼ M 1212 ¼ ; ð24Þ 2ð1  mm Þ
4Geff
12

D E 3  4m m
S2323 ¼ ; ð40Þ
1 ð1  m23 Þ
Eeff eff
Lply ¼ ; ð25Þ 8ð1  mm Þ
1111 Eeff
1m eff
23 2 2
eff m eff 2
12
E1
1
S3131 ¼ S1212 ¼ : ð41Þ
 eff
 4
m
E2
D E D E Eeff
2 1  ref eff 2
12 All other elements of the Eshelby tensor, not listed above or
E1
Lply  ;
ply
2222 ¼ L3333 ¼
ð26Þ not obtained from the listed above by symmetry relations
Eeff
1 þ m23 1  meff
eff
23  2 eff m12
2 eff 2
Sijkl = Sjikl = Sijlk, are zero (the symmetry relation Sijkl = Sklij is not valid
E1
for the Eshelby tensor).
D E D E
2 m12
Eeff eff
4. The Mori–Tanaka method
Lply ply
1122 ¼ L1133 ¼ eff
; ð27Þ
1m m
eff E2 eff 2
23 2 eff 12
E1 In accordance with the Mori–Tanaka theory [15,16], the homog-
  enized material stiffness tensor is approximated by
eff " #1
meff
23 þ Eref m12
E
D E Eeff
2
2 eff 2
MT
X
N X
N
L ¼ cr Lr Tr cr0 Tr0 ð42Þ
Lply
2233 ¼
 ; ð28Þ
1
eff r¼0 r 0 ¼0
1 þ m23 1  m23  2 2eff meff
eff eff E 2
12 E1
with
D E Tr ¼ ½I þ Sr ðM0 Lr  IÞ1 ð43Þ
eff
Lply
2323 ¼ G23 ; ð29Þ
and Sr being the Eshelby tensor of the inhomogeneity. In these for-
D E D E mulae, r enumerates composite phases and cr is the volume fraction
Lply
3131 ¼ Lply
1212 ¼ Geff
12 : ð30Þ of phase r. As phase r = 0 we consider the matrix while the rest of
phases is represented by fibers.
Thereby, for the effective engineering constants of the unidirec-
Applying this result for the unidirectional fiber-reinforced ply,
tional ply, we find:
we find
1
D
1
E; LMT ¼ ½ð1  f ÞLm þ f Lf Tf ½ð1  f ÞI þ f Tf  ; ð44Þ
Eeff
1 ¼ ð31Þ
Mply
1111 where f is the volume fraction of fibers. For the engineering con-
stants of the homogenized material, this provides:
1
1 ¼ fEf ;1 þ ð1  f ÞEm þ 2f ð1  f ÞZ 1 ðmf ;12  mm Þ ;
2
Eeff eff
2 ¼ E3 ¼
D E; ð32Þ Eeff eff
ð45Þ
ply
M 2222
Eeff eff
2 ¼ E3
D E
meff
12 ¼ m13
eff
¼  M ply eff
1122 E1 ; ð33Þ
¼ 1 =ð1  mm Þ
Eeff 2

eff

;
1mf ;23
1 þ mf ;23  ð1 þ mm Þ þ fZ 1 1þmm
E1 Ef ;2 eff Ef ;1
D E
1
1m2m
þ 2f eff Em Em Em
 E2 þ Ef ;2
Z2 f ;1

meff
23 ¼  Mply eff
2233 E2 ; ð34Þ ð46Þ

1 Z eff
Geff
23 ¼
D E; ð35Þ meff
12 ¼ m13 ¼ mm þ 2f ðmf ;12  mm Þð1  m2m Þ;
eff 1
ð47Þ
4 M ply
2323
Em

Eeff f Ef ;1
Geff eff
D
1
E: meff
23 ¼ 1 
2
2þ2 ð1  m2m Þ
12 ¼ G13 ¼ ð36Þ
Eeff 1  f Em
4 M ply
1212
1
! !
1 ðfEf ;1 ð1  mm Þ þ ð1  f ÞEm ð1  m12 ÞÞ
Z eff Eeff
2 2

We should emphasize here that we do not make a priori any  12  1

ð1 þ mm Þ ;
Ef ;1 Em Em
assumptions about whether the homogenized material is trans-
E2
eff ð48Þ
versely isotropic in the sense Geff
23 ¼ or not. On the contrary,
2ð1þm23 Þ
eff

the answer to this question will be obtained in the result of our Em

Geff
23 ¼ !1 ; ð49Þ
analysis.
The homogenization procedure considered is the Mori–Tanaka 2ð1 þ mm Þ þ f 1f
8ð1m2m Þ
þ Em
1
2ð1þmm Þ
Gf ;23
approximation. Essential for this approximation is the knowledge
of the Eshelby tensor. We consider fibers composing the unidirec-
tional ply to be cylindrical inhomogeneities embedded in the iso- Geff eff
12 ¼ G13
tropic matrix. In this case, the Eshelby tensor is [1,2]:  1 !
Em Gf ;12
¼ 1 þ f  4f 1 þ f þ 2ð1  f Þ ð1 þ mm Þ ;
2ð1  f Þð1 þ mm Þ Em
5  4m m
S2222 ¼ S3333 ¼ ; ð37Þ
8ð1  mm Þ ð50Þ
4 S.G. Abaimov et al. / Composite Structures 142 (2016) 1–6

Table 2
Properties of the homogenized material (the volume fraction of fibers f = 55%).

Eng. constants Eeff Eeff eff

1 , GPa 2 ¼ E3 , GPa

Approx. VU Ch CCA HT MT MRoM RL VU Ch CCA+ AD HT MT CCA- MRoM RL

(a)
P-100/PMR-15 428. 428. 428. 428. 428. 428. 7.30 5.63 5.34 5.34 5.2 4.96 5.23 5.23 4.90 4.60
T650/PMR-15 135. 135. 135. 135. 135. 135. 7.21 9.49 7.57 7.54 7.4 7.21 7.02 7.02 6.89 5.67
E-Glass/PMR-15 41.3 41.1 41.1 41.1 41.1 41.1 6.94 41.3 11.3 12.1 12. 12.5 9.89 9.89 10.9 6.94
Flax/PMR-15 34.6 34.5 34.5 34.5 34.5 34.5 6.87 7.36 6.56 6.45 6.4 6.12 6.15 6.15 5.95 5.23

Eng. constants meff

12 ¼ m13
eff
meff
23

Approx. VU MT CCA HT Ch MRoM RL MRoM Ch VU RL CCA- MT CCA+

(b)
P-100/PMR-15 0.290 0.290 0.290 0.290 0.290 0.290 0.374 0.324 0.320 0.407 0.340 0.437 0.449 0.449
T650/PMR-15 0.314 0.326 0.326 0.328 0.328 0.328 0.374 0.334 0.336 0.354 0.354 0.431 0.471 0.471
E-Glass/PMR-15 0.231 0.282 0.282 0.290 0.290 0.290 0.367 0.352 0.357 0.231 0.367 0.381 0.492 0.492
Flax/PMR-15 0.296 0.304 0.304 0.306 0.306 0.306 0.367 0.316 0.314 0.343 0.339 0.407 0.435 0.435
Eng. constants Geff Geff eff
23 , GPa 12 ¼ G13 , GPa

Approx. VU Ch MRoM CCA HT MT RL VU Ch MRoM CCA HT MT RL

(c)
P-100/PMR-15 2.00 2.02 1.85 1.82 1.80 1.80 1.72 11.9 3.98 3.80 3.50 3.50 3.50 2.49
T650/PMR-15 3.50 2.83 2.58 2.47 2.39 2.39 2.10 13.2 4.04 3.88 3.56 3.56 3.56 2.51
E-Glass/PMR-15 16.8 4.16 4.02 3.76 3.31 3.31 2.54 16.8 4.16 4.02 3.67 3.67 3.67 2.54
Flax/PMR-15 2.74 2.50 2.26 2.19 2.14 2.14 1.95 2.30 2.24 2.03 2.00 2.00 2.00 1.83

Table 3
Properties of the homogenized material (the volume fraction of fibers f = 70%).

Eng. constants Eeff Eeff eff

1 , GPa 2 ¼ E3 , GPa

Approx. VU Ch CCA HT MT MRoM RL VU Ch CCA+ AD HT MT CCA- MRoM RL

(a)
P-100/PMR-15 544. 543. 544. 543. 544. 543. 10.9 6.04 5.80 5.79 5.6 5.52 5.70 5.70 5.44 5.16
T650/PMR-15 171. 171. 171. 171. 171. 171. 10.7 10.9 9.08 9.08 9.0 8.88 8.55 8.55 8.42 7.06
E-Glass/PMR-15 51.5 51.4 51.4 51.4 51.4 51.4 9.94 51.5 16.4 18.1 21. 18.9 14.6 14.6 15.8 9.94
Flax/PMR-15 43.1 43.0 43.0 43.0 43.0 43.0 9.75 8.27 7.51 7.43 7.4 7.20 7.16 7.16 6.97 6.21

Eng. constants meff

12 ¼ m13
eff
meff
23

Approx. VU MT CCA HT Ch MRoM RL MRoM Ch VU RL CCA- MT CCA+

(b)
P-100/PMR-15 0.267 0.267 0.267 0.267 0.267 0.267 0.373 0.311 0.307 0.362 0.325 0.389 0.399 0.399
T650/PMR-15 0.305 0.314 0.314 0.316 0.316 0.316 0.372 0.322 0.323 0.324 0.341 0.389 0.425 0.425
E-Glass/PMR-15 0.226 0.260 0.260 0.267 0.267 0.267 0.360 0.342 0.346 0.226 0.360 0.336 0.465 0.465
Flax/PMR-15 0.279 0.286 0.286 0.288 0.288 0.288 0.361 0.298 0.296 0.308 0.321 0.364 0.387 0.387
Eng. constants Geff Geff eff
23 , GPa 12 ¼ G13 , GPa

Approx. VU Ch MRoM CCA HT MT RL VU Ch MRoM CCA HT MT RL

(c)
P-100/PMR-15 2.22 2.22 2.07 2.05 2.04 2.04 1.95 14.8 5.66 5.43 5.16 5.16 5.16 3.52
T650/PMR-15 4.13 3.43 3.19 3.11 3.00 3.00 2.63 16.5 5.80 5.59 5.30 5.30 5.30 3.57
E-Glass/PMR-15 21.0 6.08 5.89 5.93 4.98 4.98 3.65 21.0 6.08 5.89 5.57 5.57 5.57 3.65
Flax/PMR-15 3.16 2.90 2.69 2.63 2.58 2.58 2.35 2.60 2.52 2.34 2.32 2.32 2.32 2.13

0 1
where  
@Geff Eeff Ef ;2
!1 
2 A / Gf ;23  : ð53Þ
m2f ;12 ð1  mf ;23 Þ ð1 þ mm Þð1 þ f ð1  2mm ÞÞ
23
2 1 þ meff 2ð1 þ mf ;23 Þ
Z eff
1 ¼ 2ð1  f Þ þ ð1  f Þ þ ; 23
Ef ;1 Ef ;2 Em
Thereby, the homogenized material is transversely isotropic if
ð51Þ and only if the fibers are transversely isotropic also.

2 ¼ Ef ;2 ð3 þ f  4mm Þð1 þ mm Þ þ ð1  f ÞEm ð1 þ mf ;23 Þ:

Z eff ð52Þ
5. Results comparison
To the extent of our knowledge, the compact (but exact) expres-
sion of Eeff
2 is presented for the first time.
In this section, following the previous studies in the literature
The interesting question to address is whether the obtained [3–7], the obtained results of the Mori–Tanaka approach (further,
expressions correspond to the transversely isotropic material. It we will use abbreviation MT) are compared with the predictions
is easy to find that provided by other theories: The Voigt upper bound [39] (further,
 S.G. Abaimov et al. / Composite Structures 142 (2016) 1–6 5

Table 4
Properties of the homogenized material (the volume fraction of fibers f = 75%).

Eng. constants Eeff Eeff eff

1 , GPa 2 ¼ E3 , GPa

Approx. VU Ch CCA HT MT MRoM RL VU Ch CCA+ AD HT MT CCA- MRoM RL

(a)
P-100/PMR-15 582. 582. 582. 582. 582. 582. 13.0 6.18 5.95 5.94 5.9 5.71 5.87 5.87 5.64 5.37
T650/PMR-15 183. 183. 183. 183. 183. 183. 12.7 11.4 9.68 9.66 9.7 9.52 9.18 9.18 9.05 7.69
E-Glass/PMR-15 54.9 54.8 54.8 54.8 54.8 54.8 11.6 54.9 19.0 21.1 28. 22.1 17.1 17.1 18.4 11.6
Flax/PMR-15 45.9 45.8 45.8 45.8 45.8 45.8 11.3 8.56 7.86 7.78 8.0 7.60 7.54 7.54 7.37 6.63

Eng. constants meff

12 ¼ m13
eff
meff
23

Approx. VU MT CCA HT Ch MRoM RL MRoM Ch VU RL CCA- MT CCA+

(b)
P-100/PMR-15 0.259 0.259 0.259 0.259 0.259 0.259 0.373 0.307 0.303 0.348 0.319 0.373 0.381 0.381
T650/PMR-15 0.302 0.310 0.310 0.311 0.311 0.311 0.372 0.317 0.317 0.315 0.335 0.376 0.406 0.406
E-Glass/PMR-15 0.225 0.253 0.253 0.259 0.259 0.259 0.356 0.336 0.340 0.225 0.356 0.324 0.452 0.452
Flax/PMR-15 0.274 0.280 0.280 0.281 0.281 0.281 0.357 0.291 0.290 0.297 0.313 0.348 0.368 0.368
Eng. constants Geff Geff eff
23 , GPa 12 ¼ G13 , GPa

Approx. VU Ch MRoM CCA HT MT RL VU Ch MRoM CCA HT MT RL

(c)
P-100/PMR-15 2.29 2.28 2.16 2.14 2.12 2.12 2.04 15.8 6.51 6.26 6.02 6.02 6.02 4.09
T650/PMR-15 4.34 3.67 3.44 3.37 3.26 3.26 2.88 17.6 6.71 6.47 6.21 6.21 6.21 4.15
E-Glass/PMR-15 22.4 7.09 6.88 7.05 5.88 5.88 4.28 22.4 7.09 6.88 6.58 6.58 6.58 4.28
Flax/PMR-15 3.30 3.05 2.85 2.81 2.76 2.76 2.53 2.70 2.62 2.46 2.44 2.44 2.44 2.26

VU), the Reuss lower bound [40] (further, RL), Chamis hypothesis
[8] (further, Ch), the modified rule of mixture [9] (further, MRoM),
the Halpin–Tsai model [10] (further, HT). As an etalon, we utilize
either the exact solution of the composite cylinder assemblage
model [11–14] (further, CCA) or the finite-difference analysis
[41] (further, AD).
The comparison of the obtained results is presented in
Tables 2–4 for the particular materials considered in Section 2.
All models give accurate predictions for Eeff
1 with the exception
of RL which is not expected to be applicable for this module. It is
worth noting that the Mori–Tanaka expression (45) for Eeff 1 has a
term which makes it different from the rule of mixture. This cor-
rection is, however, small in practical cases: For all the variants
in Tables 2–4, the difference is unobservable when the results
are presented with three digits.
For Eeff
2 , the Ch, HT, MT, and MRoM approximations give close
predictions with the Ch hypothesis generally overestimating the
value of the transverse modulus and the MRoM model underesti-
compact expressions for all engineering constants. These
expressions are exact as much as the Mori–Tanaka approximation
is valid. We believe that the obtained results are not only useful for
practical engineering purposes but also are of importance when
somebody is looking for the exact expression of the transverse
modulus – they may present a hint what the exact expression
would look like.
Acknowledgments
The work reported here was funded by the Center for Design,
Manufacturing and Materials, Skolkovo Institute of Science and
Technology (Skolkovo, Russia) in the framework of the Master
Research Agreement No 335-MRA between KU Leuven and
Skoltech. S.V. Lomov is a Toray Professor (Toray Chair for Compos-
ite Materials, KU Leuven).
References

mating it. The MT theory gives the best prediction for
all models are compared with the exact CCA solution. For
MT theory prediction coincides with the CCA+ bound. For Geff
23
and Geff
12 , the closest predictions are provided by the MT and HT
methods.
6. Conclusions
While the elastic stiffness constants can be utilized in applied
structural analysis, the general practice is to apply the engineering
constants on the base of experimental studies or theoretical
homogenization.
Several empirical or statistical micromechanical models have
been proposed for the homogenization of the unidirectional
fiber-reinforced composites. However, the more accurate methods
generally do not present engineering constants in the closed form,
especially when the transverse elastic modulus is discussed. One of
these theories is the Mori–Tanaka approach which so far also has
lacked a compact form expression for the transverse modulus. In
this study, we overcome this drawback considering the most
general case of fiber properties and presenting the closed form,
m eff
12 when [1] Mura T. Micromechanics of defects in solids. Martinus Nijhoff; 1987.
m eff
23 , the [2] Nemat-Nasser S, Hori M. Micromechanics: overall properties of heterogeneous
materials. North-Holland, Amsterdam: Elsevier Science; 1993.
[3] Herakovich CT. Mechanics of fibrous composites. John Wiley & Sons; 1998.
[4] Searles K, Odegard G, Kumosa M. Micro- and mesomechanics of 8-harness
satin woven fabric composites: I – evaluation of elastic behavior. Composites A
2001;32:1627–55.
[5] Younes R, Hallal A, Fardoun F, Chehade FH. Comparative review study on
elastic properties modeling for unidirectional composite materials. In: Hu N,
editor. Composites and their properties. Croatia: InTech; 2012. p. 391–408.
[6] Melro AR, Camanho PP, Pinho ST. Influence of geometrical parameters on the
elastic response of unidirectional composite materials. Compos Struct
2012;94:3223–31.
[7] Aboudi J, Arnold SM, Bednarcyk BA. Micromechanics of composite materials: a
generalized multiscale analysis approach. Elsevier; 2013.
[8] Chamis CC. Simplified composite micromechanics equations for hydral,
thermal and mechanical properties. NASA; 1983. TM-83320.
[9] Hopkins DA, Chamis CC. A unique set of micromechanical equations for high
temperature metal matrix composites. Testing technology of metal matrix
composites. Philadelphia: ASTM; 1988. STP964, p. 159–76.
[10] Halpin JC, Tsai SW. Effects of environmental factors on composite
materials. Air Force Materials Laboratory; 1969. TR-67-423.
[11] Hashin Z. The elastic moduli of heterogeneous materials. J Appl Mech
1962;29:143–50.
[12] Hashin Z, Rosen BW. The elastic moduli of fiber-reinforced materials. J Appl
Mech 1964;31:223–32.
[13] Hill R. Theory of mechanical properties of fibre-strengthened materials: I.
Elastic behavior. J Mech Phys Solids 1964;12:199–212.
6 S.G. Abaimov et al. / Composite Structures 142 (2016) 1–6
[14] Christensen RM, Lo KH. Solutions for effective shear properties in three phase
sphere and cylinder models. J Mech Phys Solids 1979;27:315–30.
[15] Mori T, Tanaka K. Average stress in matrix and average elastic energy of
materials with misfitting inclusions. Acta Metall 1973;21:571–4.
[16] Benveniste Y. A new approach to the application of Mori–Tanaka’s theory in
composite materials. Mech Mater 1987;6:147–57.
[17] Kröner E. Berechnung der elastischen Konstanten des Vielkristalls aus den
Konstanten des Einkristalls. Z Phys 1958;151:504–18.
[18] Budiansky B. On the elastic moduli of some heterogeneous materials. J Mech
Phys Solids 1965;13:223–7.
[19] Hill R. A self-consistent mechanics of composite materials. J Mech Phys Solids
1965;13:213–22.
[20] Hashin Z. Assessment of the self consistent scheme approximation:
conductivity of particulate composites. J Compos Mater 1968;2:284–300.
[21] Roscoe R. The viscosity of suspensions of rigid spheres. Brit J Appl Phys
1952;3:267–9.
[22] Roscoe R. Isotropic composites with elastic or viscoelastic phases: general
bounds for the moduli and solutions for special geometries. Rheol Acta
1973;12:404–11.
[23] Boucher S. On the effective moduli of isotropic two-phase elastic composites. J
Compos Mater 1974;8:82–9.
[24] McLaughlin R. A study of the differential scheme for composite materials. Int J
Eng Sci 1977;15:237–44.
[25] Cleary MP, Chen I-W, Lee S-M. Self-consistent techniques of heterogeneous
media. J Eng Mech Div ASCE 1980;106:861–87.
[26] Norris AN. A differential scheme for the effective moduli of composites. Mech
Mater 1985;4:1–16.
[27] Hashin Z. The differential scheme and its application to cracked materials. J
Mech Phys Solids 1988;36:719–34.
[28] Nemat-Nasser S, Hori M. Elastic solids with microdefects. Micromechanics and
inhomogeneity - the T Mura 65th anniversary volume. New York: Springer;
1990. p. 297–320.
[29] Sevostianov I, Kachanov M. Explicit cross-property correlations for anisotropic
two-phase composite materials. J Mech Phys Solids 2002;50:253–82.
[30] Sevostianov I, Yilmaz N, Kushch V, Levin V. Effective elastic properties of
matrix composites with transversely-isotropic phases. Int J Solids Struct
2005;42:455–76.
[31] Chen T, Dvorak GJ, Benveniste Y. Mori–Tanaka estimates of the overall elastic
moduli of certain composite materials. J Appl Mech 1992;59:539–46.
[32] Goggin PR. The elastic constants of carbon fibre. J Mater Sci 1973;8:233–44.
[33] Wagoner G, Bacon R. Elastic constants and thermal expansion coefficients of
various carbon fibres. In: 19th biennial conf on carbon. Penn State University:
The American Carbon Society; 1989. p. 296–7.
[34] Barbero EJ. In: Introduction to composite materials design. Taylor & Francis;
1999.
[35] E-Glass Fibre: AZoM; 2015. <www.azom.com>.
[36] Kalia S, Kaith BS, Kaur I. Cellulose fibers: bio- and nano-polymer
composites. Springer; 2011. p. 737.
[37] Flax and hemp fibres: a natural solution for the composite industry: JEC
Composites; 2012.
[38] Mirzadeh F, Reifsnider KL. Micro-deformations in C3000/PMR15 woven
composite. J Compos Mater 1992;26:185–205.
[39] Voigt W. Ueber die Beziehung zwischen den beiden Elastizitätskonstanten
isotroper Körper. Ann d Phys 1889;274:573–87.
[40] Reuss A. Berechnung der Fließgrenze von Mischkristallen auf Grund der
Plastizitätsbedingung für Einkristalle. ZAMM 1929;9:49–58.
[41] Adams DF, Doner DR. Transverse normal loading of a unidirectional composite.
J Compos Mater 1967;1:152–64.