Professional Documents
Culture Documents
Journal Pre-Proof
Journal Pre-Proof
PII: S0263-8223(22)00695-X
DOI: https://doi.org/10.1016/j.compstruct.2022.115940
Reference: COST 115940
Please cite this article as: S. Bornemann, S. Henning, K. Naumenko et al., Strength analysis of
laminated glass/EVA interfaces: Microstructure, peel force and energy of adhesion. Composite
Structures (2022), doi: https://doi.org/10.1016/j.compstruct.2022.115940.
This is a PDF file of an article that has undergone enhancements after acceptance, such as the
addition of a cover page and metadata, and formatting for readability, but it is not yet the definitive
version of record. This version will undergo additional copyediting, typesetting and review before it
is published in its final form, but we are providing this version to give early visibility of the article.
Please note that, during the production process, errors may be discovered which could affect the
content, and all legal disclaimers that apply to the journal pertain.
of
a FolienwerkWolfen GmbH Guardianstraße 4, 06766, Bitterfeld-Wolfen, Germany
b Fraunhofer-Institutfü Mikrostruktur von Werkstoffen und Systemen, Walter-Hülse-Straße 1, 06120, Halle, Germany
c Otto-von-Guericke-University Magdeburg, Institute of Mechanics, PF 4120, 39016, Magdeburg, Germany
d Fraunhofer-Center for Silicon-Photovoltaics CSP Otto-Eißfeldt-Straße 12, 06120, Halle, Germany
pro
Abstract
The aim of this study is to analyze the adhesive strength of laminated glass with an interlayer material
based on ethylene-vinyl acetate (EVA) copolymer. To investigate the microstructure within the
entire cross-section of glass laminates, cryo-fracture surfaces of the samples were prepared. SEM
re-
images illustrate the boundary layers on the glass surfaces with the thicknesses within the range
from few nm to 100𝜇m. Tensile and 180◦ peel tests were performed to estimate the adhesive
strength. No significant differences in peel force values were obtained for the EVA/tin side and
EVA/air side interfaces. A direct approach to evaluate the energy of adhesion based on the results
of tensile tests and peel tests is developed. The evaluated energy of adhesion is nearly the same for
temperatures of 20◦ C and 40◦ C and increases by a factor of 2 for the temperature of -20◦ C.
lP
Key words: Float glass, EVA interlayer, peel test, energy of adhesion, cryo-fracture surface
1. Introduction
rna
Laminated structures composed from float glass skin layers and a core layer from soft polymers
like polyvinyl butyral (PVB) or ethylene-vinyl acetate (EVA) are widely used in engineering, for
example in automobile glass [1, 2], laminated glass panels [3] as well as in photovoltaic modules
[4, 5]. Laminated glass structures are frequently subjected to various quasi-static forces, such as
snow weight, wind pressure, dynamic loadings, for example hailstorm and wind gust as well as
non-stationary thermal environment [6]. As a result, viscoelastic deformation [7, 8] and ageing
processes [9] in encapsulant materials should be taken into account in design and analysis of
Jou
∗ Corresponding
author
Email addresses: steffen.bornemann@folienwerk-wolfen.de (Steffen Bornemann),
sven.henning@imws.fraunhofer.de (Sven Henning), konstantin.naumenko@ovgu.de (Konstantin
Naumenko ), matthias.pander@csp.fraunhofer.de (Matthias Pander),
nishanth.thavayogarajah@folienwerk-wolfen.de (Nishanth Thavayogarajah),
mathias.wuerkner@ovgu.de (Mathias Würkner)
laminated glass structures. Furthermore, among various mechanical characteristics, the adhesive
strength between glass layers and encapsulant material is crucial to ensure the long-term reliability.
In [10] the adhesion strength between PVB and glass is evaluated based on through-cracked tensile
tests and a finite element analysis. The results show that the energy of adhesion is temperature
independent in the range between 20◦ C and 60◦ C. In [11] pummel tests are conducted to evaluate
the interfacial adhesion capacities of various types of PVB.
of
In [12] 90◦ and 180◦ peel tests are performed to evaluate the adhesive strength of laminates with
EVA - double Polyethylene Terephthalate structures. The influences of peel angle, peel speed
and curing state of EVA after the lamination are discussed. Degradation of adhesion between
the back sheet and the EVA encapsulant in PV modules is analyzed in [13] based on 90◦ peel
pro
tests, carried out at specified time intervals. In [14–16] delamination strength for laminated glass
with interlayers from PVB and EVA and their different nominal thicknesses are compared. It
is found that inter-layer thickness has a significant influence on the delamination strength. It is
documented in [15] that after tempering, laminates with PVB interlayer have a comparatively lower
delamination strength than laminates with EVA interlayer.
The peel force and the energy of adhesion are important material parameters to calibrate cohesive
zone models, which are applied to simulate delamination processes [3, 17–21]. The peel force
re-
measured from a peel test provides only a first insight in the adhesive strength. As discussed in
many papers the peel force depends on the temperature, on the thickness of the film and the type
of peeling. These dependencies are mostly attributed to the behavior of the peel arm rather than
the adhesion properties. Therefore robust approaches are required to estimate the characteristics
of the adhesive interface, such as energy of adhesion, from the results of peel tests.
The aim of this study is to analyze the interfacial strength of a novel interlayer film, swhich is based
lP
on EVA copolymer. In addition to the previous studies on adhesion of glass-EVA we address the
following problems
• To analyze the microstructure of the entire cross-section of a polymer film in the laminate
and thus possibly existing gradients, a novel approach based on cryo-fracture surfaces of
laminate samples is developed
rna
• Tensile and peel tests are performed for the novel EVA interlayer material in a wide temper-
ature range from -20◦ C to 40◦ C. Based on the test data the energy of adhesion is estimated.
• The interfacial strength is analyzed for tin side/EVA and air side/EVA laminates
• To evaluate the energy of adhesion a direct procedure based on the results of tensile tests and
peel tests is developed.
Jou
step of the processing of EVA interlayer films. The amount of the crosslinked EVA polymer chains
is represented by the gel-rate. For achieving safety properties of laminated glasses made with EVA
laminating film, a degree of crosslinking of at least 80% is recommended. This crosslinking step
changes the properties of the EVA interlayer film and thereby that of the laminated glass. The
effect of the crosslinking on the mechanical properties of the EVA laminating film can be analyzed
by the tensile test.
of
The laminating film can be processed within a temperature range of 105◦ C to 160◦ C, and is suitable
for processing in vacuum ovens as well as in a nip roller/autoclave process. Good adhesion to the
glass, outstanding resistance to different climate conditions, durability and high transparency are
some of the important parameters to evaluate the quality of the laminated safety glass products.
pro
Mechanical stability of glass laminates with the EVA interlayer with respect to dynamic loadings,
in ball drop and pendulum tests as well as good resistance against heat, humidity and ultraviolet
radiation can be mentioned among many other properties.
Glass laminates considered in this study are manufactured by the vacuum lamination process.
This process is mainly used for the production of laminated safety glass with EVA or PVB films.
The process can be applied for the production of smaller laminated glass formats or for special
laminates, typically up to format sizes of around 2m×4m. The vacuum lamination process can also
The investigated soda-lime glass is according to DIN EN 572-1. It is flat glass produced by the
float method and especially suitable for laminated glass panels. Different surfaces of the glass sheet
should be taken into account, since one surface during manufacturing is in contact with tin (called
tin side) and the other with air (called air side). For the tin and air side different distributions of
lP
the flaws after the processing, and as a result distinct strength properties are documented. Sources
of flaws for both tin and air sides are thermal gradients experienced by the glass ribbon during the
float-glass manufacturing process. For the tin side, tin can diffuse into glass via an ion-exchange
process with silicon, and the tin penetration on the tin side may range from 10𝜇m to 50𝜇m [23].
Flaws are usually formed around tin-oxide particles adhering to the tin side surface. The mean flaw
rna
size of the air side is usually smaller by a factor of 1.5-2.0 relative to that of the tin side [24]. Flexure
strength values obtained from double ring (ring-on-ring) bending tests are presented in [25, 26].
The air side characteristic flexural strength is typically greater by a factor up to 2.0 relative to that
of the tin side. Due to different defect distributions one may expect different adhesion properties
for the tin side/EVA and air side/EVA interfaces. Therefore laminations and tests will be performed
with respect to different sides of float glass.
3. Microstructural analysis
Jou
In order to analyze the entire cross-section of a polymer film and thus possibly existing gradients
in the microstructure, cryo-fracture surfaces of the samples were prepared. For this purpose,
the film sections or sections of the laminated glass were immersed in liquid nitrogen for several
minutes. In this deeply cold state, forced fractures were then produced, which, after suitable
preparation (assembly, conductive sputter-coating with a thin conductive platinum layer), were
directly examined with the scanning electron microscope (SEM) FEI Quanta 650 ESEM-FEG
3
Journal Pre-proof
a) b)
of
pro
200𝜇m 25𝜇m
Fig. 1.
Fig. 1. SEM
SEM images
images ofof aa cryogenic
cryogenic fracture
fracture surface.
surface. a)
a) Overview
Overview with
with the
the entire
entire film
film cross-section
cross-section (middle)
(middle) and
and
adheringglass
adhering glass plates
plates (left
(left and
and right
right in
in the
the picture),
picture), b)
b) spherulitic
spherulitic texture
texture in
in the
the middle
middle ofof the
the sample
sample
operated with
operated with an
an accelerating
accelerating voltage
voltage of
re-
of 15kV.
15kV. Specimen
Specimen with with the
the glass
glass layers
layers thickness
thickness of of 3mm
3mm
and the thickness of the interlayer ranging from 0.35mm to 1.2mm were analyzed. Figure 11
and the thickness of the interlayer ranging from 0.35mm to 1.2mm were analyzed. Figure
illustrates SEM
illustrates SEM images
images ofof aa cryogenic
cryogenic fracture
fracture surface
surface including
including thethe entire
entire film
film cross-section
cross-section andand
adhering glass plates and the microstructure of the middle part. This method is also suitable for
adhering glass plates and the microstructure of the middle part. This method is also suitable
lP
for
examiningthe
examining theintact
intactcompound
compound structure
structure of
of polymer
polymer filmfilm and
and glass.
glass. In
In particular,
particular, this
thisoption
optionenables
enables
the investigation
the investigation ofof interfaces
interfaces between
between glass
glass and
and polymer.
polymer. At At high
high magnifications,
magnifications, spherulitic
spherulitic
superstructures (microstructure) and lamellar arrangements (nanostructure) could
superstructures (microstructure) and lamellar arrangements (nanostructure) could be be observed.
observed. In In
addition, particulate deposits (particulate additives, nucleation seeds) were found
addition, particulate deposits (particulate additives, nucleation seeds) were found on the fracture on the fracture
surfaces.
surfaces.
rna
This investigation
This investigation method
method hashas proven
proven toto be
be particularly
particularly suitable
suitable for
for glass
glass laminates.
laminates. On On the
the one
one
hand, an area of any size can be examined here, i.e. the entire film cross-section
hand, an area of any size can be examined here, i.e. the entire film cross-section can be viewed can be viewed
while maintaining
while maintaining the
the original
original composite.
composite. ThisThis is is of
of particular
particular interest,
interest, as
as itit enables
enables interface
interface
effects between glass and polymer, such as adhesion or delamination phenomena or
effects between glass and polymer, such as adhesion or delamination phenomena or variations
variations inin
the partially crystalline structures on the glass surfaces, to be recorded. Gradients
the partially crystalline structures on the glass surfaces, to be recorded. Gradients or orientations of or orientations of
thepartially
the partially crystalline
crystalline structure
structure occurring
occurring over
over the
the film
film cross-section
cross-section can can be
be mapped.
mapped. On Onthe
theother
other
hand, this method allows the recording and measurement of the partially crystalline
hand, this method allows the recording and measurement of the partially crystalline superstructures superstructures
such as spherulites or epitaxial effects on substrate surfaces, which are particularly important for
Jou
such as spherulites or epitaxial effects on substrate surfaces, which are particularly important for
light scattering and thus the reduction of the transparency of the composites.
light scattering and thus the reduction of the transparency of the composites.
In addition to the representation of the partially crystalline structure (crystallites, lamellae,
In addition to the representation of the partially crystalline structure (crystallites, lamellae,
spherulitic superstructures) in the case of the laminated systems, an attention was paid to the
spherulitic superstructures) in the case of the laminated systems, an attention was paid to the
formation of boundary layers with a morphology influenced by the presence of the glass surfaces.
formation of boundary layers with a morphology influenced by the presence of the glass surfaces.
It is noticeable that asymmetrical structures were formed in some preparations: the boundary layers
It is noticeable that asymmetrical structures were formed in some preparations: the boundary layers
on the glass surfaces can be significantly different in one and the same preparation, depending on
on the glass surfaces can be significantly different in one and the same preparation, depending on
44
Journal Pre-proof
a) b)
of
pro
10𝜇m 5𝜇m
Glass Glass
c)
re-
lP
5𝜇m
rna
Fig. 2. SEM images of glass-EVA interface and EVA morphology. a) glass-EVA structure, b) higher resolution image
Fig. 2. SEM images of glass-EVA interface and EVA morphology. a) glass-EVA structure, b) higher resolution image
illustrating glass-EVA interface c) spherulitic microstructure of EVA film
illustrating glass-EVA interface c) spherulitic microstructure of EVA film
the glass side viewed. All specimen show a more or less pronounced layer, which is characterized
the glass side viewed. All specimen show a more or less pronounced layer, which is characterized
Jou
by a directed growth of the crystalline lamellae starting from the glass substrate and which in some
by a directed growth of the crystalline lamellae starting from the glass substrate and which in some
cases merges abruptly and in other cases smoothly into the regular spherulitic structure, see Fig.
cases merges abruptly and in other cases smoothly into the regular spherulitic structure, see Fig.
2. The measured layer thicknesses varied between a few nm to approximately 100𝜇m, whereby
2. The measured layer thicknesses varied between a few nm to approximately 100𝜇m, whereby
these values can show considerable differences on both sides even within a sample. Figure 3 shows
these values can show considerable differences on both sides even within a sample. Figure 3 shows
variations of the interface layer for three selected specimen. There was no evidence of delamina-
variations of the interface layer for three selected specimen. There was no evidence of delamina-
tion defects on the SEM images at the boundary between the glass substrate and the film after the
tion defects on the SEM images at the boundary between the glass substrate and the film after the
cryo-fracture. Microstructural observations indicate that good interfacial strength properties are
cryo-fracture. Microstructural observations indicate that good interfacial strength properties are
5
Journal Pre-proof
of
5𝜇m 5𝜇m 5𝜇m
pro
Glass Glass Glass
Fig. 3.
Fig. 3. Variations
Variations of
of glass-EVA
glass-EVA interface
interface layer thickness in three samples
expectedfor
expected for the
the considered
considered glass
glass laminates
laminates with EVA interlayer.
4.4. Tensile
Tensile and
and peel
peel tests
tests
Tensile properties
Tensile
universal
properties were
testing
were determined
machine
determined in
with
universal testing machine with a temperature
and width
and width of of 25mm
25mm were were prepared
a
prepared for
re-
in aa tensile test based on DIN EN 527-3 in a Zwick
temperature
for the
chamber. Film strips with
the tensile testing. The thickness of the films
the length
Zwick Z050
of 150mm
films was in
Z050
150mm
in the
the
range 0.32mm - 0.38mm. Tests
range 0.32mm - 0.38mm. Tests were performed were performed fore three temperature levels (-20 ◦◦
C, 20 ◦◦
C and
C and
40◦◦C)
40 C)and
andthree
three constant
constant testtest speeds
speeds (6, (6, 60 and 600mm/min) for each temperature level. For For room
room
lP
temperature measurements the initial length of specimen was 100mm. It was reduced to 50mm for
temperature measurements the initial length of specimen was 100mm. It was reduced to 50mm for
thetests
the testsin
inthe
thetemperature
temperature chamber.
chamber. For For each
each temperature
temperature and and strain
strain rate
rate levels
levels 10 10 specimen
specimen were were
tested.
tested.
Thepeel
The peel tests
tests were
were alsoalso carried
carried out out in
in the
the Zwick
Zwick Z050 Z050 machine.
machine. The The foil
foil isis peeled
peeled directly
directly from
from
the glass with an angle of 180 . The film was laminated onto the glass and a predefined area was
the glass with an angle of 180 . The film was laminated onto the glass and a predefined area
◦
◦ was
separated with
separated with aa Teflon
Teflon film,film, asas illustrated
illustrated in in Fig.
Fig. 4. 4. From
From oneone plate,
plate, test
test strips
strips with
with aa width
width of of
rna
temperaturelevels
temperature levels of of -20
-20◦ C,
C, 20
20◦ C C and
and 4040◦◦C.
C. For
For each
each temperature
temperature levellevel 1010 specimen
specimen were were peeled
peeled
from the tin side and 10 specimen from the air side of float glass. The peel tests were performed at
from the tin side and 10 specimen from the air side of float glass. The peel tests were performed at
aacross-head
cross-headspeed speed of of 100mm/min.
100mm/min. Figure Figure 55 shows
shows the the experimental
experimental setup setup ofof tensile
tensile and
and peel
peel tests.
tests.
Figure 66 shows
Figure shows the the experimental
experimental engineering
engineering stressstress vs.
vs. engineering
engineering strain
strain curves
curves forfor EVA
EVA strips
strips
obtained under different temperatures and loading rate of 60mm/min. The biggest scatter of
obtained under different temperatures and loading rate of 60mm/min. The biggest scatter of data
data
is observed in the high strain range before the fracture. Figure 7 illustrates
is observed in the high strain range before the fracture. Figure 7 illustrates tensile curves obtained tensile curves obtained
atatroom
room temperature
temperature under under three
three loading
loading rates.
rates. A A weak
weak dependence
dependence of of the
the tensile
tensile behavior
behavior on on the
the
6
Journal Pre-proof
a) b)
250mm 250mm
25mm 25mm
55mm
of
110mm
125mm
250mm
pro
Glass
Teflon
EVA re-
Fig. 4. Peel test specimen. a) For testing at room temperature b) for testing in temperature chamber
Fig. 4. Peel test specimen. a) For testing at room temperature b) for testing in temperature chamber
displacement rate
displacement rate within
within the
the considered
considered range
range from
from 6mm/min
6mm/min to to 600mm/min
600mm/min is is detected.
detected. Figure
Figure
lP
8 shows force-displacement curves obtained from peel tests for three temperature levels and four
8 shows force-displacement curves obtained from peel tests for three temperature levels and four
strips for each temperature. For temperature levels of -20◦ C and 40◦ C specimen with shorter peel
strips for each temperature. For temperature levels of -20◦ C and 40◦ C specimen with shorter peel
length were applied, see Fig. 4b), leading to lower maximum displacement values. For -20◦◦C the
length were applied, see Fig. 4b), leading to lower maximum displacement values. For -20 C the
results of peel tests show relatively large scatter. Figure 9 provides a comparison of the determined
results of peel tests show relatively large scatter. Figure 9 provides a comparison of the determined
peel forces. The differences in adhesion between the air and tin sides at room temperature and at
peel◦ forces. The differences◦in adhesion between the air and tin sides at room temperature and at
40 C are negligible. At -20 C the peel force for the EVA/tin side interfaces exhibit slightly lower
rna
40◦ C are negligible. At -20◦ C the peel force for the EVA/tin side interfaces exhibit slightly lower
peel forces compared to EVA/air side interfaces. However, at -20◦◦C the big scatter of experimental
peel forces compared to EVA/air side interfaces. However, at -20 C the big scatter of experimental
data is observed. Because of nearly the same peel force values no difference will be made between
data is observed. Because of nearly the same peel force values no difference will be made between
the tin and air sides of glass in evaluating the energy of adhesion.
the tin and air sides of glass in evaluating the energy of adhesion.
5. Energy of adhesion
5. Energy of adhesion
The adhesion characteristics of the film play an important role in the strength analysis of glass
The adhesion characteristics of the film play an important role in the strength analysis of glass
Jou
laminate structures. The experimental investigation of the strength is carried out with the help
laminate structures. The experimental investigation of the strength is carried out with the help
of fixed-arm peel tests. In this case one of the two peeling arms has an unchangeable (fixed)
of fixed-arm peel ◦tests. In this case one of the two peeling arms has an unchangeable (fixed)
peeling angle of 0 , while the peeling angle of the second peeling arm is variably adjustable [27].
peeling angle of 0◦ , while the peeling angle of the second peeling arm is variably adjustable [27].
The force-displacement diagrams (peel curves) recorded from the fixed arm peel tests are used to
The force-displacement diagrams (peel curves) recorded from the fixed arm peel tests are used to
determine the peel force 𝐹peel . Furthermore, the fracture mechanics parameters are required for
determine the peel force
the characterization peel . Furthermore,
of the𝐹peeling process andthe
for fracture mechanics
the simulation of theparameters are required
laminate behavior. for
In this
the characterization of the peeling process and for the simulation of the laminate behavior.
paper, the method for determining the adhesion energy is discussed on the basis of experimental In this
paper, the method for determining the adhesion energy is discussed on the basis of experimental
77
Journal Pre-proof
a) b)
of
pro
re-
lP
data from the 180 ◦ peel tests and the stress-strain curves of EVA films.
rna
data from the 180 ◦ peel tests and the stress-strain curves of EVA films.
Figure 10 illustrates the configuration of the film for the case of the fixed arm peel test with a peel
Figure
angle10of illustrates
180◦ . During the configuration
the experiment, ofthe
thepath
filmofforthe
thepeeling
case ofarm
the𝑢fixed arm peel
is specified andtest
thewith
forcea peel
𝐹
angle of 180 ◦ . During the experiment, the path of the peeling arm is specified
is measured. Furthermore, the change in the length of the peel arm Δ𝑙 and the crack length 𝑎 can 𝐹
𝑢 and the force
is measured.
be determined. Furthermore,
To determinethethe
change in in
change thelength
lengthΔ𝑙,ofexperimental
the peel armdataΔ𝑙 and
fromthe cracktests
tensile length
on film𝑎 can
bestrips
determined.
are also To determine
available. the change
Figure 11 showsin force-displacement
length Δ𝑙, experimentalcurvesdata from tensile
obtained from peeltests on at
tests film
strips
an ambient temperature of 40 C. In addition, the force-displacement curves determined from the at
are also available. Figure◦ 11 shows force-displacement curves obtained from peel tests
antensile
ambient temperature
tests at the sameof 40◦ C. In addition,
temperature are plotted.theBased
force-displacement curves
on the experimental determined
data, presented in fromFig.the
tensile tests
11, the peelattest
the can
samebetemperature
divided into are
the plotted.
followingBasedranges: on the experimental data, presented in Fig.
Jou
11, the peel test can be divided into the following ranges:
• Deformation of the peel arm. In this range, the force-displacement diagrams determined
from the peel
• Deformation oftests and the
the peel tensile
arm. testsrange,
In this coincide.
the Slight deviations can diagrams
force-displacement be attributed to the
determined
different test speeds (100mm/min in peel tests and 60mm/min in tensile tests).
from the peel tests and the tensile tests coincide. Slight deviations can be attributed to the
different
• Peelingtest speeds
process. (100mm/min
After reaching theinpeel
peelforce
tests𝐹and 60mm/min in tensile tests).
peel , the separation occurs. A plateau appears
• Peeling process. After reaching the peel force 8 𝐹 , the separation occurs. A plateau appears
peel
8
Journal Pre-proof
stress, MPa
40
-20◦ C
-20◦ C
-20◦ C
-20◦ C
20◦ C
of
30
20◦ C
20◦ C
20◦ C
40◦ C
40◦ C
20 40◦ C
pro
40◦ C
10
Fig.
0
Fig. 6.
6. Stress-strain
200
Stress-strain curves
curves for
for EVA
re-
EVA strips
400
strain, %
strips for
for the
the loading
loading rate
600
rate 60mm/min
60mm/min and
and different
800
differenttemperatures
temperatures
lP
on the force-displacement diagram. The displacement of the peel arm 𝑢 is the sum of the
on the force-displacement diagram. The displacement of the peel arm 𝑢 is the sum of the
change in length Δ𝑙 and the crack length 𝑎.
change in length Δ𝑙 and the crack length 𝑎.
In addition to the peel force, the adhesion energy 𝛶 represents an important parameter for the bond
In additionIntoorder
strength. the peel force, the adhesion
to determine this valueenergy
from the 𝛶 represents an important
test data, the parameter
energy balance for the
equation forbond
the
strength. In order
peeling process to be
can determine this[18,
formulated value from the
28–31]. testprovides
This data, thetheenergy balance
fracture equation
mechanics for the
criterion:
peeling process can
crack propagation be formulated
(delamination) [18, when
occurs 28–31]. the This
rate ofprovides the fracture
energy release mechanics
has reached criterion:
a critical level
rna
crack propagation
(adhesion energy).(delamination)
In what follows occurs when
we will the rate
explain theofdirect
energy release has
approach reached the
to compute a critical
energylevel
of
(adhesion energy).
adhesion from In what follows
experimental data of we
the will
180 explain
◦ the In
peel test. direct approach
the case to compute
that the peel arm the energy of
is absolutely
adhesion
rigid (thefrom experimental
deformation data ofsmall)
is negligibly the 180 ◦ peel test. In the case that the peel arm is absolutely
Rivlin has derived the following formula [28]
rigid (the deformation is negligibly small) Rivlin2𝐹has peel
derived the following formula [28]
𝛶= , (1)
2𝐹𝑏peel
where 𝑏 is the width of the film. Taking into 𝛶= account (1)
𝑏 the small elastic deformation of the peeling
,
arm, Kendall derived the following equation in [29]
Jou
where 𝑏 is the width of the film. Taking into account the 2 small elastic deformation of the peeling
arm, Kendall derived the following equation 2𝐹 1 𝐹peel
𝛶 = in [29]
peel
+ , (2)
𝑏 2 𝑏2 𝐸 ℎ
2
where 𝐸 is the Young’s modulus of the film 2𝐹peel
material1 𝐹andpeel ℎ is the film thickness. In the case of
large elastic deformation of the peel arm, 𝛶 =the following+ , (2)
𝑏 2 𝑏 2formula
𝐸ℎ is developed in [31]
2𝐹peel 𝐹peel
where 𝐸 is the Young’s modulus 𝛶of=the film +material 𝜀 −and is the film thickness. In the case(3)
ℎ𝑊ℎ(𝜀), of
𝑏 𝑏
99
Journal Pre-proof
stress, MPa
30
6mm/min
60mm/min
600mm/min
of
20
pro
10
20◦ C
0
0 200 400 600 800
strain, %
Fig. 7.
Fig. 7. Stress-strain
re-
Stress-strain curves
curves for
for EVA
EVA strips
strips at
at 20
20◦◦ C
C and
and different
different rates
rates
whereelastic
large 𝜀 is the engineering
deformation strain
of the and
peel arm, the is
𝑊 (𝜀) the strainformula
following energyisdensity. Assuming
developed in [31] small linear-
elastic deformation of the peel arm from Eq. (3) follows Eq. (2).
2𝐹peel 𝐹peel
lP
In order to compute the adhesion𝛶energy = 𝛶 from
+ the𝜀 available
− ℎ𝑊 (𝜀),experimental data, Eq. (3) can (3) be
transformed as follows 𝑏 𝑏
where 𝜀 is the engineering strain and 𝑊 (𝜀)2𝐹 is the strain energy
𝐹peel density. Assuming small linear-
peel
elastic deformation of the 𝛶 =peel
𝛶1 +arm𝛶2 ,from
𝛶1 Eq.
= (3) follows
, 𝛶2 =Eq. (2).𝜀 − ℎ𝑊 (𝜀) (4)
𝑏 𝑏
In order to compute the adhesion energy 𝛶 from the available experimental data, Eq. (3) can be
The part 𝛶1 as
transformed corresponds
follows to the adhesion energy under Rivlin’s assumption of an absolutely rigid
rna
peel arm and can be calculated for the given 2𝐹peel force. The
𝐹peelcorrection part 𝛶2 results from the
peel
complementary energy𝛶for= the 𝛶1 +large
𝛶2 , elastic
𝛶1 =deformation., 𝛶2 Figure
= 12
𝜀 −illustrates
ℎ𝑊 (𝜀) the stress-strain curve(4)
of EVA and the averaged peel force at 40◦ C. The 𝑏 significant 𝑏deformation of the peel arm of more
The
thanpart
300% corresponds
𝛶1 before to of
the start thepeel
adhesion
processenergy
can beunderseen. Rivlin’s
Thereforeassumption
the Kendallof formula
an absolutely rigid
(2) cannot
peel armfor
be used andthecan be calculated
evaluation for the of
of the energy given peel force.
adhesion. For theThe correction
stress value ofpart 𝛶2 results
the film strip 𝜎∗from
= 𝑏ℎthe,
𝐹peel
complementary energy for the large elastic deformation. Figure 12 illustrates the
the corresponding strain 𝜀 ∗ is determined from the stress-strain curve. The strain energy density stress-strain curve
of
𝑊 EVA
(𝜀 ∗ ) isand
thethe
areaaveraged peel
under the force at 40curve
stress-strain
◦ C. The significant deformation of the peel arm of more
and is computed as follows
Jou
than 300% before the start of peel process can be seen. Therefore the Kendall formula (2) cannot
be used for the evaluation of the energy of adhesion. ∫𝜀∗ For the stress value of the film strip 𝜎 = 𝐹peel ,
∗ 𝑏ℎ
the corresponding strain 𝜀∗ is determined ∗ ) = the 𝜎(𝜀)d𝜀
𝑊 (𝜀from stress-strain curve. The strain energy density (5)
𝑊 (𝜀 ∗ ) is the area under the stress-strain curve and 0 is computed as follows
force/width, N/mm
5
-20◦ C
-20◦ C
-20◦ C
4 -20◦ C
20◦ C
of
20◦ C
20◦ C
20◦ C
3 40◦ C
40◦ C
40◦ C
pro
40◦ C
2
Fig.
0
Fig. 8.
0 200
8. Force-displacement
Force-displacement curves
curves for
re-400
for EVA/tin
displacement, mm
EVA/tin side
side glass
600
glass strips
strips from
from peel
peel test
800
test at
atdifferent
differenttemperatures
temperatures
Then the correction term of the adhesion energy can be evaluated as follows
Table 1. Peel force and maximum strain of the peel arm for EVA/glass interface at three temperature levels
rna
𝛶2 = 𝑊compl ℎ (7)
Figures
Then the13 and 14 show
correction thethe
term of experimental
adhesion energy data can
determined fromasthe
be evaluated tensile tests and peel tests
follows
for temperatures of 20 C and -20 C, respectively. With the decrease of the temperature level the
◦ ◦
peel can
tests forcebeincreases,
observed,while the 14a.
see Fig. strainTherefore
of the peelboth
armthedecreases.
minimum The
andmaximum
the maximum strainvalues
of theofpeel
the
arm force
peel beforewill
the be
separation
used. Thetakes the value of
experimental data215% at peel
of the 20 Cforce
◦ and and
100% theatstrain
-20 C.
◦
of Therefore
the peel armlarge
for
deformation of the peel arm should
three temperature values are collected in Tablebe considered in evaluation the energy of adhesion for all
considered temperatures. ◦ 1.significant scatter of experimental data from peel
Based on the experimentalFor peeland
data tests at -20
with EqsC,(4) – (6) the representative values of the adhesion
tests can be observed, see Fig. 14a. Therefore both the minimum and the maximum values of the
energy were computed. Table 2 provides a summary of the results. It can be seen that with
peel force will be used. The experimental data of the peel force and the strain of the peel arm for
the increase in temperature, the values of the peel force, and thus the Rivlin adhesion energy 𝛶1
three temperature values are collected in Table 1.
decrease. At the same time, the complementary energy, and thus the correction term 𝛶2 , increases.
11
11
Journal Pre-proof
20◦ C
4 20◦ C
4 40◦ C
40◦ C
-20◦ C
N/mm
-20◦ C
force/width,N/mm
3
3
of
peelforce/width, 2
2
pro
peel
0
0
air side tin side
air side tin side
Fig. 9.
Fig. 9. Peel force
force for EVA/glass
EVA/glass strips for
for three temperature
temperature levels at different sides
Fig. 9.Peel
Peel forcefor
for EVA/glassstrips
strips forthree
three temperaturelevels
levelsatatdifferent
differentsides
sides
re-
Δ𝑙
𝑙
𝑙
𝐹 Δ𝑙
𝐹
lP
𝑢 𝑎
𝑢 𝑎
Fig. 10. Reference and actual configuration of the film in the 180◦◦peel test
Fig. 10. Reference and actual configuration of the film in the 180 peel test
Fig. 10. Reference and actual configuration of the film in the 180◦ peel test
Based on the experimental data and with Eqs (4) – (6) the representative values of the adhesion
rna
Based on the experimental data and with Eqs (4) – (6) the representative◦ values of◦ the adhesion
As a result,
energy werethecomputed.
values of the adhesion
Table energya 𝛶summary
2 provides for the temperatures of 20
of the results. It Ccanandbe40seen
C are
thatalmost
with
energy were computed. Table 2 provides a summary of the results. It can be seen that with
the same, although the values of the peel force are different. Therefore,
the increase in temperature, the values of the peel force, and thus the Rivlin adhesion energy 𝛶the correction term 𝛶
the increase in temperature, the values of the peel force, and thus the Rivlin adhesion energy 𝛶121
related
decrease.to the
At complementary
the same time, the energy due to largeenergy,
complementary deformation of the correction
and thus peel arm must termbe𝛶 considered.
, increases.
decrease. At the same time, the◦complementary energy, and thus the correction term◦𝛶22 , increases.
For
As athe temperature
result, the valueslevel of -20
of the C, a significant
adhesion energy 𝛶 forincrease in the adhesion
the temperatures of 20energy
◦ C and can
40 C beare
observed.
almost
As a result, the values of the adhesion energy 𝛶 for the temperatures of 20◦ C and 40◦ C are almost
the same, although the values of the peel force are different. Therefore, the correction term 𝛶2
the same, although the values of the peel force are different. Therefore, the correction term 𝛶2
related
6. to the complementary energy due to large deformation of the peel arm must be considered.
Conclusions
related to the complementary energy due to large deformation of the peel arm must be considered.
For the temperature level of -20 ◦◦C, a significant increase in the adhesion energy can be observed.
Jou
For the temperature level of -20 C, a significant increase in the adhesion energy can be observed.
The aim of this study was to analyze the adhesive strength of EVA/glass interfaces. To investigate
the
6. microstructure
Conclusions of the film within the entire cross-section, cryo-fracture surfaces of the samples
6. Conclusions
were prepared. SEM images illustrate the boundary layers on the glass surfaces with the thicknesses
The
withinaimtheof this study
range was to analyze the adhesive strength of EVA/glass
can showinterfaces. To differences
investigate
The aim of this from
studyfew wasnm to 100𝜇m.
to analyze The thickness
the adhesive values
strength of EVA/glass considerable
interfaces. To investigate
theboth
on microstructure of the film
sides even within within the entire cross-section, cryo-fracture surfaces of the samples
the microstructure of◦ the afilm
sample.
within the entire cross-section, cryo-fracture surfaces of the samples
were
Tensileprepared.
tests andSEM180 images
peel illustrate
tests werethe boundarytolayers
performed on the
estimate glass
the surfaces
adhesive with theIn
strength. thicknesses
peel tests
were prepared. SEM images illustrate the boundary layers on the glass surfaces with the thicknesses
within the deformation
significant range from few of nm
the to 100𝜇m.
peel arm Thetothickness
(up 380%) isvalues can show
detected. considerable
Approximately the differences
same peel
within the range from few nm to 100𝜇m. The thickness values can show considerable differences
12
12
12
Journal Pre-proof
𝐹
, N/mm
3 𝑏
peel test
peel test
peel test
of
tensile test
2 tensile test
pro
1
40◦ C
0
0 100
re- 200
𝑢, mm
300 400
Tableand
Table 1. Peel force 2. Energy
maximumof adhesion for EVA/glass
strain of the interface
peel arm for at different
EVA/glass interfacetemperatures
at three temperature levels
on bothwere
forces sidesobtained
even within
for thea sample.
EVA/tin side and EVA/air side interfaces.
Tensile
A directtests and 180
approach
◦ peel teststhe
to evaluate were performed
energy to estimate
of adhesion basedthe
onadhesive strength.
the results In peel
of tensile teststests
and
significant deformation of the peel arm (up to 380%) is detected. Approximately
peel tests is presented. An advantage of the proposed procedure is that a finite element analysisthe same peel
forces
of wereprocess
peeling obtained
is for
notthe EVA/tinThe
required. sideestimated
and EVA/air sideof
energy interfaces.
adhesion is approximately the same
A direct approach to evaluate the energy of adhesion
for temperatures of 20 C and 40 C and increases by factor of
◦ ◦ based on2the
forresults of tensile tests
the temperature of -20and◦ C.
Jou
peel results
The tests is show
presented.
that the Ancorrection
advantage term
of the𝛶2proposed
related toprocedure is that a finite
the complementary element
energy dueanalysis
to large
of peeling process is not required. The estimated energy of adhesion is
deformation of the peel arm must be considered in evaluation of the adhesion energy. approximately the same
for temperatures
The obtained valuesof 20 ◦ C and
of the 40 C and
◦
adhesion increases
energy are theby factor
basis of 2 forand
of current thefuture
temperature
researchofon -20glass
◦ C.
The resultssubjected
laminates show that the correction
to static termloadings.
and dynamic 𝛶2 related to the complementary energy due to large
deformation of the peel arm must be considered in evaluation of the adhesion energy.
The obtained values of the adhesion energy are the basis of current and future research on glass
laminates subjected to static and dynamic loadings.
13
13
Journal Pre-proof
𝐹
, MPa
𝑏ℎ
4
stress-strain diagram
𝐹peel
𝑏ℎ
of
3
complementary energy
pro
2
40◦ C
0
0 1
re- 2
𝜀
3
𝜀∗
4 5
Fig. 12.
Fig. 12. Stress-strain
Stress-strain diagram
diagram and
and complementary
complementary energy
energy evaluation
evaluation for
for EVA
EVA at
at 40
40◦CC
◦
lP
𝑇,◦ C 𝛶1 , kJ/m2 𝛶2 , kJ/m2 𝛶, kJ/m2
Declaration of Competing Interest
-20 6.79. . . 7.75 0.69. . . 1.07 7.47. . . 8.81
20 have no3.14
The authors declare that they 0.928financial interests
known competing 4.07 or personal relationships
that could have appeared to40influence2.45 1.52 in this paper.
the work reported 3.97
rna
Acknowledgements
Table 2. Energy of adhesion for EVA/glass interface at different temperatures
The research reported in this paper has been supported by EU - EFRE (Europäischer Fonds für
regionale Entwicklung)
Declaration of CompetingSachsen-Anhalt
Interest (Grant No. 1904/00064/00065/00066). The support is
gratefully acknowledged.
The authors declare that they have no known competing financial interests or personal relationships
that could have appeared to influence the work reported in this paper.
References
Jou
[1] C. Brokmann, C. Alter, S. Kolling, Experimental determination of failure strength in automotive windscreens
Acknowledgements
using acoustic emission and fractography, Glass Structures & Engineering 4 (2019) 229–241.
[2] M. Haldimann,
The research A. Luible,
reported M. Overend,
in this paper hasStructural use of glass, by
been supported volume
EU 10, Iabse, 2008.
- EFRE (Europäischer Fonds für
[3] J. Pelfrene, J. Kuntsche, S. Van Dam, W. Van Paepegem, J. Schneider, Critical assessment of the post-breakage
regionale Entwicklung) Sachsen-Anhalt (Grant No. 1904/00064/00065/00066). The support
performance of blast loaded laminated glazing: experiments and simulations, International Journal is
of Impact
gratefully acknowledged.
Engineering 88 (2016) 61–71.
[4] M. Aßmus, K. Naumenko, H. Altenbach, A multiscale projection approach for the coupled global–local structural
analysis of photovoltaic modules, Composite Structures 158 (2016) 340–358.
14
14
Journal Pre-proof
a) b)
𝐹 𝐹
, N/mm , MPa
𝑏 𝑏ℎ
6 6
peel test stress-strain diagram
peel test 𝐹peel
peel test 𝑏ℎ
tensile test
of
4 tensile test 4
pro
2 2
20◦ C 20◦ C
0 0
0 200 400 600 0 1 2 3
𝑢, mm 𝜀
Fig.13.
13.Experimental
Experimentaldata
datafrom
fromtensile
tensileand
andpeel
peeltests
testsfor
forEVA
EVAatat20
20◦◦C.
C. a)
a) Force-displacement
Force-displacement curves,
curves, b)
b) stress-strain
stress-strain
Fig.
diagram
a)
𝐹
References
𝑏
and
, N/mm
peel force value used to evaluate the re-
complementary energy
diagram and peel force value used to evaluate the complementary energy
b)
𝐹
𝑏ℎ
, MPa
60 20
peel test
[1] C. Brokmann, C. Alter, S. Kolling, Experimental determination stress-strainof failurediagram
strength in automotive windscreens
peel test
usingpeelacoustic emission and fractography, Glass Structures & Engineering 𝐹peel 4 (2019) 229–241.
lP
max
test
[2] M. Haldimann,
peel test A. Luible, M. Overend, Structural use of glass, volume 𝐹peel 10, Iabse, 2008.
𝑏ℎ
min
15
[3] J. Pelfrene, J. Kuntsche, S. Van Dam, W. Van Paepegem, J. Schneider,
tensile test 𝑏ℎ Critical assessment of the post-breakage
tensile test
40
performance of blast loaded laminated glazing: experiments and simulations, International Journal of Impact
Engineering 88 (2016) 61–71.
[4] M. Aßmus, K. Naumenko, H. Altenbach, A multiscale projection 10 approach for the coupled global–local structural
analysis of photovoltaic modules, Composite Structures 158 (2016) 340–358.
[5]
20 K. Naumenko, V. A. Eremeyev, A layer-wise theory of shallow shells with thin soft core for laminated glass and
rna
a) b)
𝐹 𝐹
, N/mm , MPa
𝑏 𝑏ℎ
60 20
peel test stress-strain diagram
peel test 𝐹peelmax
peel test
𝑏ℎ
peel test 𝐹peelmin
15
of
tensile test 𝑏ℎ
40 tensile test
10
pro
20
-20◦ C -20◦ C
0 0
0 20 40 60 80 100 0 0.5 1 1.5 2
𝑢, mm 𝜀
Fig.14.
Fig. 14.Experimental
diagram
Experimentaldata
and peel force
datafrom
fromtensile
values
tensileand
used to
andpeel
peeltests
evaluate the re-
testsfor
forEVA
EVAat
at -20
complementary
-20◦◦C.
diagram and peel force values used to evaluate the complementary energy
debonding analysis
C. a)
a) Force-displacement
energy
Force-displacement curves, b) stress-strain
mini-modules duringofdamp-heat
laminated glass, Theoretical
exposure, ProgressandinApplied FractureResearch
Photovoltaics: Mechanics 108Applications
and (2020) 102668. 22 (2014)
[18] 796–809.
M. Nase, M. Rennert, K. Naumenko, V. A. Eremeyev, Identifying traction–separation behavior of self-adhesive
[14] A.polymeric
Vedrtnam, films from in Experimental
S. Pawar, situ digital images under t-peeling,
and simulation studiesJournal of theand
on fracture Mechanics
adhesionand testPhysics of Solids
of laminated 91
glass,
lP
(2016) 40–55.
Engineering Fracture Mechanics 190 (2018) 461–470.
[19] A.
[15] J. Pelfrene,
Vedrtnam, S. Van Dam, W. Van
Experimental andPaepegem,
simulationNumerical
studies onanalysis of the peel
delamination test for
strength ofcharacterisation
laminated glassofcomposites
interfacial
debonding
having in laminated
polyvinyl glass,
butyral and International
ethyl vinyl acetateJournal of Adhesion
inter-layers and Adhesives
of different 62 (2015) 146–153.
critical thicknesses, Defence Technology
[20] 14
S.(2018)
Pettersson, J. Engqvist, S. Hall, N. Toft, H. Hallberg, Peel testing of a packaging material laminate studied
313–317.
[16] A. Vedrtnam, C.tomography
by in-situ x-ray Bedon, M. and cohesive S.
A. Youssef, zone modeling, Effect
Chaturvedi, International Journal oftemperature
of non-uniform Adhesion and Adhesives
exposure 95
on the
(2019) 102428.
out-of-plane bending performance of ordinary laminated glass panels, Composite Structures 275 (2021) 114517.
[21] W.
[17] H.Gao,
Yin, X.
L. Liu,
Liang, Y. Wei,T.Z.
S. Chen, Q.Peng,
Bui, S.S.Yoshimura,
Chen, Determination
A cohesiveofzone the based
interface properties
DE/FE couplingin an elastic film/substrate
approach for interfacial
rna
[25] H.
[21] M.Yin,
H. L.
Krohn,
Liang, J. Y.
R.Wei,
Hellmann,
Z. Peng, D.S.L.Chen,
Shelleman, C. G. Pantano,
Determination G. E. Sakoske,
of the interface properties Biaxial flexurefilm/substrate
in an elastic strength and
dynamic fatigue of soda–lime–silica float glass, Journal of
system, International Journal of Solids and Structures 191 (2020) 473–485.the American Ceramic Society 85 (2002) 1777–1782.
[26] M. H. Krohn, J. R. Hellmann, D. L. Shelleman, C. G. Pantano, G. E. Sakoske, Effect
[22] Folienwerk Wolfen GmbH, Evguard: special ethylene-vinyl acetate (eva) film for laminated glass and laminated of enameling on the
strength
safety andWorld
glass, dynamicWidefatigue of soda–lime–silica
Web page, float glass, Journal
https://www.evguard.de/en/, 2021. of the American Ceramic Society 85
[23] M.(2002) 2507–2514.
Grujicic, W. Bell, B. Pandurangan, B. Cheeseman, P. Patel, P. Dehmer, Effect of the tin-versus air-side plate-
[27] glass
M. Nase, B. Langer,
orientation on theW. Grellmann,
impact responseFracture mechanics
and penetration on polyethylene/polybutene-1
resistance of a laminated transparent peelarmour
films, structure,
Polymer
Testing 27 (2008) 1017–1025.
Proceedings of the Institution of Mechanical Engineers, Part L: Journal of Materials: Design and Applications
[28] 226
R. S. Rivlin,
(2012) The effective work of adhesion, in: G. I. Barenblatt, D. D. Joseph (Eds.), Collected Papers of R.S.
119–143.
16
16
Journal Pre-proof
[24] X. Nie, W. W. Chen, A. A. Wereszczak, D. W. Templeton, Effect of loading rate and surface conditions on the
flexural strength of borosilicate glass, Journal of the American Ceramic Society 92 (2009) 1287–1295.
[25] M. H. Krohn, J. R. Hellmann, D. L. Shelleman, C. G. Pantano, G. E. Sakoske, Biaxial flexure strength and
dynamic fatigue of soda–lime–silica float glass, Journal of the American Ceramic Society 85 (2002) 1777–1782.
[26] M. H. Krohn, J. R. Hellmann, D. L. Shelleman, C. G. Pantano, G. E. Sakoske, Effect of enameling on the
strength and dynamic fatigue of soda–lime–silica float glass, Journal of the American Ceramic Society 85
of
(2002) 2507–2514.
[27] M. Nase, B. Langer, W. Grellmann, Fracture mechanics on polyethylene/polybutene-1 peel films, Polymer
Testing 27 (2008) 1017–1025.
[28] R. S. Rivlin, The effective work of adhesion, in: G. I. Barenblatt, D. D. Joseph (Eds.), Collected Papers of R.S.
Rivlin, volume 2, 1997, pp. 2611–2614.
pro
[29] K. Kendall, Thin-film peeling – the elastic term, Journal of Physics D: Applied Physics 8 (1975) 1449–1452.
[30] A. J. Kinloch, C. C. Lau, J. G. Williams, The peeling of flexible laminates, International Journal of Fracture 66
(1994) 45–70.
[31] V. A. Eremeyev, K. Naumenko, A relationship between effective work of adhesion and peel force for thin
hyperelastic films undergoing large deformation, Mechanics Research Communications 69 (2015) 24–26.
re-
lP
rna
Jou
17
Journal Pre-proof
CRediT authorship contribution statement
of
Mathias Würkner: Investigation, Review & Editing
pro
re-
lP
rna
Jou