IOP Conference Series: Materials Science and Engineering

PAPER • OPEN ACCESS

Fatigue damage growth in notched laminates

To cite this article: A Muc et al 2019 IOP Conf. Ser.: Mater. Sci. Eng. 544 012038

View the article online for updates and enhancements.

This content was downloaded from IP address 193.111.186.115 on 09/07/2019 at 18:33

MEACM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 544 (2019) 012038 doi:10.1088/1757-899X/544/1/012038

Fatigue damage growth in notched laminates

A Muc1, M Chwał1 and P Kędziora1

1
Institute of Machine Design, Cracow University of Technology, Warszawska 24, 31-
155 Cracow, Poland
E-mail: olekmuc@mech.pk.edu.pl

Abstract: The issue discussed in the current paper deals with experimental and finite element
analysis of fatigue problems for notched rectangular multilayered plates. The fatigue damage
growth and fracture at holes in reinforced laminates loaded in tension is presented. The
analysis concerns the ±45°glass/epoxy laminated plates with several types of holes such as
vertical elliptical, horizontal elliptical and circular. Also a few remarks on the static analytical
solutions and fatigue live predictions are presented. Here, the initial results are only
demonstrated, which are fundamental for the further analysis. The experiments shown that
independently on the shape of the hole the fatigue process includes three stages, i.e. the matrix
cracking, the delamination between the plies and the final failure. The first stage of the failure
can be described with the help of the standard finite element codes.

1. Introduction
A great deal of research has been devoted to a study of the mechanism of fatigue, and yet there is still
not a complete understanding of the phenomenon and the effect of fatigue loading on the crack
propagation under the complex stress of boundary and loading conditions. The behavior and fatigue
degradation of notched composite laminates under static and fatigue loading exhibits many features
which usually are not observed for homogeneous materials such as metals. As the most significant
features of the fatigue phenomena in composite laminates, one can mention the following:
 during fatigue loading of a notched laminate, high axial shear stresses exist near the notch root;
these high shear stresses will cause the material in the vicinity of the notch to be degraded much
more rapidly than throughout the rest of the laminate,

Figure 1. Fatigue failure modes in notched Figure 2. Fatigue failure modes in notched
unidirectional glass/epoxy laminates (eight layers unidirectional glass/epoxy laminates (eight layers
±450) - 1300000 cycles (prior to final failure) ±450) – 1500000 cycles (prior to final failure)
– a vertical ellipse. – a horizontal ellipse.

Content from this work may be used under the terms of the Creative Commons Attribution 3.0 licence. Any further distribution
of this work must maintain attribution to the author(s) and the title of the work, journal citation and DOI.
Published under licence by IOP Publishing Ltd 1
MEACM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 544 (2019) 012038 doi:10.1088/1757-899X/544/1/012038

 frequently, the direction of crack growth in a fiber composite laminate is not perpendicular to the
applied load, even for perpendicular slit notches (figure 1); it is worth to note that that effect is
usually connected with the shape of the hole – Ref. [1] (compare figure 1 and 2),
 static failure modes may be completely different from the mode of failure observed during or after
a program of cyclic loading,
 experiments demonstrate evidently the interaction of delamination and cracks (see Ref. [2] and
figure 3).

Figure 3. Failure form (splitting) in notched Figure 4. A multilayered laminated plate with an
unidirectional glass/epoxy laminates (eight layers arbitrary shaped hole under in-plane loading.
±450) – a circular hole

In general, the problems considered herein deal with the rectangular multilayered plates having
centrally located holes and subjected to the remote tension – figure 4. Here fatigue damage growth and
fracture at holes in reinforced laminates loaded in tension have been presented. In the beginning, a few
remarks on the experimental results are demonstrated. Then, the possibilities of the stress estimations
around holes for static tension are briefly discussed. At the end of the paper, various models used for
the fatigue life predictions are studied and they are supplemented with the finite element results.

2. A short review of the fatigue life prediction methods

2.1. Static analytical solutions
For plates having circular, elliptical or irregular shaped holes under static tension (or compression), the
analytical solutions can be found with the use of the Airy function, compatibility equation and
complex functions. Lekhnitskii [3] used one mapping function for isotropic plates. Lu et al. [4]
introduced three different mapping functions for generally anisotropic plates. The broader discussion
of those problems and solutions can be found in Muc et al. [5-8].

2.2. Fatigue life prediction

Fatigue cracks appear in components after a certain number of load cycles at nominal stress levels
which are often below the tensile strength of the material. The stress-life method [9-10] was the first
approach used in an attempt to understand and quantify metal fatigue. This approach requires
extensive experimental work and does not take into account the actual damage mechanisms. The
classical S-n correlation is given by an equation of the form:
Log(n)=A-BLog(Seq); Seq=Smax(1-R)C – D; R=Smin/Smax (1)
where Smax is the maximum stress, Smin is the minimum stress and n is the life in cycles. The quantities
A, B, C and D may be viewed as material properties which have to be determined through constant
amplitude stress-controlled fatigue testing. It is common practice to define the S-n diagram for
constant R-value. Echtermeyer [11] gave the parameters of a standard R=-1 strain-life log-log S-n
curve, for mainly [0°, 90°] lay-ups of polyester and vinylester-glass laminates, and some criteria for its
use. In the case of the existence of a fatigue limit, eq. (1) is sometimes formulated as:
S = S∞ + B/nx (2)

2
MEACM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 544 (2019) 012038 doi:10.1088/1757-899X/544/1/012038

where S∞ is the fatigue or endurance limit stress – see Och [12]. Sims and Brogdon [13] used a
modified power law relationship to fit various datasets as follows:
n= [b/(S-a+cA-y)](1/x) , A=(1-R)/(1+R) (3)
and a, b, x, and y are fitting parameters. Caprino and Giorlea [14] formulated a two-parameter S-n
formulation from a strength degradation equation:
US – Smax= aSmax(1-R)(nc-1) (4)
where US means the ultimate strength. Epaarachchi and Clausen [15] have recently expanded this to a
three-parameter S-n formulation, which takes into account the dependency of stress ratio influence on
dominant fibre angle, and frequency effects:
US –Smax= aSmax(1-)1.6- |sinθ|[Smax/US] 0.6- |sinθ| ( nc-1)/fc (5)
The parameters ψ, θ, and f quantify the R-value dependence, dominant fibre angle, and frequency, resp.
In general, the above methods contain an assumed form of R-value dependency and should only be
used with care where this dependency is not experimentally verified. In most cases, the mean S-n
curve is used to describe fatigue data. However, for certification purposes it may be worthwhile to
describe the S-n data regarding percentiles other than this 50% survivability line – see, e.g., Och [12].
The next group of fatigue life prediction is directly connected with the S-n method although fatigue
life is estimated with the use of the fatigue failure criteria or equivalently damage parameter or
parameters (in general denoted by the symbol d) [16]. The fatigue life is established with the use of the
failure criteria (e.g., in the form of the Tsai-Wu criterion):
FijSiSj+FjSj=1, i,j=1,2,6 (6)
however, the coefficients are defined with the use of the three S-n curves. The application of the above
method is presented by Muc et al. [17]. In view of the above concepts dealing with the fatigue damage
analysis it is worth to mention Reifsnider’s approach–the remaining (residual) strength of a “critical
element” governs the life of the entire structure [18]. When composing his construction from
individual layers, the designer would like to have design rules that are based on strains rather than
stresses. Strains are equal for all layers (in the structural reference frame of a uniformly, uni-axially
loaded laminate – the notched problem). Therefore, cyclic strain-controlled tests can better
characterize fatigue behavior of material than cyclic stress-controlled tests, particularly in the low
cycle fatigue region and/or in notched members. The strain-life curves are often called low cycle
fatigue data because much of the data are for less than 105 cycles. The total strain amplitude can be
resolved into elastic and plastic strain components from the steady-state hysteresis loops. Both the
elastic and plastic curves can be approximated as straight lines. According to this approach, the fatigue
life of a member subjected to fully reversed constant amplitude strain-controlled loading is given by
an equation of the form:
  e  p  f
    2N 
'

  
b ' c
2N f f
(7)
f
2 2 2 E

where: Δε/2 - total strain amplitude, σf’ - fatigue strength coefficient, b - fatigue strength exponent, εf’
- fatigue ductility coefficient, c - fatigue ductility exponent, E - modulus of elasticity, 2Nf - reversals to
failure. The strain-life approach results in better predictions in cases where considerable plastic
deformation is present in the critical regions of a component. When the deformations are
predominantly in the elastic region, this approach yields the same predictions as the much simpler
stress-life approach.

3. Materials and methods

The samples applied in the fatigue tension experiments were made of 8-layer of unidirectional
glass/epoxy composite. The mechanical properties of the composite are given in table 1.

3
MEACM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 544 (2019) 012038 doi:10.1088/1757-899X/544/1/012038

Table 1. Mechanical properties of unidirectional glass/epoxy composite

Fiber E1 E2 G12
Material ν12
orientation [GPa] [GPa] [GPa]
o
Unidirectional glass/epoxy ±45 46.4 14.9 5.2 0.27

All specimens were made in autoclave under the pressure 0.4MPa and cured at the temperature 135ºC.
The dimensions of the laminated plates are shown in figure 4. The total thickness of the laminate was
equal to t=2.2 mm and the width and the height were equal to Lx=Ly=250 mm. The diameter of the
circular hole was equal to 50mm, whereas the semi-axes of the ellipses were following: vertical ellipse
a=17.5 mm, b=35.7 mm; horizontal ellipse a=35.7 mm, b=17.5 mm.
The experimental fatigue tests were performed on the MTS Landmark 370 servo hydraulic testing
machine with a FlexTest 40 controller and the MTS 793 System Software. The experimental settings
of fatigue tests are presented in table 2.

Table 2. Fatigue tests settings

Hole
Circular Elliptical horizontal Elliptical vertical
The average force Pm [kN] 44 38 50
Amplitude (Pmax-Pm) [kN] 4 4 4
Frequency f [Hz] 15 30 30

4. Results and discussion

The experiments have been carried out for three shapes of cut-outs: circular, horizontal elliptical and
vertical elliptical – figure 4 and table 2. As it may be observed the whole process of fatigue failure is
identical for all shapes of the holes (figure 5). In general, it is composed of three parts:
 the first stage I is associated with the initiation of the crack (at four points around the hole) and
further (rather rapid) development of the stiffness degradation due to the matrix cracking – see
figure 1 and 2;

3
Strain [%]

Elliptical horizontal
Circular
1 Elliptical vertical

0
0 4 8 12 16
Number of cycles nx105
Figure 5. The process of fatigue Figure 6. The final fatigue Figure 7. The propagation of
failure – the strain corresponds to failure mode of the plate with fatigue failure.
the total strain of the plate, i.e. the circular cutout.
ε=(Ly-Ly0)/Ly0 where Ly – the
actual plate length and Ly0 – the
initial plate length.
 the second stage II is connected with the gradual (almost linear) increase of the strain that is
caused by the delamination failure of the plies – figure 3;
 the last (third) III stage is connected with a rapid increase of strains and the final fatigue
failure – see figure 6.

4
MEACM 2018 IOP Publishing
IOP Conf. Series: Materials Science and Engineering 544 (2019) 012038 doi:10.1088/1757-899X/544/1/012038

With the use of the FE modeling or the analytical approach, it is possible to capture the initiation of
the first-ply-failure of the notched plate using, e.g., the Hashin or the Tsai-Wu criterion. In the stage I
the matrix degradation can also be evaluated with the aid of the FE codes - the representative results of
such an analysis are demonstrated in figure 7.
However, using the standard FE codes, it is impossible to detect the final fatigue failure of the notched
laminates since the classical codes do not incorporate the evaluation and the final failure of the mixed
modes that are connected with the simultaneous matrix cracking and delamination modes – see figure
5 and the discussion in Ref. [2]. Briefly speaking, the variety of fracture criteria (Paris, Wallker,
Collpriest, Newmann etc. – see Ref. [18]) cannot be applied herein.

5. Summary
In the paper, the problems connected with the fatigue failure process of rectangular plates made of
eight plies oriented at ±450, subjected to axial tension and having circular or two elliptical (vertical and
horizontal) holes is discussed from both experimental and theoretical point of view. It is necessary to
emphasize that the initial results are only demonstrated which will be fundamental for the further
analysis. The experiments show that independently on the shape of the hole the fatigue process
consists of three stages, i.e. the matrix cracking, the delamination between the plies and the final (very
rapid) failure – fibres cracking and debonding. The first stage of the failure can be described with the
use of the standard FE codes. The consistent characterization of the next stages of failure requires
further investigations. The possible theoretical approaches and methods of the solution of the latter
problem are also discussed herein.

References
[1] Durchlaub EC and Freeman RB 1974 Design data for composite structure safe life prediction,
AFML-TR-73-225
[2] Hallett SR, Jiang WG, Khan B and Wisnom MR 2008 Compos Sci Technol 68 80
[3] Lekhnitskii SG 1968 Anisotropic plates, New York, Gordon and Breach
[4] Lu AZ, Zhang N, Zhang XL and Li WS 2015 Int J Geomech 15
[5] Muc A 2000 Comput Struct 76 211
[6] Muc A and Ulatowska A 2012 Compos Struct 94 1413
[7] Muc A and Romanowicz P 2017 Compos Struct 178 27
[8] Muc A, Barski M, Chwal M, Romanowicz P and Stawiarski A 2018 Compos Struct 189 117
[9] Bannantine JA, Corner JJ and Handrock JL 1990 Fundamentals of metal fatigue analysis,
Prentice Hall
[10] Broek D 1989 The practical use of fracture mechanics, Kluwer Academic Publishers.
[11] Echtermeyer AT 1994 Proc EWEA Conference and Exhibition, p 391
[12] Och F 1983 Fatigue Strength, AGARD-AG-292, Helicopter Fatigue design guide, p 109
[13] Sims D and Brogdon V 1977 ASTM STP 636 185
[14] Caprino G and Giorlea G 1999 Compos A 30 299
[15] Epaarachchi JA and Clausen PD 2003 Compos A 34 313
[16] Degrieck J and Paepagem V 2001 Appl Mech Rev 54 279
[17] Muc A, Kedziora P and Krawiec Z 2011 Procedia Eng 10 1315
[18] Reifsnider K and Stinchcomb WW 1986 Composite Materials: Fatigue and Fracture, ASTM
STP 907, ed HT Hahn, American Society for Testing and Materials: West Conshohocken, PA.
p 298