Design and Optimization of Laminated Composite Structure using 3-D Failure Criterion

Sampath Kumar L 1,

Abstract: Sizing of laminates for practical applications requires a reliable criterion for design. Stressing and failure
mechanisms in fiber- reinforced plastic laminates are indeed complex and so the development of reliable and standard failure criteria
requires special attention. In general, a failure criterion is formulated based on a set of experimentally evaluated strength values. Typical
examples in this category include Maximum Stress criterion, Maximum Strain criterion, Tsai-Wu criterion; Etc contains an excellent review
of several criteria currently in use. In this approach, it is postulated that the failure occurs at a point in the laminate, when stresses or strains
satisfy a certain condition at that point. An important limitation of this approach is that it has no provision to include the effect of the
operating failure mechanism during the failure process
It is well known that different failure mechanisms are operational, at different points in the laminates depending upon the stress
state. Although the failure process involves a combination of crack systems, representative of fiber breaks, matrix cracks, fiber buckling,
matrix crushing, fiber disbonds and delaminations, one can still expect that a few of them, certainly not all, will dominate during initiation
of failure. As a simplification, one of the mechanisms can be considered as the initiating failure mechanism, at the time of the failure
initiation.

In mathematics, the simplest case of optimization, or mathematical programming, refers to the study of problems in
which one seeks to minimize or maximize a real function by systematically choosing the values of real or integer variables
from within an allowed set. Optimization is the process of simultaneously optimizing two or more conflicting objectives
subjected to certain constraints. Optimization problems can be found in various fields: product and process design, finance,
aircraft design, the oil and gas industry, automobile design, or wherever optimal decisions need to be taken in the presence of
trade-offs between two or more conflicting objectives. Maximizing profit and minimizing the cost of a product; maximizing
performance and minimizing fuel consumption of a vehicle; and minimizing weight while maximizing the strength of a
particular component are examples of multi-objective optimization problems.
Keywords ELL,ETT ,Ett : Longitudinal, transverse, normal elastic moduli of laminate, E f - Youngs Modulus fiber, Em - Youngs Modulus
matrix, E11 - Youngs Modulus in the Longitudinal directions, E22 - Youngs Modulus in the Transverse directions, E33- Youngs Modulus in
the Transverse directions

1.

INTRODUCTION
A thin rectangular composite laminated plate of length a in x- direction, width b in the y -direction, and thickness h in the

z-direction is considered in the present work. The plate is assumed to be constructed of arbitrary number, n, of linearly elastic
orthotropic layers. Each layer consists of homogeneous fibre reinforced composite material and the plate is simply supported on
all four edges. A rectangular Cartesian coordinate system x, y and z is used to describe the infinitesimal deformations of an Nlayer laminated composite plate. The laminate consists of n plies with the individual thicknesses hi and the layer orientation angles

i i 1, 2,3......n as

shown in Figure 2.1. Total thickness of the plate is h and bottom and top surfaces are located at

z h 2 and z h 2 respectively. We assume that the middle surface of the undeformed plate coincides with the xy- plane.
The principal fiber direction is oriented at an angle

to the x-axis. The layer wise principal material coordinate system, denoted

by 1, 2 and 3 are used to find the stresses and strains in order to analyze lamina failure. The 1 and 2 axes are parallel and

perpendicular

to

the

x-y

plane

respectively,

and

the

3-axis

is

aligned

parallel

to

the

global

z-axis.

Figure 1: Geometry of a simply supported thin laminated plate

Sizing of laminates for practical applications requires a reliable criterion for design. Stressing and failure mechanisms in
fiber- reinforced plastic laminates are indeed complex and so the development of reliable and standard failure criteria requires
special attention. In general, a failure criterion is formulated based on a set of experimentally evaluated strength values. Typical
examples in this category include Maximum Stress criterion, Maximum Strain criterion, Tsai-Wu criterion; Etc contains an
excellent review of several criteria currently in use. In this approach, it is postulated that the failure occurs at a point in the
laminate, when stresses or strains satisfy a certain condition at that point. An important limitation of this approach is that it has no
provision to include the effect of the operating failure mechanism during the failure process
It is well known that different failure mechanisms are operational, at different points in the laminates depending upon the
stress state. Although the failure process involves a combination of crack systems, representative of fiber breaks, matrix cracks,
fiber buckling, matrix crushing, fiber disbonds and delaminations, one can still expect that a few of them, certainly not all, will
dominate during initiation of failure. As a simplification, one of the mechanisms can be considered as the initiating failure
mechanism, at the time of the failure initiation.
The effect of different stress components in initiating failure depends on initiating the failure mechanisms. For example,
considering the case when longitudinal tensile stress is close to XT, fiber breakage will be dominating failure mechanism, since all
other components of stress are much less than the corresponding strengths. In this mode, while the transverse stress T may
influence this failure, the interaction effect due to LT is negligible. For such a case, the Yamada-Sun criterion may be considered
appropriate. To develop more representative criteria, it will be necessary to use a representative damage model and fracture
mechanics. Similarly when T is close to YT, and other stress components are far less than the corresponding strength values,
matrix cracking may be expected to be the dominant failure mechanism. This case has been discussed in great detail, by HartSmith [5]. When T is close to XC, the ductile shear failure of the matrix could be appropriate failure mechanism and the criteria
based on this mechanism, was developed for Grumman aircraft. Figure 3 shows a possibility of dividing the failure envelope, into
areas in a qualitative sense, with specific initiating failure mechanism. Thus, division of the stress space in the failure envelope
into different regions with specific failure mechanism and development of suitable design criteria, specifically applicable for each
region, appears to be an interesting avenue for the development of rational failure criteria for fibrous composites. Division of the
stress space in the failure envelope is an essential prerequisite to implement failure mechanism-based design criteria. The concept
is indicated in Fig. 3. A great deal of further research is needed to accurately develop such maps. In this study, it is assumed that a
simple criterion such as the Maximum Stress criterion could be used, as a first step, to identify the initiating failure mechanism, by
comparing the ratios L/X T or L/XC and T/YT or T/YC. Depending upon the identified failure mechanism, appropriate failure

criteria can be used to design the laminates. The use of such an approach, which we refer to here as the failure mechanism-based
approach is likely to be useful in ensuring the reliability of designs. In order to bring out differences in using some of the failure
criteria currently in use and to demonstrate application of the failure mechanism-based approach, a rectangular panel subjected to
boundary displacements has been considered. By varying boundary displacements over a wide range, predictions of various
failure criteria are compared. Comparison of the `Failure Load Fractions' indicated substantial differences.
The failure mechanism in each case is indeed a complex crack system, involving one or more of the mechanisms, such as fiber
breaks, disbonds, delaminations, matrix cracks, etc. Thus, a consideration of a representative damage model is necessary to be
able to accurately model failure mechanism and to estimate corresponding failure load. Basic principles to achieve this goal are
well known, yet no standard criteria have emerged in the literature. As a step towards that direction, the edge failure has been
considered and failure load fraction has been estimated assuming edge delamination as the initiating failure mode. Through a
comparison of the results with the predictions of the conventional failure criteria, the need for development of failure mechanismbased approaches for design of laminates has been indicated.
2.

Failure predictions compared for different failure criteria:

Four well-known failure criteria, given under, are considered for a comparative study. Material moduli and strength

parameters are given in Table 2.

2.1 Maximum stress failure criterion:
Lamina fails when any one of the stress components

LT

reaches its corresponding strength value. There is no

interaction between stresses. In the Maximum Stress criterion, the stress in each direction of the local (material) system is
compared separately with the material strength in that direction. Therefore, there is no interaction between the stresses in
different directions. In a three-dimensional analysis, the failure envelope for this criterion is represented by:
.
in which the

and

allowables depend on the sign of

. . . . . (1)
,

and

respectively. (Generally, tensile and

compressive allowable are different for orthotropic materials.)

The reserve factor is calculated as follows:
.. . . . (2)
The Maximum Stress Criterion is very simple, intuitive, and easily implemented. Moreover, it allows the consideration of
different strength parameters for tensile and compressive strengths and the determination of the failure mode
2.2 TsaiWu Failure criterion:
The Tsai-Wu criterion is based on the polynomial failure theory proposed by Gol'denblat and Kopnov. The Tsai-Wu
failure surface is given by
Where i,j=1,2,.. 6 for a three dimensional stress state. The F parameters are obtained experimentally through uniaxial stress
tests in each direction. The stress invariants for transversely orthotropic symmetry are used in a polynomial expansion up to terms
of second degree. The following form is then the failure criterion:

. . . . . (3)
. . . . . (4)

Where,

. . . . . (5)
The strength constants FLL, FTT , FL , FT,, FLt , FTt , FSS are related lamina strength values, which are given table 1.
2.3

Tsai-Hill failure criterion:

The Tsai-Hill failure criterion is used as the basic composite material failure theory to estimate the material failure.

Three-dimensional Tsai-Hill criterion stress invariants for transversely orthotropic symmetry are used in a polynomial expansion
up to terms of second degree. The following form is then the failure criterion:

. . . . . (6)
. . . . . (7)
Where,

. . . . (8)
The strength constants FLL, FTT , FL , FT,, FLt , FTt , FSS are related lamina strength values, which are given table 2.
2.4 Hoffman Failure Criterion:
Hoffman failure criterion is similar as the Tsai-Wu failure criterion only the strength evaluated values are different. The
Hoffman failure criterion is used as the basic composite material failure theory to estimate the material failure. Three-dimensional
Hoffman criterion stress invariants for transversely orthotropic symmetry are used in a polynomial expansion up to terms of
second degree. The following form is then the failure criterion:

. . . . . (9)

. . . . (10)
Where,

. . . . (11)
The strength constants FLL, FTT , FL , FT,, FLt , FTt , FSS are related lamina strength values, which are given table 4. 1.
Table 4.1 Strength constants of Tsai-Wu, Tsai-Hill and Hoffman Failure criterion
Tsai-Wu Criterion (MPa)
Strength Parameters
Strength

Hoffman Criterion (MPa)

Strength Parameters
Strength

Tsai-Hill Criterion (MPa)

Strength Parameters
Strength Constants

FLL = (Xt * Xc )-1

Constants
3.465 x10-7

FLL = (Xt * Xc )-1

Constants
3.465 x10-7

FLL = 1/(Xt )^2

3.465 x10-7

FTT = (Yt * Yc )-1

7.352 x 10-5

FTT = (Yt * Yc )-1

7.352 x 10-5

FTT = 1/(Yt )^2

1.0416 x 10-4

FL = 1/Xt -1/Xc

-1.628 x10-4

FL = 1/Xt -1/Xc

-1.628 x10-4

FL = 0

FT = 0

FT = 1/Yt -1/Yc

9.705x10

-3

FT = 1/Yt -1/Yc

9.705x10

-3

FLT =-0.5(FLLFTT)0.5

1.485x10-9

FLT =-0.5(FLL)

-1.8225x10-7

FLT =-0.5(FLL)

-1.8225x10-7

FLt =-0.5(FLLFtt)0.5

1.485x10-9

FLt =-0.5(FTT)

-3.676x10-5

FLt =-0.5(FTT)

-3.676x10-5

FTt =FTT (FILSS/2)

7.3352x10-5

FTt =FTT (FILSS/2)

1.485x10-9

FTt =FTT (FILSS/2)

1.0416 x 10-4

FSS = 1/SLT2

1.5625x10-4

FSS = 1/SLT2

1.5625x10-4

FSS = 1/SLT2

1.5625x10-4

Figure (1) shows the [0,30,-30,60,-60,90]S laminate under consideration, bounded by surfaces 1 and 10 normal to X-axis
2 and 2` normal to Y-axis and top and bottom surfaces normal to Z-axis. 1 and 1` are subjected to normal displacement of U 2 and
tangential displacement of V2, each being constant over the entire surface. Similarly surfaces 2 and 2` receive a normal
displacement of V1 and a tangential displacement of U 1. Top and bottom surfaces are stress free. Such a situation may be viewed
as representative of a segment of a panel under plane loading of a thin-walled structure, such as an aircraft wing. Although in
reality the boundary displacements do vary to a certain extent both along the width and depth, the study of this example is
believed to be useful to bring out the comparative performance of various failure criteria. The standard ANSYS-12 finite element
software with 20 nodded 3-D Solid Layered elements with three degrees of freedom per node has been used to obtain numerical
results. In the thickness direction, element thickness is kept the same as ply thickness. A typical finite element mesh is shown in
Fig. 4(b) and (c). Total 900 elements, 1922 nodes and 5766 degrees of freedom are considered for numerical study. The panel is
1200 mm X 1200 mm with 1.5 mm total thickness (0.125 mm ply thickness).
Figure (1) shows the [0,30,-30,60,-60,90]S laminate under consideration, bounded by surfaces 1 and 10 normal to X-axis
2 and 2` normal to Y-axis and top and bottom surfaces normal to Z-axis. 1 and 1` are subjected to normal displacement of U 2 and
tangential displacement of V2, each being constant over the entire surface. Similarly surfaces 2 and 2` receive a normal
displacement of V1 and a tangential displacement of U 1. Top and bottom surfaces are stress free. Such a situation may be viewed
as representative of a segment of a panel under plane loading of a thin-walled structure, such as an aircraft wing. Although in
reality the boundary displacements do vary to a certain extent both along the width and depth, the study of this example is
believed to be useful to bring out the comparative performance of various failure criteria. The standard ANSYS-12 finite element
software with 20 nodded 3-D Solid Layered elements with three degrees of freedom per node has been used to obtain numerical
results. In the thickness direction, element thickness is kept the same as ply thickness. A typical finite element mesh is shown in
Fig. 4(b) and (c). Total 900 elements, 1922 nodes and 5766 degrees of freedom are considered for numerical study. The panel is
1200 mm X 1200 mm with 1.5 mm total thickness (0.125 mm ply thickness).
3 Failure Load Fractions:
Failure Load Fraction (FLF) is a constant which is defined as the ratio of applied stress to the failure strength, which is
multiplied to find out the safe operational load. In this case, the applied boundary displacements cause failure of the ply. The
procedure for estimating it for different criteria is stated under.
Table 2
(i)

Properties of laminates for Carbon/Epoxy lamina

Unidirectional Laminate Properties

a. Youngs Modulus (GPa)
Longitudinal Direction (L)

126

Transverse Direction (T)

11

Radial Direction (t)

11

b. Shear Modulus (GPa)

c.

(ii)

GLT

6.6

GTt

6.6

GLt
Poissons Ratios

3.929

LT

.28

Tt

.40

Lt
Strength of the Laminate (GPa)

.28

a.

b.

Tensile Strength
Longitudinal ( SLt )

1.95

Transverse (STt )
Compressive Strength

0.068

Longitudinal (SLc )

1.48

c.
d.
e.

Transverse (S c )
In-Plane Shear Strength (SLT )
Transverse (Interlaminar) Shear Strength (SILSS )
Normal Tensile (peal) Strength (STZ)

0.20
0.08
0.075
0.060

3.1 Maximum stress failure criterion: Failure Load Fraction (FLF) is smallest of
(i) Xt/L or Xc/L

(ii) Yt/T or Yc/T (iii) Zt/T or Zc/T

(iv) S/LT

(v) ILSS/ Lt

(vi) ILNS/ Tt

3.2 Tsai-Wu failure criterion: Failure load fraction, is obtained as the solution of the quadratic equation

Where,

3.3 Tsai-Hill Failure Criterion: Failure load fraction, is obtained as the solution of the quadratic equation

Where,
0

3.4 Hoffman failure Criterion: Failure load fraction, is obtained as the solution of the quadratic equation

Where,

Table (18) provide the results obtained for specified sets of boundary displacements U 1, U2, V1 and V2. Points 1, 2, 3, 4, 5,
and 6 represent the mid-points of the six layers (see Fig. 4(a)) and the corresponding FLFs are shown in these tables. The ply with
the lowest of these values fails first and is referred as a First Ply Failure Load Fraction (FPFLF) and these quantities are indicated
with a star (*) in the tables. A perusal of these tables clearly indicates differences in the predictions by these criteria.
Table (18) shows the ratios of the actual stress divided by the failure stress of the failing ply for various displacement
boundary loading conditions and this data is used here to identify the dominant failure mechanism. Failure is possible, when any
one of the ratios reaches one. However, since at each point there is more than one component, the component which gives the
highest stress ratio may be expected to initiate failure. For example in the case of U 1 =1, U2 = V1 = V2 =0 case, T / YT is the largest

and assuming these stress ratios grow in a similar fashion, T /YT reaches unity first, and so the initiating mechanism may be
identified as matrix cracking. Tables 1 to 17 indicates all possible initiating failure mechanisms corresponding to different cases.
Z

A 1

00

300

-300

600

-600

Interface -B
Interface -C
Interface -D
Interface -E
Interface -F

90

Interface G

Section A-A
Figure 2 Laminated Plate Section
Z
TOP SURFACE
00
300
-300
600
600
900
MIDDLE LAYER
Figure 3 Finite Element Discretization in ZX-Plane (Symm).

Figure 4 Initiating failure mechanisms at various points on the failure envelop for carbon/epoxy laminate
T
Matrix Cracks

L
Fiber

Fiber

Buckling

Breaks

C
Matrix Crushing
Figure 5 A typical division of failure envelope into regions with different initiating failure mechanisms.

Sl
No
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
Sl. No

Table 3
Boundary Displacements

Stress ratios in the first failing ply

U1x = 1; V1y = 0; U2x = 0; V2y =0

U1x = 0; V1y = 0; U2x = 0; V2y =1
U1x = 1; V1y = 0; U2x = 0; V2y =1
U1x = 1; V1y = 0; U2x = 0; V2y =0.2
U1x = 1; V1y = 0; U2x = 0; V2y =0.4
U1x = 1; V1y = 0; U2x = 0; V2y =0.6
U1x = 1; V1y = 0; U2x = 0; V2y =0.8
U1x = 0; V1y = 1; U2x = 1; V2y =0
U1x = 1; V1y = 1; U2x = 1; V2y =1
U1x = 1; V1y = 1; U2x = 1; V2y =0
U1x = 0; V1y = 1; U2x = 1; V2y =1
U1x = -1; V1y = 0; U2x = 0; V2y =0
U1x = 0; V1y = 0; U2x = 0; V2y = -1
U1x = -1; V1y = 0;U2x = 0; V2y = -0.2
U1x = -1; V1y = 0;U2x = 0; V2y = -0.4
U1x = -1; V1y = 0;U2x = 0; V2y = -0.6
U1x = -1; V1y = 0;U2x = 0; V2y = -0.8

26.24
26.24
9.002
41.043
21.718
14.765
11.185
.3070
.3039
.3040
.3040
191.702
13.608
9.179
9.134
9.089
8.489

2.816
2.816
2.031
2.455
2.333
2.223
2.213
.0950
.0969
.0950
.0950
3.788
11.805
5.450
3.836
2.959
8.846

0
0
0
0
0
0
0
4.595
.5594
.5594
.5543
6.502
6.502
0
0
0
0

Table 4 Optimum Weight for Uniaxial load different failure criterion

Failure Criterion
Total Ply numbers
Stacking Sequences

First failing
Ply location
1
6
1
1
1
1
1
1
1
1
6
2
2
6
6
6
6

Thickness

Weight

2D Maximum Stress Failure Criterion

22

[011] S

0.05000

104.5000

3D Maximum Stress Failure Criterion

14

[07]S

0.0750

102.5000

2D Tsai-Wu Failure Criterion

24

[012]S

0.0500

114.0000

3D Tsai-Wu Failure Criterion

18

[09]S

0.0700

119.5000

2D Hoffman Failure Criterion

24

[012]S

0.0500

114.0000

3D Hoffman Failure Criterion

18

[09]S

0.0675

115.0000

4.

Conclusion:
Failure of the composite laminates involves different failure mechanisms depending upon the stress state, and so different

failure mechanisms become dominant at different points on the failure envelope. Although the failure process involves a

combination of crack systems, representative of fiber breaks, matrix cracks, fiber buckling, matrix crushing, fiber disbonds, and
delaminations, one of these can be considered as the initiating failure mechanism at the time of the failure initiation.
Most of the existing failure criteria aim at covering all possible initiating failure mechanisms. Predictive capability of
such criteria varies with load and stress state. On the other hand, as against a single failure criterion concept to predict failures,
one can consider a set of criteria, each specifically tailored for a specific initiating failure mechanism. Obviously the second
approach is likely to perform much better, as the designer chooses the criteria, based on an initial/preliminary study of stresses and
initiating failure mechanisms. The present work is intended to focus attention on the need for such failure mechanism-based
failure criteria.
The free edge failure in laminates under extension is considered to provide a comparison between conventional failure
criteria prediction with delamination onset as the initiating failure mechanism model. Significant differences in the predicted
failure load fractions are noted. These studies bring out a need to develop failure mechanism-based failure criteria and their use in
designing to eliminate the uncertainties associated with the failure criteria.
The result is obtained by running the code for 200, 250 and 300 iterations and the repeated result were tabulated in Table
21. The results obtained after optimization for weight of laminated plate for uniaxial load with different failure criterion such as
Maximum stress failure criterion ( 2D & 3D), Tsai-Wu failure criterion (2D & 3D) and Hoffman failure criterion (2D & 3D)
concludes that the weight obtained for Maximum stress failure criterion for 2D is more than that of 3D failure criterion and w.r.t to
Tsai-Wu and Hoffman 2D failure criterion, the Optimum weight obtained is less than the Tsai-Wu and Hoffman 3D failure
criterion. Hence, we can conclude that the Maximum failure criterion is gives better result for 3D and the other failure criterion,
Tsai-Wu and Hoffman gives good result for 2D cases.
References
1. G. Narayana Naik, A.V. Krishna Murthy, A failure mechanism-based approach for design of composite laminates,
2.

Department of Aerospace Engineering, Indian Institute of Science,1999.

A.V.Krishna Murthy, G.Narayana Naik, and S.Gopalakrishnan, Towards a rational failure Criterion for Unidirectional

3.

Composite laminae, Department of Aerospace Engineering, Indian Institute of Science, 2005.

G.Narayana Naik, A.V.Krishna Murthy, S.Gopalakrishnan, A failure mechanism based failure theory for laminated
composites including the effect of shear stress, Department of Aerospace Engineering, Indian Institute of Science,july,

4.

2004.
U.lcardi, S.Locatto, A.Longo, Assessment of recent Theories for Predicting failure of Composite Laminates,

5.

Departmento di Ingegneria Aeronautica e Spaziale Politicnico di Torino,Italy. March, 2007.

Isaac M.Daniel, Jyi-Jiin Luo, Patrick M.Schubel, Mechanical and Failure Characterization of Textile Composites,

6.

Northwestern University, 2007.

Richard M. Cristensen, A Survey of and Evaluation Methodology for Fiber Composite Materials Failure theories,

7.

Department of Aeronautics and Astronautics, Sanford University, Stanford, Calif.,USA,2001.

I.M.Daniel, Failure of Composite Materials, Robert R. McCormick School of Engineering and Applied Science,

8.

Northwestern University, Evanston, USA, 2007.

W.C.CUI, M.R.Wisnom and M.Jones, A comparision of failure criteria to predict delamination of unidirectional

9.

glass/epoxy specimens waisted through the thickness University of Bristol, UK, Dec,1991.
S.N. Omkar *, Dheevatsa Mudigere, G. Narayana Naik, S. Gopalakrishnan, Vector evaluated particle swarm
optimization (VEPSO) for multi-objective design optimization of composite structures Department of Aerospace

Engineering, Indian Institute of Science,June,2009.

10. Elisa Valentina Onet and Ecaterina Vladu, Nature inspired algorithms and artificial intelligence ,2006.
11. L.B. Zhang, C.G Zhou,X.H. Liu, Z.Q. Ma, M. Ma and Y.C. Liang, Solving Multi Objective Optimization Problems
Using Particle Swarm ptimization Jilin University, Changchun 130012, P. R. China,2003.
12. Gurdal Z, Haftka RT, Hajela P. Design and optimization of laminated composite materials. Wiley-Interscience; 1999.
13. ] Gurdal Z, Haftka RT. Optimization of composite laminates. In: Presented at the NATO advanced study institute on
optimization of large structural systems, Germany; 1991.
14. Bruyneel M. A general and effective approach for the optimal design of fibre reinforced composite structures. Compos
Sci Technol2006:130314.

15.

Shin DK, Gurdal Z, Griffin OH. Minimum weight design of laminated plates for post-buckling performance. In:

16.

Proceedings of the 32nd IAA/ASME/ASCE/AHS/ASC structures, structural dynamics and materials conference; 1991.
Rowlands RE. Strength (failure) theories and their experimental correlation: Hand book of composites fracture

mechanics of composites. In: Sih GC, Skudra AM, editors, vol. 3. Amsterdam: Elsevier, 1985:71127.
17. Benzeggagh ML, Khellil K, Chotard T. Experimental determination of Tsai failure tensorial terms Fij for unidirectional
composite materials. Comp Sci and Tech 1995;55:14556.
18. Hart-Smith LJ. Should fibrous composite failure modes be interacted or superposed. Composites 1993;24(1):5355.
19. Yamada Sun CT. Analysis of laminate strength and its distribution. J Comp Mater 1978;12:27584.
20. Hart-Smith LJ. A Reexamination of the analysis of in-plane matrix failures in fibrous composite laminates. Comp Sci
21.

and Tech 1996;56:10721.

O'Brien TK. Characterization of delamination onset and growth in a composite laminate. Damage in composite
materials, ASTM STP 1982;775:14067.
Brewer JC, Lagace PA. Quadratic Stress criterion for initiation of delamination. J Comp Mater 1988;22:1141155.