Professional Documents
Culture Documents
Analysis of Composite Laminates Theories and Their Applications 1St Edition Dinghe Li Full Chapter PDF
Analysis of Composite Laminates Theories and Their Applications 1St Edition Dinghe Li Full Chapter PDF
https://ebookmass.com/product/theories-of-development-concepts-
and-applications-6th-edition-william-crain/
https://ebookmass.com/product/applied-analysis-of-composite-
media-analytical-and-computational-approaches-drygas/
https://ebookmass.com/product/damage-modeling-of-composite-
structures-strength-fracture-and-finite-element-analysis-1st-
edition-pengfei-liu/
https://ebookmass.com/product/applied-analysis-of-composite-
media-analytical-and-computational-approaches-piotr-drygas/
Metal Oxide-Based Nanofibers and Their Applications
Esposito V.
https://ebookmass.com/product/metal-oxide-based-nanofibers-and-
their-applications-esposito-v/
https://ebookmass.com/product/inorganic-and-composite-fibers-
production-properties-and-applications-kyosev/
https://ebookmass.com/product/dynamics-of-molecular-excitons-
theories-and-applications-nanophotonics-1st-edition-seogjoo-j-
jang/
https://ebookmass.com/product/theories-of-development-concepts-
and-applications-6th-edition-ebook-pdf/
https://ebookmass.com/product/fascinating-fluoropolymers-and-
their-applications-1st-edition-ameduri-bruno-ed/
ANALYSIS OF
COMPOSITE LAMINATES
ANALYSIS OF
COMPOSITE LAMINATES
Theories and Their Applications
DINGHE LI
College of Aeronautical Engineering
Civil Aviation University of China
Tianjin, China
Elsevier
Radarweg 29, PO Box 211, 1000 AE Amsterdam, Netherlands
The Boulevard, Langford Lane, Kidlington, Oxford OX5 1GB, United Kingdom
50 Hampshire Street, 5th Floor, Cambridge, MA 02139, United States
Copyright © 2022 China Science Publishing & Media Ltd. Published by Elsevier Inc. All rights reserved.
ISBN: 978-0-323-90804-7
I have many people to thank. I am grateful to Professor Xiong Zhang for academic guid-
ance during my doctoral study at Tsinghua University and his constructive suggestions
about this book, which aided me in writing the text. I also appreciate Professor Jianxin
Xu and Professor Guanghui Qing for the academic guidance in my early research at
the Civil Aviation University of China. I am grateful to Professor Yan Liu and Profes-
sor Xinming Qiu for their academic guidance during my doctoral study at Tsinghua
University. I wish to thank the members of my research group on computational com-
posite mechanics at the Civil Aviation University of China. I received valuable assistance
from my M.A. students including Ruipeng Wang, Feng Zhang, Jiuyue Yang, Dong Xu,
Zhengguang Xiao, Wukui Shan, Liangyi Li, Zhengming Wang, Shuo Ma, and Zhaoxin
Yun. I express my deepest appreciation to Zhengming Wang for her assistance in editing
the text. Finally, I wish to express my gratitude and apologies to my family, Li Zhang
and Yinguang Li, for enduring this past year when much of my time and energy should
have been devoted to them rather than to this book.
v
Contents
Biography xv
Preface xvii
Acronyms xix
vii
viii Contents
Index 511
This page intentionally left blank
Biography
xv
Acronyms
AM analytical method
C2 computational continua
CLPT classical laminated plate theory
CCC carbon-carbon composite (carbon matrix)
CDM continuum damage mechanics
CTDM central difference method
CFRP carbon fiber reinforced plates
CMC ceramic matrix composite
CUF Carrera’s unified formulation
CUC computational unit cell
DCB double cantilever beam
DDZM discrete damage zone model
DLWM discrete layerwise methods
DNS direct numerical simulation
DOF degree-of-freedom
EST equivalent single-layer theory
IIM interaction integral method
ILWM integrated layerwise methods
LMAM layerwise multiscale analysis method
LSM level set method
LWT layerwise theory
LW/SE laywise/solid-element
FDM finite difference method
FWDM forward difference method
Full-XLWM full extended layerwise method
FSDT first-order shear deformation theory
FEM finite element method
HSDT high-order shear deformation theory
HrSDT higher-order shear deformation theory
GLPT generalized laminate plate theory
GFRP glass fiber reinforced plates
PMC polymer matrix composite
MMC metal matrix composite
MCEP minimum complementary energy principle
MCTSC maximum circumferential tensile stress criterion
MPEP minimum potential energy principle
MQ modified multiquadrics
RBF radial basis functions
RLWT Reddy’s layerwise theory
RPIM radial point interpolation method
SAM semianalytical method
SDT shear deformation theories
SIF stress intensity factor
SERR strain energy release rate
xix
xx Acronyms
Nowadays composite materials play a very important role in all types of engineering
structures, such as aerospace, automotive, underwater structures, medical productions,
electronic, and sports equipments. Computational mechanic analysis is the base of the
dramatic development of composite materials. Composite structures can be studied
by using two basic analysis schemes: micromechanical methods and macromechanical
methods. In general, micro-mechanical analysis is aimed at predicting and understanding
the average properties in terms of the detailed microscopic behavior of the mate-
rial, rather than generating accurate design data; while macromechanical analysis draws
mainly on the results obtained from physical and mechanical testing of unidirectional
composites. If we need to analyze the macro and microresponse simultaneously, mul-
tiscale analysis methods are necessary. This book is focused on the macromechanical
analysis and multiscale analysis of composite engineering structures. In the early days
of laminated composites, the technique used for analyzing conventional plates was ex-
tended to analyze these new structures. The zig-zag effects and Cz0 -Requirements poses
a serious challenge to the early traditional analytical methods. Although the compos-
ite analysis theories have made great progress, there are still many challenges because
of the complex characteristics and wide applications of composites: their complex in-
tegrated molding process, complex anisotropic constitutive relations, complex damage
mechanism, complex multiphysical loading, and complex multiscale effective, namely
C 5 challenges. Unfortunately, it is almost impossible to accurately consider all the chal-
lenges using the existing refined analysis methods.
The author and his research team have focused on solving the basic mechanical prob-
lems in composite engineering structures, especially the theories of composite laminated
beams, plates, and shells; they are called the composite laminated theories in this book.
The research team has developed a composite structure analysis software system, with
a total code of more than 200,000 lines. The software system has a large number of
solvers based on shear deformation theories, state space theory, the layerwise method,
extended layerwise method, computational continua, multipoint multilevel grid refine-
ment method, and taking the fracture, damage, multiscale and multiphysics analysis
problems as its advantages. Analysis objects of this software system include: beams,
plates, shells, stiffened plates and shells, sandwich plates and shells, and multilayer sand-
wich plates and shells. Loading types include: mechanic, thermal, electric, chemical, and
their coupling loading. The software system adopts advanced storage and solution tech-
nologies, and requires less memory and hard disk than commercial software. For some
specific challenging engineering problems, the efficiency and accuracy of the proposed
software are better than these of the commercial software as well.
xvii
xviii Preface
Dinghe Li
Tianjin
CHAPTER 1
Analysis of Composite Laminates Copyright © 2022 China Science Publishing & Media Ltd. Published by Elsevier Inc.
https://doi.org/10.1016/B978-0-32-390804-7.00009-1 All rights reserved. 1
2 Analysis of Composite Laminates
heat generated during the reentry of a spacecraft could exceed 1500◦ C, which is beyond
the temperature limits of any monolithic or composite material, especially low-melting
point polymers. In 1956, an asbestos fiber was added into a phenolic resin as a possible
reentry nosecone material, and the composites of metal matrix was also regarded as a
solution. In these composite materials, the inorganic, ceramic fiber, or particulate phase
were employed to improve the heat resistance of lightweight metals and to lower their
thermal expansion coefficient. The space race provided an impetus for the development
of carbon and boron fibers as well. They were developed around the same time, but
the carbon took the lead in the 1960s due to its superior processing capabilities and
lower cost. The graphite fibers were of use only in polymer matrices at this time.
Because of the reactivity of carbon with aluminum and magnesium, the use of graphite
fibers as reinforcement for metal matrices was not possible. The strength of boron fibers
exceeded that of carbon fibers, and it is more suitable to military applications where the
cost was no concern, and made no extension into other industries. The compound of
aramid fibers was developed in 1964. Aramids belong to the nylon family of polymers;
the key structural feature is aromatic rings linked by amide groups.
The third stage is searching for new application fields and synergy of properties,
whereas the spacecraft and aircraft demands prompted new high modulus fibers in the
1960s. The composites made with the aforementioned expensive fibers had to find civil
applications in the 1970s when the space and military demands declined. The sports and
automobile industries became more important markets. At the same time, the analysis
and design methods based on the computational composite mechanics drove further
applications, such as the civilian and military aircraft. The carbon fibers were used ex-
tensively in sporting products beginning in the 1970s, with graphite tennis rackets and
golf clubs replacing the wooden and steel materials. During the same period, the com-
posites of a ceramic matrix was developed and applied. It must be reinforced by high
temperature fibers, such as SiC, because low-melting fibers would be destroyed at the
high processing temperatures required for ceramic sintering. On the other hand, the
brittle ceramics need a reinforcing phase to improve the toughness.
The fourth generation is the hybrid materials, nanocomposites, and biocomposites
(green-composites). In the 1990s, both academic and industrial researchers started to
extend the composite paradigm into smaller and smaller scales. From the macroscopic
scale to the molecular scale, it results into hybrid materials. They are an intentional
and complimenting combination of two or more materials with new properties. Ac-
cording to their criterion, the difference between hybrid materials and composites is
their functions and/or properties. Toward the nanoscale, it results into nanocomposites.
Nanocomposites are solid materials that have multiple phase domains and at least one of
these domains has a nanoscale structure. A biocomposite is a material composed of two
or more distinct constituent materials (one being naturally derived), which are com-
bined to yield a new material with improved performance over individual constituent
Composite analysis overview 3
With decades of development, the composite materials have been called the shape
of aerospace’s future. In the civil aviation industry, A380, A350, and B787 represent the
application of composite materials. The A380 is the first aircraft that has a central wing
box with carbon fiber reinforced plates (CFRP), representing a weight saving up to one
and a half tonnes compared to the most advanced aluminum alloys. The main challenge
is the wing root joint, where composite components could be up to 45 mm thick.
For this specific application, Airbus reaps a large benefit from the A340-600 CFRP
keel beams (16 meters long and 23 mm thick), each of which carries a force of 450
tonnes. A monolithic CFRP design has also been adopted for the fin box and rudder,
as well as the horizontal stabilizer and elevators. The size of the CFRP horizontal
tail plane is close to the size of wings of the A320, so the main challenge becomes
the size of the components. As for the center wing box, the size of the components
justifies the intensive use of automated tape-laying technology. Furthermore, the upper
deck floor beams and the rear pressure bulkhead is made of the CFRP. The advanced
glass/aluminum composites was used on the upper portion of fuselage, and it is called
the glare laminate. The glare laminate is a fiber metal laminate composed of several very
thin aluminum layers interspersed with glass fiber prepreg layers, and they are bonded
together with a matrix such as epoxy. This new aero structure material is widely used
in fatigued parts due to its outstanding features in anticracking and tension.
Boeing was the first to commit to a composite fuselage and wing for the B787. Such
extended applications of composite materials (about 50% of the structure) would give
the efficiency improvement needed by the airline operators, such as about 20% more
fuel efficient and 20% fewer emissions. The B787 is the first major commercial airplane
to have a composite fuselage, wings, and most other airframe components. Each B787
contains approximately 35 metric tons of CFRP, made with 23 metric tons of carbon
fiber. The fuselage is constructed in tubular segments, which are then joined together
during the final assembly. The applications of composites in the B787 saves 50,000 rivets
per plane; and each rivet site would have required maintenance checking as a potential
failure location. Other advantages of using composite materials is that a typical bonded
repair on the standard aluminum airframe sometimes requires more than 24 hours of
downtime, but Boeing has created a new way of repairing composite materials in less
than an hour. This leads to quicker turnarounds and offers the opportunity for a quick
in-and-out for minor damage that might make an aluminum airplane unable to fly.
The A350 is the first Airbus airplane with both fuselage and wing structures made
primarily of the CFRP. The A350 also has a slightly higher proportion of composites
(about 52% of the structure) in its construction than that of the B787. Most parts of the
aircraft are a mixture of composite and alloy. In the wings, the main metal components
are interspar ribs (the spars being the supporting structures and running along the wing
from tip to root; the ribs running across the width of the structure). In the fuselage, the
outer skin panels are constructed by CFRP, and the frame includes aluminum strips to
ensure that lightning strikes can be dissipated.
Composite analysis overview 5
The fiber/matrix interface is known as the interphase. The properties of matrix and
the bond of fiber/matrix interface dominate the mechanical properties of composite
materials. The matrix and matrix/fiber interaction also have a significant effect on the
crack propagation of the composite materials. If the matrix shear strength and modulus
and the fiber/matrix bond strength are too high, a crack may propagate through the
fiber and matrix without turning, so the composite materials will behave as a brittle
material and show clean fracture surfaces. If the bond strength is too low, the matrix
will act as a fiber bundle and the composite materials will be weak. For the intermediate
bond strength, the cracks propagating in resin or fiber would turn at the matrix/fiber
interface and extend along the fiber. The composite materials that failed in this mode
will show considerable fiber pull-out and the fracture surface will be very rough. This
results in considerable energy absorption.
If the intact fibers are available behind the crack tips and connect at the crack faces,
the crack bridging mechanism is operative. Hence, the load would be shared by the
bridging fibers and crack tips, and the stress intensity factor (SIF) on the crack tip would
be reduced significantly. A higher amount of bridging fibers leads to the lower SIF on
the crack tip, and the resistance to crack growth increases with crack length increasing.
The extension of a transverse crack bridged by the intact fibers leads to the debonding
and fibers pull-out. This will increase the fracture toughness of composite materials.
The strength and fracture toughness of fiber reinforced composites is determined by the
interplay between the damage processes in different elements of composite materials.
Therefore the fiber bridging effect of transverse crack and delamination is a very im-
portant and challenging problem to understand the damage mechanism of composite
materials, and many analytic methods have been established [2,3].
1.2.2 Plies
A ply, which consists of fibers and matrix and also known as a lamina or layer, is shown
in Fig. 1.1. The continuous fibers in the plies can be oriented in a single direction,
or multiple directions for the woven composites. In the aircraft structures, the plies
can be in a prepreg form or dry fabric form. The prepreg is a common form. It is
preimpregnated by the fibers and a resin in a semicured state, where the fibers and resin
are combined but are still flexible enough to be laid in tooling. The laminate is fully
cured prepreg layers, and then the cured resin is referred to as the matrix mentioned
in previous section. The prepreg forms also include: tape, slit tape, fabric, sheet, and
tow/roving. The prepregs are common used for large aircraft structures and can consist
of dry fabric layers that are impregnated with an uncured and low viscosity resin. This
is known as a wet layup, as it may be considered for repairs but are not commonly used
for the original structures in large aircraft.
For a unidirectional individual ply, the fibers are all aligned in a single direction. In
tape form, it may also be called a unitape or a tape ply. Fabrics may be in prepreg or dry
Composite analysis overview 7
Figure 1.1 Ply in composite materials. (a) Unidirectional ply and (b) Fabrics.
form. For woven fabrics, such as a plain weave fabric shown in Fig. 1.1(b), the weaving
bends the fibers. Furthermore, the fibers are not as closely packed as for unidirectional
plies. These characteristics reduce the static strength and stiffness, especially for the
compression loading. For the tensor loading, the stiffness of woven fibers would increase
due to the fiber stretching. For the manufacturing of composite structures, a fabric ply
is more easily draped over a compound surface than a unidirectional prepreg ply.
1.2.3 Laminates
The composite laminate is an assembly of individual layers to provide required me-
chanical properties, including in-plane stiffness, bending stiffness, strength, thermal
expansion coefficients, and so on. It is result of the designability of composite mate-
rials; see Figs. 1.2 and 1.3. The multiple cross-ply laminate is presented in Fig. 1.2; the
two-dimensional woven laminates and three-dimensional fabrics are shown in Fig. 1.3.
A layup is a processing method where plies are placed on top of each other and is defined
by the material system, orientations, and stacking sequence of the plies. Although all of
the plies in a laminate could be oriented with fibers aligned in the same direction, the
mechanic properties would be undesirable. Therefore the practical laminates are usually
multidirectional. The laminate coordinate system is the x-y-z system, namely global
coordinate system, as opposed to the ply’s local 1-2-3 coordinate system; see Fig. 1.2.
The orientation angle of individual ply is with respect to the x-y-z system. In the ply’s
local coordinate system, the 1-direction (longitudinal) is aligned with the fibers, the
2-direction (transverse) is normal to the fiber direction, and the 3-direction is along the
thickness and coincides with z-axial of laminate coordinate.
In the laminates, the individual layers generally are orthotropic or transversely
isotropic, so the laminate exhibits anisotropic, orthotropic, or quasiisotropic properties.
The quasiisotropic laminate exhibits isotropic in-plane response but is not restricted to
isotropic out-of-plane response. Depending upon the stacking sequence of individual
layers, the laminate may exhibit coupling between in-plane and out-of-plane responses.
8 Analysis of Composite Laminates
The orientation and position of each individual ply in a laminate may be expressed
in multiple ways. Although there are no universal rules, some of the common ways of
representing the ply orientations and laminate stacking sequence are presented in this
subsection, and they will be used in the following chapters. The 0◦ ply has its 1-direction
aligned with the x-direction in the global coordinate, and the 90◦ ply has its 1-direction
aligned with the y-direction in the global coordinate; see Fig. 1.2. In general, the first
ply in the sequence is located at the most negative z-position.
Composite analysis overview 9
The degree symbol can be omitted for convenience. Some examples, which are not
necessarily indicative of practical laminates, are given by
[0/0/0/0] = [04 ]
[0/+45/90/−45/−45/90/+45/0] = [0/+45/90/−45]s
[0/+45/90/+45/0] = [0/+45/90]s
(1.1)
[0/90/0/90/0/90] = [0/90]3
[0/+45/−45] = [0/±45]
[0/0/+45/−45/−45/+45/0/0] = [02 /±45]s
where s indicates that the laminate is symmetric about the given ply.
Figure 1.4 Micromechanical methods and macromechanical methods of composite structures. (a) Mi-
cromechanical methods; (b) Mesomechanical methods; (c) Macromechanical methods.
Another one is the elasticity approach aimed at the upper and lower bound solutions,
exact solutions for very specific systems or numerical solutions.
Except the aforementioned two basic analysis schemes, the multiscale methods,
which are aimed at bridging material and structural scales (see Fig. 1.5), have been wildly
applied into the analysis of composite laminates. The multiscale simulation scheme takes
advantage of the fact that composite structures are made up of laminates, which in turn
are obtained by stacking individual plies with different fiber orientation. This leads to
three different entities (ply, laminate, and component) whose mechanical behaviors are
characterized by three different length scales, namely fiber diameter, ply, and laminate
thickness, respectively. Fiber diameters are of the order of 5–10 µm, while ply thick-
nesses are in the range 100–300 µm and standard laminates are several mm in thickness
and above. This clear separation of length scales is very useful to carry out multiscale
modeling by computing the properties of one entity (e.g., individual plies) at the rele-
vant length scale, homogenizing the results into a constitutive model, and passing this
information to the simulations at the next length scale to determine the mechanical
behavior of the larger entity (e.g., laminate). Thus the multiscale modeling is carried
out through the transfer of information between different length scales rather than by
coupling different simulation techniques.
In this book, we are mainly focused on the analysis methods of macroscopic re-
sponses, such as the static responses, dynamic responses, sensitivity coefficients, and
responses of fracture and damage. The multiscale analysis methods based on the macro-
scopic methods are also presented, such as the beam, plate, and solid elements. The
micromechanical is not discussed in following chapters of this book. Therefore the ba-
sic equations, which will be employed to analyze the macroscopic responses, are briefly
introduced.
Composite analysis overview 11
Figure 1.6 A three-dimensional deformational body and the equilibrium of elemental volume.
boundary u . On the other part of boundary σ , the distributed force per unit area t̄i
(i = x1 , x2 , x3 ) is subjected, and it also called traction. The body force fi is subjected to
the elemental volume dV . A load acting at a point is represented by its three components
Pi , also called point loads. Under the forces, the body deforms, and the deformation of
a point is given by three components of displacement ui . Fig. 1.6(b) shows a vanishing
small element of a deformable body, together with the stresses. The stresses on the
mutually parallel planes are the same but with an infinitesimal change in value. Summing
forces at the directions x1 , x2 , and x3 , including the body force fi , we then get three
differential equilibrium equations. They are given by
σji,j + fi = 0, i, j = 1, 2, 3 (1.2)
where i is a free index. This means that there are three subequations in the above
differential equilibrium equation, while j is a dummy index, summing within its range.
For the elastic dynamic problems of composite laminates, the inertia force can be
regarded as body force according to the d’Alembert principle. Therefore the differential
Composite analysis overview 13
∂ 2 ui
σji,j + fi = ρ (1.3)
∂ t2
where t denotes time. ρ is the density.
Considering the moment equilibrium of elemental volume in Fig. 1.6(b), we can
obtain the equivalent theorem of shearing stress (symmetry of stress tensor) as
x = x(X ). (1.5)
In the rectangular coordinate system, the Green strain tensor are given by
1 ∂ ui ∂ uj ∂ um ∂ um
Eij = + + (1.6)
2 ∂ Xj ∂ Xi ∂ Xi ∂ Xj
For the small deformation problem of composite laminates, we can obtain the strain-
displacement relationship as
∂ u1 1 ∂ u1 ∂ u2
ε11 = ε12 = ε12 = +
∂ x1 2 ∂ x2 ∂ x1
∂ u2 1 ∂ u2 ∂ u3
ε22 = ε23 = ε32 = + (1.7)
∂ x2 2 ∂ x3 ∂ x2
∂ u3 1 ∂ u1 ∂ u3
ε33 = ε13 = ε31 = +
∂ x3 2 ∂ x3 ∂ x1
In Fig. 1.6(a), there are displacement boundary conditions and traction conditions.
For the displacement boundary condition u , the displacement field should be equal to
the constraint ūi , namely
ui = ūi (1.8)
14 Analysis of Composite Laminates
In the early days of composite laminates, the previous techniques used for analyzing
conventional plates were extended to analyze this new structures directly. The classi-
cal laminated plate theory (CLPT) is an extension of the classical plate theory based
on Kirchhoff assumptions. In order to analyze the composite structures accurately and
effectively, many theories have been established in the last half century [9], such as
equivalent single-layer theory (EST), layerwise theory (LWT), Zig-Zag theory (ZZT),
three-dimensional elastic methods, and so on. The CLPT and first-order shear defor-
mation theory (FSDT) are the two most popular ESTs, and widely used in design,
analysis, and optimize the composite engineering structures; they are also employed
in the commercial finite element software. The ESTs can provide good results for the
global responses of very thin composite laminated plates and shells, for example, gross
deflections, critical buckling loads, fundamental vibration frequencies, and associated
mode shapes, but have poor results for the thick composite laminates, especially the lo-
cal responses, such as the distribution of ply-level stresses. The CLPT cannot represent
the realistic distributions of displacements and stresses along the thickness direction. The
major drawback of ESTs is introduced by the continuous assumption of in-plane dis-
placements, and it cannot present the Zig-Zag distribution of in-plane displacements.
The LWT and ZZT were established to overcome the drawback of ESTs, and they have
become the most important developed direction of composite laminated theories.
16 Analysis of Composite Laminates
According to the Carrera’s point of view, we can divide the available theories of
monocoque plates and shells into the following three groups [4]: (1) continuum based
or stress resultants based models; (2) asymptotic type approaches; (3) axiomatic type ap-
proaches. In the framework of the first approach, the three-dimensional continuum of
plates or shells is seen as a surface on which correspondence stress resultants are defined;
the two-dimensional approximations are then introduced at a certain level and inte-
gration in the thickness direction is then performed. These approaches do not present
any difficulties in the formulation of nonlinear theories in both the case of geometric
and physical nonlinear behavior. In the framework of asymptotic approaches, the three-
dimensional governing equations are expanded in terms of a perturbation parameter
(usually the thickness to length ratio is used) and theories related to the same order
in perturbation parameter are derived. In the axiomatic framework, an certain displace-
ment and/or stress fields are postulated in the thickness direction; it permits to introduce
intuitive approximations into the plate/shell behavior while no results can be established
as far as the convergence of intuitive models to three-dimensional solutions.
1.3.4 Challenges
Taking the B787 as an example, the application of a composite would result into a
weight saving up to 15 tons theoretically, but in fact, the weight of a B787 is only
reduced by 4.5 tons compared with the traditional metal components; although B787
represents the highest level of composite application in civil aircrafts. From the author’s
point of view, there are three main reasons: (1) lack of experience in using this new ma-
terials; (2) lack of effective experimental schemes; (3) theoretical and numerical methods
can not meet the requirements of engineering problems. The last one will be discussed
in the following.
The composite laminates is a multilayered plate or shell; see Fig. 1.7. The challenges
of analysis method are from following aspects (namely, C 5 challenges): (1) exhibiting
the so-called complex anisotropy, in-plane anisotropy and transverse anisotropy, Zig-
Zag effects and interlaminar continuity (Cz0 -Requirements) [4]; (2) extremely complex
fracture and damage mechanisms; (3) complex structural forms resulted from integral
forming technology; (4) complex multiscale effect, there are three different entities: ply,
laminate, and component, and their mechanical behaviors are characterized by three
different length scales; (5) complex multiphysical loads, including mechanical, thermal,
curing fields and so on.
The in-plane anisotropy means different mechanical and physical properties in dif-
ferent in-plane directions, it could show higher transverse flexibility with respect to
in-plane deformability, namely shear and normal components. The laminates are the
high transversely deformable structures as well, exhibiting high values of Young’s moduli
orthotropic ratio, E11 /E22 = E11 /E33 ; and low values for the transverse shear modulus
ratio, G12 /E22 ≈ G23 /E22 . The transverse anisotropy would lead to higher transverse
Composite analysis overview 17
shear and normal stress deformability than the isotropic cases. In addition, the in-plane
anisotropy is related to the other complex problems: (1) coupling between shear and
axial strains; (2) coupling between in-plane and out-of-plane strains for the unsymmet-
rically laminated panels [4,10].
From a view of computation, due to the transverse discontinuous mechanical prop-
erties, the displacement fields in the thickness direction are extremely complex Zig-Zag
sharp; they must to be exhibited to represent the rapid change of their slopes in corre-
spondence to each interface; see Fig. 1.8. In-plane stresses σ11 , σ22 , and σ12 are in general
be discontinuous at each interface, while the transverse stresses σ13 , σ23 , and σ33 must be
continuous at each interface for equilibrium, it is called as the interlaminar continuity of
transverse stresses. The displacements and transverse stresses are C 0 -continuous functions
in the thickness direction. In the most general case, the displacement and transverse stress
have discontinuous first derivatives with correspondence to each interface, since the me-
chanical properties change along the thickness direction. The Zig-Zag and interlaminar
continuity are referred to as Cz0 -Requirements. The fulfillment of Cz0 -Requirements is
a crucial point in the development of any composite laminated theories.
The knowledge of damage and fracture mechanisms plays one of the most important
roles for the practical design of composite structures. Furthermore, the damage in fiber
reinforced composite materials occur at a number of length scales, from the micro-
scopic, barely observable fiber/matrix debonding, to macroscopic transverse cracking
and delamination. The layered, orthotropic, sometimes inhomogeneous and multima-
terial characteristics allow the possibility for occurrence of various failure modes under
different loading conditions. And various failure modes are exist in the damaged com-
posite structures simultaneously. In general, the failure modes of composites can be
categorized into four classes: fiber failure, interfacial delamination, matrix cracking and
fiber/matrix debonding; see Fig. 1.9. There is no method to predict effectively the ini-
tiation and growth of these damage, so does the effects of these damage on the structural
integrity.
Figure 1.9 Composite damage forms. (a) Debonding; (b) Fiber breakage; (c) Matrix crack and (d) De-
lamination.
The structural plates/shells systems stiffened by beams are widely applied in the
construction of aircraft fuselage and wings. For instance, the fuselage structure is a semi-
monocoque construction consisting of a thin shell stiffened by axial elements (stringers
and longerons) and transverse elements (frames or rings). The composite sandwich
structures, which consist of two thin but stiff composite laminated facesheets bonded
to a lightweight and thick core with low in-plane modulus are widely used in aero-
nautics and astronautics owing to the low weight and high rigidity. The complexity
of the overall and local behavior of sandwich structures has aroused a large number of
Composite analysis overview 19
computational methods. The mentioned two basic composite structures poses serious
challenges to the existing analysis methods. Furthermore, with the rapid development
and wide application of composite integral forming technology, the forms of stiffened
composite structures, grid structures, and sandwich structures are becoming more and
more complex.
Due to the obvious multiscale effect, the macroscopic properties and response of
composites depends on the microscopic and mesoscopic details, such as properties of
individual layers, bonding between layers, stacking sequences and so on. For compos-
ites design, a good understanding is required for the relation between micromechanical
and global structure properties. The traditional beam, plate and shell elements do not
provide detailed stresses in the matrix and fiber phases, while the direct numerical simu-
lation (DNS) that resolves microstructural details is too expensive for practical problems.
Therefore it is necessary to develop practical multiscale analysis models for the compos-
ite structures.
The composite laminated structures are often under the combined action of mul-
tiphysical field loads in the process of manufacturing and service period, such as
mechanic-thermal-curing coupling, mechanic-moisture-thermal coupling, mechanic-
thermal-electric of smart composites. The accurate analysis of multiphysical field be-
havior has become an unavoidable problem in the design of composite structures. The
computational complexity increases with the increase of multiphysical variables.
3. The complex fracture and damage problems can be solved properly by the three-
dimensional composite laminated theories, such as LWT, with the traditional frac-
ture and damage concepts and methods used in the three-dimensional finite element
method (FEM).
4. Similar to the three-dimensional FEM, the traditional multiscale analysis methods
can be introduced into the LWT to obtain the microscopic and mesoscopic re-
sponses in the composite plates/shells. Based on this scheme, the multiscale analysis
may be directly applied into the analysis and design of practical engineering struc-
ture of composite materials; it is difficult for the existing multiscale analysis methods
based on the ESTs.
5. The existing methods should directly applied to the complex composite structures
manufactured by integral forming technology and second molding, it is necessary
to established some special schemes and to develop special software.
6. Although the composite laminated theories were developed originally for the com-
posites, it can be extended to other isotropic or anisotropic multilayered structures,
such as functional graded materials, thermal barrier coating, multilayered anticorro-
sive coating and so on. The multilayer structures with different characteristics exist
widely in nature and engineering practices, so the laminated theories will still be a
hot issue of computational mechanics.
7. The existing composite laminated theories can be employed to solve the ply-level
residual responses introduced by the multiphysic fields, including the mechanical,
thermal, curing, hygro and so on.
8. The existing composite laminated theories should be improved to solve the general
multidisciplinary complex engineering problems, such as multiphysic/scale analysis,
multiphysic/scale/damage analysis, multiphysic/scale/fracture analysis and so on.
where : denotes the double dot product. Since both of stress σ and stain ε are the
second-order tensor, according to the quotient rule of tensor, C is the fourth-order
tensor and is termed stiffness tensor.
In general, there are 34 = 81 scalar components in a fourth-order tensor. In the
absence of body couples, the principle of conservation of angular momentum requires
the stress tensor to be symmetric, namely, σij = σ ji , so Cijkl must be symmetric in the
first two subscripts. Hence, the number of independent material stiffness components
reduces to 6 × (3)2 = 54. Since the strain tensor is also symmetric, εkl = εlk , Cijkl must
be symmetric in the last two subscripts as well, the number of independent stiffness
components reduces to 6 × 6 = 36.
If the material is hyperelastic, the stress can be obtained based on a strain energy
density function U0 (εij ) as
∂ U0
σij = = Cijkl εkl + σij0 (1.12)
∂εij
and we have
∂ 2 U0
= C ijkl (1.13)
∂εij ∂εkl
Since the order of differentiation is arbitrary in above equation, Cijkl = Cklij ; the
number of independent stiffness components reduces to 21. Usually, we express equa-
tion (1.11) into Voigt–Kelvin notation, the single subscript is used for stresses and strains,
and two subscripts are used for the stiffness coefficients, for example,
Now the coefficients Cij must be symmetric for the hyperelastic material, the most
general elastic material has 21 independent stiffness coefficients. Assuming that the
stress-strain relations (equation (1.15)) are invertible, thus the strain-stress relations are
given by
εi = Sij σj + εi0 (1.16)
where Sij is the compliance parameter with [S] = [C ]−1 .
22 Analysis of Composite Laminates
For the general hyperelastic materials, there is the most complex normal and shear
coupling effect; see equation (1.15). Although the normal and shear coupling effect also
appears in the monoclinic materials, only the shear strain in symmetric plane x1 − x2
can rise to the normal stress; see equation (1.19). It can be found from above equation
that there is no normal and shear coupling effect for the orthotropic materials.
In general, the most simple mechanical-property characterization tests are performed
with a known load or stress. Hence, it is convenient to write the inverse of relations in
equation (1.20), namely [S] = [C ]−1 . In general, the material properties are determined
in a laboratory in terms of the engineering constants such as Young’s modulus, shear
modulus and so on. These constants are measured using simple tests like uniaxial tension
test or pure shear test. Because of their direct and obvious physical meaning, engineering
constants are used in place of the more abstract stiffness coefficients Cij and compliance
coefficients Sij . The strain-stress relationship in term of engineering constants is given
by
⎡ 1 ν21 ν31 ⎤
− − 0 0 0
⎢ E11 E22 E33 ⎥
⎢ ⎥
⎧ ⎫ ⎢ ν12 1 ν32 ⎥⎧ ⎫
⎪ ε1 ⎪ ⎢− − 0 0 0 ⎥ ⎪σ1 ⎪
⎪
⎪ ⎪
⎪ ⎢ E11 ⎥⎪ ⎪
⎪
⎪ ε2 ⎪⎪ ⎢ ν E22 E33 ⎥⎪⎪σ ⎪ ⎪
⎪
⎪ ⎪
⎪ ⎢ 13 ν23 1 ⎥⎪⎪ 2⎪⎪
⎨ε ⎬ ⎢− − 0 0 0 ⎥ ⎨σ ⎪
⎪ ⎬
⎢ ⎥ 3
= ⎢ E11 E22 E33
3
⎥ (1.21)
⎪
⎪ ε4 ⎪⎪ ⎢ 1 ⎥⎪ σ ⎪
⎪
⎪ ⎪
⎪ ⎢ 0 0 0 0 0 ⎥⎪⎪ 4⎪ ⎪
⎪
⎪ ε ⎪ ⎢
⎪ G23 ⎥⎪⎪
⎪ σ5 ⎪
⎪
⎪
⎪
⎩ ⎭ ⎢
5 ⎪ ⎥⎪ ⎪
ε6 ⎢ 1 ⎥ ⎩σ6 ⎭
⎢ 0 0 0 0 0 ⎥
⎢ G13 ⎥
⎣ 1 ⎦
0 0 0 0 0
G12
The above equation presents the relation of compliance coefficients Sij to the engi-
neering constants. Since the compliance matrix [S] is the inverse of the stiffness matrix
[C ] and the inverse of a symmetric matrix is symmetric. The stiffness coefficients also
can be expressed in terms of the engineering constants as well
where
E
= (1.26)
(1 + ν)(1 − 2ν)
where α and β take the values of 1 and 2. Although σ33 = 0, ε33 is not zero.
Composite analysis overview 25
Note that the reduced stiffness involve four independent material constants,
E11 , E22 , v12 , and G12 . The transverse shear stresses are related to the transverse shear
strains in an orthotropic material by the following relations
σ4 Q44 0 ε4
= , Q44 = G23 , Q55 = G13 (1.31)
σ5 0 Q55 ε5
JUIN 1755. »
NOTICE HISTORIQUE
SUR LES TRAVAUX AUXQUELS IL A DONNÉ LIEU
E port de Caen est aussi ancien que la ville. Dès lʼan
1026, il avait assez dʼimportance pour que la dîme des
produits de sa douane fût attribuée par Richard II,
comme une donation sérieuse, à lʼabbaye de Fécamp.
Au temps du duc Guillaume, sa prospérité fut encore
augmentée par la conquête de lʼAngleterre, qui amena
nécessairement un échange de productions entre la Normandie et le
royaume nouvellement conquis.
Jusque-là, les navires nʼavaient eu pour principale station que le
cours du Grand-Odon, depuis lʼendroit où cette rivière se jetait dans
lʼOrne, cʼest-à-dire vers le point où est actuellement le pont des
Abattoirs, jusquʼau pont de Darnetal, appelé plus tard pont St-Pierre.
La première amélioration du port fut entreprise par le duc
Robert, fils de Guillaume le Conquérant, vers lʼannée 1104. Après
avoir renforcé lʼOdon dʼune branche de lʼOrne, à laquelle la postérité
reconnaissante a conservé le nom de canal du duc Robert, le duc fit
creuser à lʼOdon un nouveau lit dans la prairie St-Gilles, pour lʼélargir
et le rejeter un peu plus haut dans lʼOrne, vers le lieu quʼon appelle
encore le rond-point. Grâce à ces travaux, des bâtiments plus forts
purent remonter jusquʼau pont St-Pierre.
Ils y vinrent en si grand nombre que, quelque dix ans après
cette première amélioration, la vue du mouvement du port excitait
lʼadmiration dʼun certain Raoul Tortaire, moine de lʼabbaye de St-
Benoît-de-Fleuri (Loiret), qui nous a laissé une curieuse relation en
vers latins du voyage quʼil avait fait en Normandie, à une date quʼon
peut fixer dʼune manière certaine entre les années 1107 et 1113.
« Le port, dit-il dans son poëme, donne asile à quelques gros
vaisseaux que lui envoie la mer, dont les ondes, dans leur flux,
suspendent presque entièrement le cours de la rivière. Ce sol,
fécond en moissons, ne connaît pas lʼombrage des forêts; la noix
gauloise, le raisin, la figue et lʼolive lui manquent; mais lʼîle
Britannique lʼenrichit des produits divers du commerce et de ce
quʼenfantent les terres baignées par la mer dʼOccident. »
Ébloui et tenté par le nombre et lʼéclat des étoffes de laine de
diverses couleurs, des tissus de lin dʼune rare finesse, des soies
moelleuses à trame serrée, et des autres marchandises quʼon
débarque sur le quai, le bon moine sʼécrie naïvement: « A la vue de
tant de richesses apportées des pays les plus divers par des
hommes, dont les vêtements sont si disparates, je me sens tout
agité et horriblement malheureux de ne pas avoir dʼargent! ».
Ces brillants produits de lʼOrient, qui faisaient regretter au bon
religieux ses vœux de pauvreté, étaient échangés contre le blé,
lʼorge, le hareng salé qui servait à lʼapprovisionnement des places
fortes, et aussi contre les pierres à bâtir tirées des carrières de
Vaucelles et de St-Julien.
e
Au XIII siècle, lʼaffluence des « navires chargés de toute sorte
de marchandises » est encore affirmée en vers latins, par Guillaume
36
Le Breton , historiographe de Philippe Auguste. Mais le
mouvement du port dut singulièrement se ralentir pendant les
malheurs de la guerre de Cent-Ans, les troubles de la Ligue du bien
public et les dévastations des guerres de religion. Durant cette
longue période de désastres, aucune amélioration nouvelle ne fut,
on le comprend, apportée à la situation du port.
Cependant, il en eût exigé dʼurgentes; car, tandis que le pays
commençait à se débarrasser de ses ennemis, le port de Caen
subissait un autre genre dʼinvasion qui devait compromettre sa
fortune et le menacer dʼune ruine prochaine. Lʼhistoire du port, à
e
partir du XVI siècle, ne se compose guère, en effet, que de la
relation des envasements successifs de lʼOrne, des projets quʼon
proposa et des travaux qui furent tentés pour remédier à cet état
périlleux pour la navigation. Ces envasements redoutables tenaient
à la nature des terrains où lʼOrne sʼétait creusé son lit capricieux. Le
sol des prairies de Caen jusquʼà la mer nʼest, en effet, que le produit
des matières que lʼeau de la rivière et le flux des marées avaient
successivement déposées dans lʼancienne baie. Des fouilles,
e
exécutées à la fin du XVIII siècle pour creuser le nouveau canal de
lʼOrne, ont donné lieu à des découvertes qui sembleraient prouver
que cette alluvion ne sʼest pas accomplie avec la lenteur que met
habituellement la nature dans son patient travail des siècles. Telle
est, du moins, lʼopinion dʼun observateur du temps, qui pense que le
sol de lʼancienne baie de Caen se serait exhaussé de 6 mètres