Polymer 55 (2014) 1583e1592

Contents lists available at ScienceDirect

Polymer
journal homepage: www.elsevier.com/locate/polymer

A quantitative analysis of the effect of interface delamination on the

fracture behavior and toughness of multilayered propyleneeethylene
copolymer/low density polyethylene lms by the essential work of
fracture (EWF)
Guansong He, Jiang Li*, Fengshun Zhang, Fan Lei, Shaoyun Guo**
The State Key Laboratory of Polymer Materials Engineering, Polymer Research Institute of Sichuan University, Chengdu, Sichuan 610065, China

a r t i c l e i n f o

a b s t r a c t

Article history:
Received 9 September 2013
Received in revised form
23 December 2013
Accepted 26 January 2014
Available online 6 February 2014

The multilayered propyleneeethylene copolymer (CPP)/low density polyethylene (LDPE) composite

sheets were prepared by the microlayered coextrusion system. The essential work of fracture (EWF)
method was rstly used to quantitatively evaluate the fracture behavior of layered materials. The
experimental results indicated that the two-dimensional layered interfaces in the multilayered materials
could play an important role in the fracture behavior. The specic essential work of fracture, we,
increased with the layers due to interfacial delamination. Additionally, the different testing speeds had a
dual effect on the increscent trend of the specic essential work of fracture, we, with increasing layers.
2014 Elsevier Ltd. All rights reserved.

Keywords:
Multilayered coextrusion
Essential work of fracture (EWF)
Fracture behavior

1. Introduction
Multi-component materials are used in increasing quantities in
all elds of society due to the advantages of their constituent
phases. Most of these materials consist of several phases, with interfaces between the phases. An interface is a zone that is spontaneously formed or consciously engineered when phases of
different character are attached to each other, and it can play a key
role in all multi-component materials irrespective of the number
and type of their components or their actual structures, and in
some cases, it may directly determine the ultimate properties [1].
The interfacial morphologies of polymer blends, which are
prepared through conventional technology, are usually spherical or
ellipsoidal. However, these simple interfaces cannot satisfy the
demand of high-performance and functional polymer materials.
Recently, some special interfacial morphologies have been constructed to realize the high performance and functionalization. For
example, concerning the barrier property, the platelet-shaped or
layered interfaces are necessary to decrease the permeability coefcient of the materials by creating a maze or tortuous path that
retards the progress of the gas molecules through the matrix resin
* Corresponding author. Fax: 86 028 85466077.
** Corresponding author. Fax: 86 28 85405135.
E-mail addresses: li_jiang@scu.edu.cn (J. Li), nic7702@scu.edu.cn (S. Guo).
0032-3861/$ e see front matter 2014 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.polymer.2014.01.039

[2]. Regarding the strengthening properties, the ber-like interfaces are the most effective choice, as they are distributed parallel to the direction of forces. In this case, perfect interfacial
adhesion is necessary to transfer a load from the matrix to the ber
[1]. Additionally, a co-continuous morphology can be formed in the
binary polymer blends when their composition is approximately
50/50. Therefore, the interfaces also become continuous. The selective location of the conductive llers at the continuous interfaces
can greatly reduce the conductive percolation threshold of the
materials due to the special double percolation effect [3e5].
However, the formation of this continuous interface is hardly
controlled through conventional blending approaches because it is
inuenced by many rigorous conditions, such as processing parameters, ratios of volume, the viscoelasticity and the interfacial
tension of two polymers in a molten state.
Recently, a multilayered coextrusion technology has been
developed to prepare multilayered materials that contain large
continuous layered interfaces. This layering is a special cocontinuous structure, that gives the multilayered materials some
outstanding barrier, damping, conductive, optical and mechanical
properties [6e10], mainly for the synergistic effect, which combines the physical properties of the individual layers through the
layered interfaces.
Among these multilayered materials, the important effect of
two-dimension layered interfaces on mechanical properties has

1584

G. He et al. / Polymer 55 (2014) 1583e1592

been intensively studied. Hiltner and coworkers found the

microlayered polycarbonate (PC) and styrene-acrylonitrile
copolymer (SAN) materials could present ductile fracture
behavior [11e14], which was achieved by the interaction between
the crazing and shear-banding that were formed at the interface.
However, concerning the effect of interface adhesion on the
mechanical properties of a multilayered system, Julia Kerns found
that the PC/PMMA could present brittle fracture behavior due to
the higher adhesion between PC and PMMA restricting the
interface delamination during yielding [15]. Therefore, the PMMA
crazes opened up into cracks and tore into the neighboring PC
layers and nally initiated brittle fracture. Instead, PC layers could
be drawn easily because the local delamination relieved the
constraint at the PC/SAN interface. Therefore, the interface
delamination signicantly affected the mechanical property.
Moreover, the energy absorption due to interfacial delamination
is of great importance when evaluating the damage and fracture
characteristics of composites.
Jung-Seok Kim studied the low-velocity impact response of
composite laminates, and the results showed that the laminate
with the largest delamination area could absorb the highest impact
energy and revealed good impact resistance [16]. Similar experimental results were found by Tien-Wei Shyr [17]. Part of the impact
energy could be absorbed by interfacial delamination in the laminates. Regarding the energy-absorbing mechanism, the results
showed that the delamination became more important when the
layers of laminate increased, and it was concluded that the layer
number was an important parameter for the energy-absorbing
mechanism in composite laminates. Nevertheless, the previously
obtained results indicating more energy absorbed due to delamination were not quantitative, but merely qualitative.
Currently, the characterization of fracture behavior for thin
sheets and lms is very complicated due to their small thickness
and ductile behavior. The LEFM (linear elastic fracture mechanics)
approach is not adequate to characterize the failure of ductile
polymers due to a large plastic zone usually existing around the
crack tip [18]. Furthermore, the J-integral approach, which is
traditionally used for the toughness characterization of ductile
polymeric materials, is also limited because the sample preparation
procedure is quite tedious and extreme care is needed to ensure a
valid critical Jc value [19]. In this situation, the essential work of
fracture (EWF) method is increasingly being used as the most
appropriate technique to evaluate the fracture behavior of various
ductile polymeric materials, especially for samples in the form of
lms or thin sheets, due to the simplicity of the experimental
process and data manipulation [20].
Although the EWF approach has been widely applied to study
fracture mechanics, its application to multilayer polymeric lms is
rarely reported, especially regarding the inuence of layered interfaces on the toughness of materials. In addition, the traditional
fracture surface is perpendicular to the tensile direction, but the
interfacial delamination surface that is created during EWF testing
for a multilayered polymer is parallel to the tensile direction.
Additionally, the energy absorption that is associated with a new
surface that is created in a different direction can be quantied by
the surface energy item of EWF. Therefore, the applied eld of the
EWF method can be expanded.
In this paper, the multilayered polypropyleneeethylene
copolymer/low density polyethylene (CPP/LDPE) composites with
alternating CPP and LDPE layers were prepared by multilayered
coextrusion technology that was developed in our lab. The
essential work of fracture (EWF) method was used to evaluate the
fracture toughness of the lms, and the morphologies of the
fracture specimens were observed by scanning electron microscopy (SEM).

2. Theory
The EWF method that was rst proposed by Cotterell and
Reddell [21], is based on Brobergs work on stable crack growth
[22], and then has been developed by Karger-Kocsis [23,24] and
Mai [25]. Based on the energy partition [26], the total energy that
is required to fracture a pre-cracked specimen can be separated
into the essential work of fracture, We, and the non-essential or
plastic work of fracture, Wp. We represents the energy that is
involved in the creation of the two new surfaces during the crack
propagation (inner fracture process zone-IFPZ). The second term,
Wp, collects the volume energy that is dissipated during the
fracture, such as plastic deformation (outer plastic deformation
zone-OPDZ). These zones are schematized in Fig. 1 for a deeply
double edge notched tension specimen (DDENT) specimen. For a
given thickness, We is proportional to the ligament length (l)
while Wp is proportional to l2. The total fracture work, Wf, can be
expressed as follows:

Wf We Wp we lt bwp l2 t

(1)

wf Wf =lt we bwp l

(2)

where l is the ligament length, t is the specimen thickness, b is a

shape factor that is associated with the dimension of the plastic
zone and we and wp are the specic essential work of fracture and
specic non-essential work of fracture or specic plastic work,
respectively. The specic essential work of fracture has been shown
to be a material property [27], and it can be used to characterize the
fracture toughness and has been demonstrated theoretically to be
equivalent to Jc [28]. The specic plastic deformation work of
fracture, bwp, which can be termed the specic non-essential work
of fracture [29], however, is not a material constant and depends on
the shape of the plastic zone surrounding the crack, the specimen
geometry, testing speed, etc. The term we is experimentally determined by the extrapolation of wf as a function of l to zero ligament
length and bwp is considered the slope of the linear t extrapolation [25,30].
In addition, except for the original fracture surface, some new
inner surfaces could be generated inside the necking and plastic

Fig. 1. The deeply double edge notched tension specimen (DDENT) that was used in
the EWF test.

G. He et al. / Polymer 55 (2014) 1583e1592

zones during the fracture process. The surface energy absorption

will then increase, leading to an increase in the essential work of
fracture We. In this case, the total surface energy item We can be
split into We,ORI and We,INN, and Eqs. (1) and (2) can be rewritten,
respectively, as follows:

Wf We Wp We;ORI We;INN Wp


we;ORI we;INN lt wp bl2 t

(3)

wf Wf =lt we;ORI we;INN bwp l

(4)

1585

3.4. Differential scanning calorimetry (DSC)

The crystalline analysis of the samples for EWF testing after the
same heat-treatment history was performed using the DSC instrument (Q20, TA Instrument). To investigate the inuence of the
same heat treatment on the crystalline structure, the results of the
rst heating process were adopted. The samples, ranging in weight
from 7 to 8 mg, were heated at a rate of 10  C/min, and then
maintained at 200  C for 5 min.

3.5. Tensile and fracture test

where we,ORI is the specic essential work of fracture that is associated with the original fracture surface and we,INN is the specic
essential work of fracture that is associated with the increased inner fracture surface.

3. Experimental
3.1. Materials
The materials that were used in this work were the CPP and
LDPE. The CPP was EPS30R from the Du Shan Zi Petroleum Chemical Company Ltd. (China), consisting of 6.5 wt % ethylene with a
melt ow rate (measured at 230  C and 2.16 kg) of 1.5 g/10 min. The
LDPE was LD100AC from Beijing Yan shan Petrochemical Co., Ltd
(China) with a melt ow rate (measured at 190  C and 2.16 kg) of
1.9 g/10 min.

3.2. Specimen preparation

Through the multilayered coextrusion system that was designed
by our lab, the 2-, 8-, 16-, 32-, 64- and 128-layer CPP/LDPE alternating multilayer materials were prepared. The thickness and
width of the extruded layered samples were 0.5 and 50 mm,
respectively. The details of preparation method have been previously reported [31e33]. The neat CPP and LDPE sheets were prepared by one extruder with the same dimension as the
multilayered specimens. The temperature prole was in the range
of 150e200  C for the CPP and 120e190  C for the LDPE. The
preferential orientation along the extrusion direction of the CPP/
LDPE sheets could have a huge effect on the crack propagation [34].
To eliminate the effect of orientation on the testing, all of the
extruded CPP/LDPE sheets together with the pure CPP and LDPE
sheets were annealed at 190  C for 10 min.
The DDENT specimens (length  width  thickness
100  35  0.5 mm), as seen in Fig. 1, were cut from the annealed
specimens. The sharp pre-cracks on both sides of the specimens were
made perpendicular to the tensile direction with a fresh razor blade.
To meet the plane stress conditions of the specimens, their ligament
range was limited (3t < l < w/3) [35]. For each material, nearly 20
specimens were tested with different ligament lengths in the range of
approximately 3e10 mm. Before the EWF testing, the ligament
lengths and the thickness of all the specimens were measured using a
microscope and a vernier caliper, respectively.

3.3. Polarized optical microscopy (POM)

First, an approximately10-mm thin slice was obtained from the
specimen by a microtome along the extrusion direction for
morphology observation. Then, the multilayered structure morphologies were observed by POM (BX51, OLYMPUS) equipped with
a video camera.

The EWF tests of the DDENT specimens were performed using

an Instron5567 tension machine (Canton, MA, USA) with a 1000 N
load cell at a room temperature of 23  C. The crosshead speed was
set at 3, 10 and 15 mm/min. The loadedisplacement curves were
recorded, and the energy that was absorbed until failure was
calculated by the computer integration of the loadedisplacement
curves. In addition, the tensile modulus of the CPP and LDPE dumbbell shaped specimens with the same thickness of 0.5 mm was
obtained using the same tension machine that was equipped with a
small strain gauge. The speed was also set at 3, 10 and 15 mm/min.

3.6. Scanning electron microscopy (SEM)

To study the interfacial interaction between the CPP and LDPE
phases during the EWF test, the fractured specimens were cryofractured in liquid nitrogen, after which the fracture morphology
was examined by SEM (JSM-5900LV, Japan).

3.7. Density measurement

The densities of the samples were tested in a GH-120M high
precision density tester (Matsuhaku, Taiwan), the measuring accuracy of which was 0.0001 g/cm3. The measured densities of the
CPP and LDPE were 0.8670  103 and 0.9132  103 kg/m3, respectively. According to the measured density of the composite, rC, the
volume fraction of the CPP (VCPP) or LDPE (VLDPE) could be found
using the following equation:

VCPP

rC  rLDPE
rCPP  rLDPE

where, rC is the density of the composite, rCPP and rLDPE represent

the densities of the CPP and LDPE, respectively.

3.8. Oil absorption value measurement

The surfaces of a material can absorb some oil when immersed
in oil. If more surfaces are created during the fracture process, a
higher oil absorption value can be achieved. To obtain the relative
area of the surfaces that were created during the EWF test, the oil
immersion method was adopted. The samples containing the
fracture and plastic zone were weighed before hand, put into a
stainless steel mesh and then immersed into silicone oil for 3 min.
(During this time, the oil can sufciently spread on the surface and
will not diffuse into the inner structure.) The samples and the mesh
were together removed from the oil, drained for 2 min, wiped with
lter paper to remove the excess oil, and then weighed. The oil
absorption value was calculated by the following formula: Oil absorption value (mg) Weight of the absorbed oil in the sampleWeight of the sample before oil absorption.

1586

G. He et al. / Polymer 55 (2014) 1583e1592

volume fractions of the CPP and LDPE did not change distinctly with
the number of layers changing.
4.2. Loadedisplacement curves
The loadedisplacement curves as a function of the ligament
lengths during EWF tests at three different testing speeds of the
DDENT specimens for the neat CPP, LDPE and multilayered specimens are shown in Fig. 3aec, respectively. Before the EWF method
is applied to assess the toughness, some requirements should be
met, such as full ligament yielding prior to crack initiation and selfsimilar loadedisplacement curves, etc. [35]. For all of the specimens
that were studied here, after the maximum load curve, a smooth
and slow drop in the load occurred, and the ligament began to
yield. After the complete yielding of the ligament, the cracks began
to propagate in a stable manner until the nal fracture of the
specimen. In addition, the maximum load and the displacement to
fracture increased regularly with the increasing ligament length,
and each group of curves had a good self-similarity. These behaviors validated the EWF testing and ensured that the cracks propagated under similar stress conditions [36].
4.3. Fracture parameters of the EWF

Fig. 2. Polarized optical micrographs of the multilayered structure morphologies for

different layered specimens.

4. Results and discussions

4.1. Phase morphology
Fig. 2 shows the POM micrographs of the multilayered morphologies. The darker layer belongs to the CPP, which had a large
spherulite and low crystallinity, while the whiter layer belongs to
the LDPE, which had a small spherulite and high crystallinity. All of
the multilayered specimens had a laminar morphology, where the
CPP and LDPE layers were aligned alternately. However, with
greater than two layers, the layer thickness deviated slightly from
the estimated value; that is to say the layers became thinner or
thicker. This deviation is because the continuity and uniformity of
the layers is strongly dependent on the viscosity ratio and interfacial instabilities of the components during coextrusion. In the
CPP/LDPE system, as the melt viscosity of the CPP is not the same as
that of the LDPE, the layer thickness uniformity is not very good.
However, their close melt viscosities can guarantee that the layers
are continuously arranged. The volume fractions of the CPP and
LDPE that were acquired by the density measurement method are
listed in Table 1. It is clear that, for all multilayered specimens, the

Table 1
The content of the CPP and LDPE in multilayered samples.
Specimens

Density (g/cm3)

Volume fraction
of the CPP (%)

Volume fraction
of the LDPE (%)

CPP
LDPE
2L
8L
16L
32L
64L
128L

0.8670
0.9132
0.8911
0.8907
0.8914
0.8918
0.8915
0.8923

1
e
47.84
48.7
47.19
46.32
46.97
45.24

e
1
52.16
51.3
52.81
53.68
53.03
54.76

The total work of fracture, Wf, could be obtained by the integration of the loadedisplacement curve shown in Fig. 3. The plot of
wf versus the ligament length l of the different testing speeds for
the multilayered structure specimens is shown in Fig. 4aec. The
wfel diagrams gave very good linear relationships for all of the
samples at the different speeds. The values of we and bwp that were
obtained from the interception and slope of the straight lines
extrapolated to zero ligament length, together with the regression
coefcient, are listed in Tables 2e4.
From the values that are listed in Tables 2e4, the linear
regression coefcient (R2), being in most cases higher than 0.96,
showed a good linear correlation. At each testing speed, the specic
plastic deformation work of fracture, bwp, did not change much
with the layers, revealing that the layer number and interface had
little effect on the plastic deformation ability. Therefore, it is not
discussed in detail here. However, for the specic essential work of
fracture we, the effect of the layer number became very obvious.
The we values of the multilayered materials were intervenient between the values of the CPP and LDPE, but increased with the layer
number. The we value of the pure CPP was higher than that of the
pure LDPE, indicating a good resistance to crack propagation.
The upper band of mechanical property, Y, (parallel model) is
given by the rule of binary mixtures [37] as follows:

Y Y1 f1 Y2 f2

(5)

where, Y is the mechanical property of the blends, Y1 and Y2 are the

mechanical properties of components 1 and 2, respectively, and 41
and 42 are the volume fractions of components 1 and 2, respectively. This equation is applicable for models in which the components are arranged parallel to the applied stress. In this work, the
CPP and LDPE layers in the alternate multilayered structure were
arranged parallel to the tensile stress, and the phase continuity
could meet the requirement of the parallel model. As the specic
essential work of fracture, we, has been shown to be a material
property for a given sheet thickness and independent of the specimen geometry, the parallel model can be used to verify the we
value of the multilayered CPP/LDPE specimen is the linear combination of the value of the individual component. Additionally, the
interface failure was not considered. The theoretical value of we can
be obtained by Equation (5), and the results are listed in Tables 2e4.

G. He et al. / Polymer 55 (2014) 1583e1592

1587

Fig. 3. Loadedisplacement curves for the neat CPP and LDPE specimens and the multilayered structure CPP/LDPE specimens at different testing speeds: (a) 3 mm/min; (b) 10 mm/
min; (c) 15 mm/min.

The theoretical value of we did not change much with the layers and
testing speed. However, the actual values of we tended to increase
with the number of layers, and this trend did not meet the parallel
model. As the CPP and LDPE were both semicrystalline polymers,
the changes of crystallization with the layer number may affect the
mechanical property of the multilayered materials. Before the EWF
testing, all of the multilayered samples had been treated with the
same heat history. Therefore, the crystallizations of the different

samples may ideally be the same. However, this should be proven

by the DSC data. Fig. 5 shows the DSC heating curves. To investigate
the inuence of the same heat treatment on the crystalline structures, the results of the rst heating process were adopted. The
calculated results from the DSC curves are listed in Table 5. For both
the CPP and LDPE, there is little change in the melting point and
enthalpy with changes in the layer number, indicating that the
crystalline properties of the specimens with different layer

1588

G. He et al. / Polymer 55 (2014) 1583e1592

Fig. 3. (continued).

numbers were the same. Therefore, the change in we with layer

number is not caused by the crystalline properties.
On the basis of the theory of EWF, we is essentially the energy
needed to create new surfaces of a material during fracture process;
therefore, the increscent trend may be related to the creation of a

new inner surface in the fracture process. Additionally, in the CPP/

LDPE system, the CPP and LDPE are partially compatible at the
interface, so interface delamination may occur when subjected to
external stress. Furthermore, the creation of the new inner surface
that is associated with interface delamination can absorb more

Fig. 4. Specic work of fracture wf against ligament length l for the multilayered CPP/LDPE specimens at different testing speeds: (a) 3 mm/min; (b) 10 mm/min; (c) 15 mm/min.
The spots with different shapes represent the experimental data, while the straight lines are tting results.

G. He et al. / Polymer 55 (2014) 1583e1592

1589

Table 2
EWF parameters of the CPP/LDPE specimens with varied number of layers at the
testing speed of 3 mm/min.
Specimens

we (KJ/m2)

bwp (MJ/m3)

R2

Theoretical
value of we (we,ORI)

we,INN

LDPE
CPP
2L
8L
16L
32L
64L
128L

12.82
52.04
31.72
33.66
35.18
36.49
39.61
41.13

8.02
10.79
9.42
9.35
9.43
9.34
9.65
9.31

0.98
0.98
0.99
0.98
0.99
0.99
0.98
0.97

e
e
31.58
31.92
31.33
30.99
31.24
30.56

e
e
0.14
1.74
3.85
5.50
8.37
10.57

Table 3
EWF parameters of the CPP/LDPE specimens with varied number of layers at the
testing speed of 10 mm/min.
Specimens

we (KJ/m2)

bwp (MJ/m3)

R2

Theoretical
value of we (we,ORI)

we,INN

LDPE
CPP
2L
8L
16L
32L
64L
128L

13.71
51.04
32.04
35.35
37.53
38.23
42.27
44.22

7.78
10.30
9.09
9.10
9.09
9.09
9.03
9.03

0.99
0.98
0.99
0.98
0.97
0.96
0.97
0.97

e
e
31.57
31.89
31.33
31.00
31.24
30.60

e
e
0.47
3.46
6.20
7.23
11.03
13.62

energy, resulting in an increase of the fracture energy. To prove this

point, the interface morphologies of the multilayered samples after
fracture were observed by SEM. Fig. 6 shows the SEM micrographs
of 2-, 32- and 128-layer specimens at three testing speeds. The CPP
and LDPE layers could be easily distinguished when deformed by
their different plastic deformation surfaces. Compared with the
LDPE layer, the CPP layer was a little whiter due to its porous
structure after plastic deformation. Signicant delamination
occurred at the interface for all of the multilayered specimens after
fracture at three different speeds. The creation of new failure surfaces at the interface by delamination could lead to additional
surface related energy absorption in the fracture process. Therefore,
the total surface related energy term, We, increased. According to
Eqs. (3) and (4), the specic essential work of fracture, we, can be
split into we,ORI and we,INN. The item we,ORI is associated with the
original fracture surface and is equal to the theoretical value that is
calculated by the parallel mode. From the values that are listed in
Tables 2e4, the item we,ORI did not change with the layer number,
but the item we,INN increased with the layer number. It is possible
that additional new surfaces were created by interface delamination with the increasing layer number. Hence, the increscent trend
of the specic essential work of fracture, we, as a result of the item
we,INN increasing with the layer number could contribute to the

Table 4
EWF parameters of the CPP/LDPE specimens with varied number of layers at the
testing speed of 15 mm/min.
Specimens

we (KJ/m2)

bwp (MJ/m3)

R2

Theoretical
value of we (we,ORI)

we,INN

LDPE
CPP
2L
8L
16L
32L
64L
128L

13.06
51.49
31.79
32.34
32.62
33.13
36.12
37.70

7.30
9.33
8.39
8.34
8.34
8.35
8.34
8.37

0.98
0.98
0.98
0.99
0.98
0.99
0.98
0.98

e
e
31.44
31.77
31.20
30.86
31.11
30.45

e
e
0.35
0.57
1.42
2.27
5.01
7.25

Fig. 5. The rst DSC heating curves of the multilayered CPP/LDPE specimens.

increase in the resistance to crack propagation, leading to an

improvement in the toughness.
4.4. The inuence of layer number and testing speed on the fracture
toughness
Fig. 7 shows the we,INN values of the multilayered specimens
versus the layer number. For the three selected testing speeds, the
we,INN values increased with the layer number. However, with the
same layer number the we,INN values were different at the three
testing speeds, and the we,INN value was the highest at 10 mm/min.
This result indicates that the degree of increase of the we,INN values
from 2-layer to 128-layer was the greatest at 10 mm/min compared
to those at 3 mm/min and 15 mm/min.
As discussed above, we,INN is associated with the creation of new
surfaces that were generated by interface delamination. During the
process of crack propagation, the energy absorption would increase
if more surfaces were created by delamination, leading to the increase of the we,INN values. Hence, the we,INN values may be directly
related with the area of the interfacial delamination surfaces. That
is to say the larger the area of interfacial delamination, the more
additional energy could be absorbed. However, this relationship
needs to be proven by the measurement of the interfacial delamination area.
When the multilayered specimens were fractured during the
EWF testing, not all of the interfaces between the CPP and LDPE
layers were completely separated. Some of the interfaces were only
partially separated or not separated. Therefore, the direct measurement of the total interfacial delamination area became very
difcult. However, a relative measurement could be achieved
through the method of oil immersion. If a multilayered specimen
has a larger delamination area, more oil can be absorbed through
the newly generated surface. Fig. 8 shows the oil absorbed values
(OAV) of the fractured multilayered specimens with the same geometry. At a certain testing speed, the oil absorbed values increased
with the layer number, and at the same time the oil absorbed values

Table 5
The rst melting parameters of the CPP/LDPE specimens.
Specimens

Melting point
of the CPP ( C)

Melting point
of the LDPE ( C)

Enthalpy of
the CPP (J/g)

Enthalpy of
the LDPE (J/g)

2L
8L
16L
32L
64L
128L

165.9
165.6
165.7
166.3
165.6
165.8

107.7
107.3
107.4
108.0
107.4
107.6

30.4
30.6
32.0
30.4
31.9
30.3

47.6
47.1
45.7
48.3
47.7
47.3

1590

G. He et al. / Polymer 55 (2014) 1583e1592

Fig. 6. SEM micrographs of 2-, 32- and 128-layer specimens after fracture at three testing speeds of 3 mm/min, 10 mm/min and 15 mm/min. The sketch in each image indicates the
area where the picture was taken.

of the same specimen were different at the three selected testing

speeds, being greatest at 10 mm/min. Obviously, the correlation
between the oil absorbed values and the layer number and testing
speed is similar to that of the we,INN values, indicating that a larger
area of interfacial delamination leads to higher we,INC values.
Fig. 9 plots the we,INN values against the OAV for all of the
multilayered specimens, independent of the testing speed. The
straight line is the linear tting result of the experimental data, and

Fig. 7. The we,INN values of the multilayered specimens with different layer numbers.

it can nearly pass through the origin. The we,INN values have a good
linear relationship with the area of the interfacial delamination
surfaces. The linear regression coefcient, R2, is 0.99. However, from
Fig. 8, both the layer number and the testing speed can affect the
area of interfacial delamination. The effect of the layer number on
the area of interfacial delamination can be easily understood. Under
the same testing conditions, the number of interfaces increases with

Fig. 8. The oil absorbed values (OAV) of the multilayered specimens at three testing
speeds.

G. He et al. / Polymer 55 (2014) 1583e1592

Fig. 9. The item we,INN against the OAV for all of the multilayered specimens.

the layer number. If interfacial delamination occurs, the specimen

that contains more interfaces will have a larger delamination area. In
comparison, the effect of testing speed can be very complex.
According to our previous work [38], when the external load
was parallel to the interface, interfacial stress between the adjacent
layers would occur due to the difference of the mechanical
response between the layers in the multilayered system. The
interfacial stress could be balanced through the static friction at the
interface. The tensile modulus of the CPP and LDPE layers at the
different speeds is shown in Table 6. The modulus of the CPP layer
was signicantly higher than that of the LDPE layer at the three
selected testing speeds. When the CPP and LDPE layers were
stretched and deformed simultaneously, the stress in the CPP layer
was obviously different from that of the LDPE layer due to the
discrepancies of their modulus. Therefore, there was a large difference in the mechanical response between adjacent layers, and
the existing interfacial stress became very large. When the interfacial adhesion of the CPP and LDPE layers was not strong enough to
restrict the static friction that was needed to balance the interfacial
stress, interfacial delamination would occur. Moreover, with
increasing the testing speeds, the modulus of the CPP and LDPE
would both signicantly increase, and the discrepancies between
them then became larger. Therefore, the interfacial stress increased
with the testing speed, resulting in the more likely occurrence of
interfacial delamination, which could lead to an improvement in
the degree of delamination and an increase in the area of interfacial
delamination. These changes explain why the we,INN value at
10 mm/min was higher than that at 3 mm/min. However, at 15 mm/
Table 6
Tensile modulus of the CPP and LDPE layers at different tensile speeds.
Modulus

CPP (MPa)

LDPE (MPa)

Discrepancies (MPa)

3 mm/min
10 mm/min
15 mm/min

996.1  11.3
1064.4  8.3
1123.1  23.7

270.0  10.7
316.1  16.4
363.2  9.2

726.1
748.3
759.9

1591

min, the we,INN value did not increase continuously. Therefore, there
may be another reason explaining the effect on the area of interfacial delamination.
As the fracture morphology of the 2-layer sample was simple
and easy to observe, the plastic deformation behaviors of the 2layer sample at three testing speeds were investigated. From the
SEM images of the plastic zone sections of the 2-layer specimens
with the same geometry at three different testing speeds, as shown
in Fig. 10, interfacial delamination (in the dashed box) occurred in
the plastic zone, and the delamination depth reduced with the
increasing testing speed. Some literature has reported a signicant
effect of testing speed on the plastic deformation work of materials.
The bwp could decrease with the increasing testing speed [39e41].
This was because the shape factor, b, associated with the dimension
of the plastic zone was likely to change with the testing speed [42].
The plastic deformation work, bwp, of the multilayered CPP/LDPE
specimens at different testing speeds is shown in Tables 2e4. The
plastic deformation work of all of the specimens decreased with the
increasing testing speed. Thus, the area of plastic zone was reduced.
Fig. 11 shows the plastic zones in the 2-layer specimen photographs. The dashed line zone indicates where interfacial delamination occurred. The metrical area of this zone is 15.85 mm2 for
3 mm/min, 11.89 mm2 for 10 mm/min and 9.35 mm2 for 15 mm/
min. It is clear that the area of the interfacial delamination zone
decreased with the increasing testing speed. Moreover, the
decrease in the area of the interfacial delamination zone could
restrict the area of interfacial delamination, causing the we,INN
values at 15 mm/min to decrease.
Therefore, the testing speed had a dual effect on the we,INN
values. With the increasing testing speed, interfacial delamination
is more likely to occur, causing the area of delamination to increase;
however, the area of plastic deformation zone is deceased, causing
the area of the delamination to decrease instead. The relationships
can be represented by the following equation:

we;INN k*S k*n*A*B

(6)

where, k is a linear coefcient, S is the area of interfacial delamination, n is the number of layers, A is a parameter related to the
extent of interfacial delamination, and B is a parameter related to
the area of interfacial delamination zone. As discussion above, the
we,INN values have a good linear relationship with the area of
interfacial delamination. The area of interfacial delamination S is
determined by the parameters n, A and B, and both A and B are
associated with the testing speed. At the same testing condition,
the we,INN value increases with the layer number increasing. And
when the parameter n is certain, with the testing speed increasing,
the parameter A increases while B decreases oppositely. Therefore,
a maximum value of A*B may exist. And the we,INN value at the
speed of 10 mm/min was best compared to the values at the speed
of 3 and 15 mm/min.

Fig. 10. SEM images showing the plastic zone sections of the 2-layer specimens with the same geometry for the three different testing speeds.

1592

G. He et al. / Polymer 55 (2014) 1583e1592

Fig. 11. Photographs showing the plastic zones of the 2-layer specimens with the same geometry for the three different testing speeds.

5. Conclusions

References

In this paper, the fracture behavior and toughness of thin

multilayered CPP/LDPE lms were successfully evaluated using the
EWF method. The layered interfaces in the multilayered materials
can play an important role in the fracture behavior of multilayered
structure materials. When the external load was parallel to the
interface, the interfacial stress between the adjacent layers would
be created due to the difference of the mechanical response of CPP
and LDPE layers. The interfacial delamination, which resulted in an
excess energy absorption, could occur because the adjacent interfacial adhesion of the CPP and LDPE layers was not strong enough to
restrict the interfacial stress. As a result, the toughness and crack
resistance of multilayered samples increase with the increase of we.
The we value is dependent on the area of the interfacial delamination, and increased linearly with its increase. The following
conclusions can be drawn:

[1] Puknszky B. Eur Polym J 2005;41(4):645e62.

[2] Sinha Ray S, Okamoto M. Prog Polym Sci 2003;28(11):1539e641.
[3] Calberg C, Blacher S, Gubbels F, Brouers F, Deltour R, Jerome R. J Phys D Appl
Phys 1999;32(13):1517e25.
[4] Triantafyllidis KS, LeBaron PC, Park I, Pinnavaia T. Chem Mater 2006;18(18):
4393e8.
[5] Gupta M, Lin Y, Deans T, Baer E, Hiltner A, Schiraldi DA. Macromolecules
2010;43(9):4230e9.
[6] Jarus D, Hiltner A, Baer E. Polymer 2002;43(8):2401e8.
[7] Sung K, Haderski D, Hiltner A, Baer E. J Appl Polym Sci 1994;52(2):147e62.
[8] Nazarenko S, Haderski D, Hiltner A, Baer E. Polym Eng Sci 1995;35(21):1682e
7.
[9] Tangirala R, Baer E, Hiltner A, Weder C. Adv Funct Mater 2004;14(6):595e604.
[10] Nazarenko S, Hiltner A, Baer E. J Mater Sci 1999;34(7):1461e70.
[11] Haderski D, Sung K, Im J, Hiltner A, Baer E. J Appl Polym Sci 1994;52(2):121e
33.
[12] Shin E, Hiltner A, Baer E. J Appl Polym Sci 1993;47(2):245e67.
[13] Sung K, Hiltner A, Baer E. J Mater Sci 1994;29(21):5559e68.
[14] Baer E, Hiltner A, Keith H. Science 1987;235(4792):1015e22.
[15] Kerns J, Hsieh A, Hiltner A, Baer E. J Appl Polym Sci 2000;77(7):1545e57.
[16] Kim J-S, Chung S-K. Compos Struct 2007;77(4):484e92.
[17] Shyr T-W, Pan Y-H. Compos Struct 2003;62(2):193e203.
[18] Williams J. Polym Eng Sci 1977;17(3):144e9.
[19] Lee C-F, Sue H-J, Fiscus DM. Polym Test 2013;32(2):256e64.
[20] Sun X, Shen H, Xie B, Yang W, Yang M. Polymer 2011;52(2):564e70.
[21] Cotterell B, Reddel J. Int J Fract 1977;13(3):267e77.
[22] Broberg K. J Mech Phys Solids 1975;23(3):215e37.
[23] Karger-kocsis J, Barany T, Moskala E. Polymer 2003;44(19):5691e9.
[24] Karger-kocsis J, Barany T. Polym Eng Sci 2002;42(7):1410e9.
[25] Wong JSS, Ferrer-Balas D, Li RKY, Mai YW, Maspoch ML, Sue HJ. Acta Mater
2003;51(16):4929e38.
[26] Gan M, Staudinger U, Thunga M, Knoll K, Schneider K, Stamm M, et al.
Polymer 2012;53(10):2085e98.
[27] Gray R. Test protocol for essential work of fracture, version 5. Les Diablerets:
ESIS TC-4 Group; 1998.
[28] Mai YW, Powell P. J Polym Sci B Polym Phys 1991;29(7):785e93.
[29] Arkhireyeva A, Hashemi S. Polymer 2002;43(2):289e300.
[30] Gan M, Satapathy BK, Thunga M, Weidisch R, Ptschke P, Jehnichen D. Acta
Mater 2008;56(10):2247e61.
[31] Shen J, Wang M, Li J, Guo S, Xu S, Zhang Y, et al. Eur Polym J 2009;45(11):
3269e81.
[32] Xu S, Wen M, Li J, Guo S, Wang M, Du Q, et al. Polymer 2008;49(22):4861e70.
[33] Wen M, Sun X, Su L, Shen J, Li J, Guo S. Polymer 2012;53(7):1602e10.
[34] Gmez-Prez J, Santana OO, Gordillo A, Maspoch ML. Polym Eng Sci
2007;47(9):1365e72.
[35] Brny T, Czigny T, Karger-Kocsis J. Prog Polym Sci 2010;35(10):1257e87.
[36] Gong G, Xie B, Yang W, Li Z, Zhang W, Yang M. Polym Test 2005;24(4):410e7.
[37] van Veen B, van der Voort Maarschalk K, Bolhuis GK, Frijlink HW. Powder
Technol 2004;139(2):156e64.
[38] Shen J, Li J, Guo S. Polymer 2012;53(12):2519e23.
[39] Karger-Kocsis J, Czigny T, Moskala EJ. Polymer 1998;39(17):3939e44.
[40] Hashemi S. Polym Eng Sci 2000;40(1):132e8.
[41] Karger-Kocsis J, Czigny T. Polymer 1996;37(12):2433e8.
[42] Hashemi S. Polym Eng Sci 1997;37(5):912e21.

(1) When the testing speed was constant, the specic essential
work of fracture, we, increased with the layer numbers,
indicating that the toughness and crack resistance increased.
In addition, the specic plastic deformation work of fracture,
bwp, did not change with the layer numbers, revealing that
the plastic deformation ability was unchanged.
(2) With the testing speed increasing from 3 to 15 mm/min, an
initially increasing and then decreasing trend of the specic
essential work of fracture, we, was observed. The we value at
the testing speed of 10 mm/min was highest, indicating a
highest toughness and crack resistance. The specic plastic
deformation work of fracture, bwp, decreased with the
increasing testing speed.
Acknowledgments
The authors are grateful to the National Natural Science Foundation of China (51227802, 50933004, 51073099 and 51121001), the
Ministry of Education Priority Funding Areas (20110181130004),
and the Program for New Century Excellent Talents in University
(NCET-10-0593) for nancial support of this work.
Appendix A. Supplementary data
Supplementary data related to this article can be found at http://
dx.doi.org/10.1016/j.polymer.2014.01.039.