Analyzing Adhesion of Epoxy/Steel Interlayer in Scratch Test

Alp Özdemir et al. Int. Journal of Engineering Research and Applications www.ijera.com
ISSN: 2248-9622, Vol. 5, Issue 11, (Part - 5) November 2015, pp.42-52
www.ijera.com 42|P a g e
Analyzing Adhesion of Epoxy/Steel Interlayer in Scratch Test
Alp Özdemir, İbrahim Kocabaş,Pavel Svanda
Department of Mechanics, Materials and Machine Parts, Jan Perner Transport Faculty, Pardubice University,
53009, Studentska 95, Pardubice, Czech Republic
ABSTRACT
The aim of this paper is to investigate use of an experimental technique to determine which parameters effects
on the interfacial durability performance of adhesive on the metallic adherends as zinc plated mild steel (S235)
by using Taguchi method. The experimental layout has been used four scratch force parameters using the L16
(41
x23
) orthogonal array. The statistical methods of signal to noise ratio (SNR) and the analysis of variance
(ANOVA) were applied to examine effects of surface treatment, adhesive type, blade angle and thickness on
scratch force and scratch energy. Besides, the surface analysis was carried out the morphological modifications
as well as to perform elemental analyses of the pre-treated surfaces. Results of this study indicate that the
thickness and surface treatment are main parameters influencing scratch force (by 52.4% and 19.9%) and
scratch energy (by 44.0 % and 25.6%), respectively.
Keywords: Adhesive, Scratch test, Surface treatment, Taguchi method, Zinc plated mild steel
I. INTRODUCTION
The scrape (scratch) test technique attempted to
grade the strength of adhesion of an adhesive to a
metallic (or relatively smooth non-metallic) adherend
by measuring the force required to remove the
adhesive from an adherend. The scratch test is
usually applied to determine the adhesive strength of
coatings deposited by chemical or physical vapor
deposition techniques[1, 2].This test technique
appears to be very useful for rapidly detecting
changes in interfacial strength of adhesive/adherend
system, and for distinguishing amongst the durability
performance of various surface
pretreatments.Numerous research efforts have been
carried out and similar commercial scratch test
equipment has been employed to evaluate coatings
adhesive strength [3, 4, 5, 6]rather than[2].According
to Knox and Cowling [2]the residual adhesive-
adherend interfacial strength was quantified by
recording the required force to remove a strip of
adhesive from the adherend surface by using a razor.
The proposed benefits of this test method are that the
adhesives are aged in “realistic” environments while
gaining results within a relatively short time
span.Knox and Cowling, [2] initial conclusion was
that the method would be unworkable due to two
main reasons; the formation of an adequately shaped
bead and in some cases the epoxy bond strength
would be too great and only impractically small
beads can be broken free before thewire/fiber-breaks.
Xie andHawthorne[3]performed the effect of
indenter geometry on the failure modes, so that
proper scratch parameters can be chosen to ensure an
adhesive failure is induced in the scratch adhesion
test.This scratch method suggests that it appears to be
very useful for rapidly detecting changes in
interfacial strength of an adhesive-adherend system,
and for distinguishing amongst the durability
performance of various surface treatments [2].
Application of adhesives is usually independent
to metallic substrate material (adherend). In adhesive
bonding, the surface of elements to be joined is
defined as the part of material where interactions
with an adhesive occur. In the many studies, it has
been demonstrated that the strength has beenaffected
by surface treatment, adhesive type, adhesive
thickness, geometry and durability [7, 8]. The surface
pretreatment enables to have a good surface
wettability, precision of properties, improved surface
developments, good activation of surface elements
being bonded and removal of all contaminantsthat
could significantly decrease adhesive joint strength
e.g. lubricants, dusts, loose corrosion layers and
micro-organisms[9, 10].
The Taguchi experimental design method is a
statistical approach that reduces the number of
experiments necessary for investigating the effects of
various parameters on the product quality and/or
quantity. This method also screens the significant
factors affecting the response from those with less
significance, and gives the optimum condition to
attain the most desirable performance[11]. Although,
there are many papers recently published on different
fields by using Taguchi method, but there is no report
available regarding to application of experimental
design analysis considering the effects of surface
treatments, thickness and adhesive type parameters
on the scratch force. The aim of this research was to
reveal use of an experimental approach to
RESEARCH ARTICLE OPEN ACCESS

definewhich parameter affectsinterfacial durability
performance of metallic adherend (S235 zinc plated
mild steel)via Taguchi method. The Taguchi L16
orthogonal array was employed to analyze
experimental scratch test results obtained from eight
experiments withtworepetitionsand four process
parameters e.g. surface treatment, adhesive type,
blade angle, and thickness. The obtained results were
analyzed by using a variance analysis
(ANOVA).Besides, the surface morphology of each
adherend after treatments was observedvia scanning
electron microscopy (SEM) with energy dispersive
X-ray spectroscopy (EDX).
II. MATERIAL METHOD
2.1. Materials
Two type of adhesives were selected; Veporal
Super (HE 20-06), a unique hybrid two-component
epoxy structural adhesive with high elongation up to
55% having excellent peel and shear strength. It is
used for structural bonding for a wide range of
substrates in the scratch tests. Shear strength, tensile
strength and strain at fracture are 13 MPa, 16 MPa
and 40%, respectively. The second type of adhesive
was a brittle type of adhesive Carbo Resin. Carbo
Resin is two component epoxy base glue with
inorganic fillers. The have good adhesion to many
materials. Curing is at normal room temperature.
Minimum shear strength after 14 days is 13 MPa.
Total curing 7 days at 20°C.
The adherend materialis a low strength mild steel
(S255) whose chemical composition is given in Table
1 based on Ref.[9]. Two values of adhesive thickness
were used 0.6 and 0.3 mm in the experiments. The
adherend thickness was constant as 1 mm for each
specimen.
Table 1.Chemical composition of mild steel (S255)
Composition [wt.%]
C Mn Si P S Cr Ni N Cu
Max. Max. 0.15-0.3 Max. Max. Max. Max. Max. Max.
0.22 0.65 0.04 0.05 0.3 0.3 0.012 0.3
2.2. Test method
To investigate the effect of surface treatment on
the adhesion strengths of Veporal Super (HE 20-06)
and Carbo Resin, a jig at the surface was created
based on Knox and Cowling‟s paper[2],(see in the
Figure1) to strip a thick film of adhesive from an
adherend see in Figure 2.
For specimens used in the scratch tests the
procedure is as followings[2]
 The required area on the adherend is prepared
for adhesive. This may include chemical
etching, sanding (shot blasting), anodic
oxidation of the surface, and followed by
treatment with a primer if required.
 The adhesive was applied to adherend surface.
 The thickness of specimen and bond-line
thickness were controlled by using wires above
the adhesive.
 The adhesive was cured according to
manufacturer‟s instructions involving a cure at
room temperature for 24 hour. The specimens
were then allowed to wait at ambient in the
laboratory environment.
 The specimen thickness was verified after
curing process.
 The tests were performed by using a scratch
tool(Fig. 1)in a tensile testing machine (ZD
10/90) at a constant crosshead speed of
25mm/min) at ambient conditions.
Figure 1.Design of scratch jig
Figure 2.Scratch test specimen (not to scale)

2.3. Surface Treatment
In accordance with, the adherends were treated
by four different surface treatment method i.e.
sandblasting, chemical etching, and anodic treatment
followed by mixture combination in this study.
2.3.1. Sandblasting- Sanding (S1)
The sandblasting process was performed using
dry sanding box with ceramic abrasive for industrial
sanding application. This procedure was carried out
for specimens, held at a distance of 2-3 cm
approximately from the nozzle as accurately as
possible and sand-blasted at a pressure of 600 kPa.
2.3.2. Chemical etching (S2)
The chemical etching process was carried out all
the zinc plated mild steel specimens immersed into
acid solution. The acid solution was prepared by
using 20 ml hydrochloric acid and 40 ml distilled
water. During the etching process,zinc layer was
completely removed from the steel substrate.
2.3.3. Anodic oxidation treatment (S3)
In anodizing treatment, the adherend was clamped
to the anode and cathode holders. The composition
of solution was arranged with same percentage as
phosphoric acid (10 ml); distilled water (30ml). The
anodizing voltage was raised to 30 V and held for
20-30 seconds. At the end of this time the adherend
was cleaned by using cold water at ambient
temperature. The anodized adherends can then be
air-dried, preferably blow-dry. Anodic oxidation
treatment produces a very thin layer on the adherend
surface. Before anodized treatment was applied, all
specimens were undergone chemical etching
process.
2.3.4. Mixture combination
Mixture surface treatment processes were consisted
of sanding, anodizing treatment after chemical
etching processes. In the first step of this surface
treatment, chemical etching was applied on all zinc
plated mild steel specimens. Secondly, all specimens
which have been undergone treatment were sanded.
Finally, anodic oxidation was applied by using
electrochemical treatment.
2.4. Taguchi matrix
The Taguchi method was used to design the
experiments. The Taguchi array contains four
factors, or variables, corresponding to the surface
treatment (A), adhesive type (B), blade angle (C)
and thickness (D). If all the possible test
combinations were to be tested, the number of tests
would be 64 (one test, no repeating) which are
impractical in terms of time andcost. The use of pre-
defined orthogonal arrays on which the Taguchi
method is based reduces the number of tests and
permits to quantify the interactions between the
variables considered. The experimental layout for
the four scratch force parameters using the L16
(41
x23
) orthogonal array is shown in Table 2.
Accordingly, eight experiments were carried out to
study effect of scratch force input parameters. Each
experiment was repeated two times in order to
reduce experimental errors. It contains 8 rows
corresponding to the number of tests with two
replicates, one column with four levels) and 3
columns with 2 levels. The first column was
assigned the surface treatment, the second to the
glue type, the third to the blade angle, and the fourth
to the adhesive thickness (see Table 3). The response
studied was scratch force (F), scratch energy (SE)
and it involves signal to noise (S/N) ratio factors.
The influence of each variable was assessed by the
average response and the analysis of variance
(ANOVA). The statistical software MINITAB
17program [12] was used.
Table 2.Experimental layout using L16 orthogonal
array
No Sample
No
Surface
Treat.
(A)
Type
(B)
Angle
(C)
Thick-
ness
(D)
1 1 1 1 1 1
2 1 1 1 1
3 2 1 2 2 2
4 1 2 2 2
5 3 2 1 1 2
6 2 1 1 2
7 4 2 2 2 1
8 2 2 2 1
9 5 3 1 2 1
10 3 1 2 1
11 6 3 2 1 2
12 3 2 1 2
13 7 4 1 2 2
14 4 1 2 2
15 8 4 2 1 1
16 4 2 1 1
Table 3.Scratch force parameters
and their levels
Parameters Level 1 Level 2 Level 3 Level 4
Surface
treat. (A)
S1 S2+ S1 S2+S3 S2+S1+ S3
Adhesive
type (B)
Soft Rigid - -
Blade
angle (C)
0° 15° - -
Thickness
(D)
0.3mm 0.6mm - -

2.5. Surface analysis
The treated surfaces were characterized for
microstructural evaluations by using analytical
scanning electron microscopy (SEM) with energy
dispersive X-ray spectroscopy (EDX) analysis using
Tescan Vega III SB electron microscope. This
surface analytical technique was used to study the
morphological modifications as well as to perform
elemental analyses of the treated surfaces.
III. RESULTS AND DISCUSSION
3.1 Microstructureevaluation
In the EDX analysis of basic material (zinc
plated mild steel), it was observed some starting of
corrosion after one day even if it is in the good
condition. In accordance with, we may apply surface
treatment on basic material preventing the weak
boundary layer occurred against to corrosion and
corrosion products (see in Figure 3(a-d)). This
problem is resulted from weak boundary layer
theory. According to [13] this theory states that bond
failure at the interface is caused by either a cohesive
break or a weak boundary layers. Weak boundary
layers can originate from the adhesive, the adherend,
the environment, or a combination of any of the
three. Weak boundary layers can occur in the
adhesive or adherend if an impurity concentrates
near the bonding surface and forms a weak
attachment to the substrate. When failure takes
place, it is the weak boundary layer that fails,
although failure appears to take place at the
adhesive-adherend interface. Weak boundary layers,
such as those found in polyethylene and metal
oxides, can be removed or strengthened by various
surface treatments. Weak boundary layers formed
from bonding environment are very common.
3.2 Scratch force analysis and failure mechanisms
Scratch forces for each configuration of
adhesive samples including different surface
treatments, adhesive type, thickness and angle of
scratching were performed experimentally. The
trends of scratch forces with respect to position of
cutting tool are demonstrated in Figure 4 and 5. Two
different behavior of fracture mechanism were
achieved as ductile and brittle response. The samples
having soft adhesive, are mainly characterized by
relatively smooth and lower force amplitudes with
low amount of oscillations as it is exhibited e.g. in
sample 1, 3 and 7 in Fig. 4(a) and 4(c), and Fig 5(c).
Hence, the failure mechanism for these samples is
mainly dominated by interfacial fracture stimulating
exponential traction and separation cohesive zone
delamination as stated in literature. For almost all
samples, the force increases up to traction limit
corresponding to peak values on the graphs then
softening mechanism takes place until the critical
distance is achieved. This phoneme was also
experienced for rigid adhesives, excepting large
amplitudes of force oscillations caused by
considerably high amount of vibrations due to brittle
cracking fracture response. Contrary, the scratch
force variation for the sample 5 has a brittle fracture
response. This adverse effect may be evaluated as
the tendency of interface adhesion to a brittle
behavior due to surface treatment factor (etching
plus anodic oxidation), yielding an adhesive failure
at slightly lower thickness of 0.3mm in Fig. 5(a).
Therefore, this mentioned brittle interface zone was
considered to generate high frequency of vibrations
accompanied with high scratch force amplitudes.
The surface treatment option, especially anodic
oxidation process had a quite negative impact on the
bonding characteristic of adhesive and adherend.
This situation was observed in sample 5 and 8 in
Fig. 5.a and 5.d. The anodic oxidation processes led
to weakening bonding strength at relatively low
adhesive thickness. The mean values of scratch
forces in the steady-state (separation) region were
illustrated inTable 4.The samples corresponding to
thickness (0.6 mm) have relatively high scratch
forces at an interval of 918 N and 1020 N. The lower
scratch force was obtained at thinner (0.3 mm)
adhesive sections which are a sign of significant
effect of thickness.

Figure 3.SEM- EDX analysis of treated surface,a-Sanding b- Chemical etching + sanding, c- Chemical etching,
d- Chemical etching + sanding + anodic oxidation
Figure 4.The scratch force-displacement graphs of samples 1-4
ba
dc

Figure 5.The scratch force-displacement graphs of samples 5-8
3.3 Scratch energy analysis
The amount of energy to drive the adhesive zone
into fracture state is a better indication of surface
adhesion properties. For this reason, the work done
during the scratch process was evaluated in terms of
area under the force-displacement curves based on
trapezoidal integration rule per unit area of adhesive
surface. This parameter is called specific scratch
energy in kJ/m2
to reach failure state. The specific
scratch energy values of each sample were given in
Table 4 and Figure 6 each represents different surface
treatments, geometry parameters and etc. The highest
specific energy values were obtained in the samples
of 2,3 and 7 corresponding to thickness 0.6mm,
expressing the substantial effect of thickness by
roughly 44%. This is a reason of stacking ability and
larger contact interaction of cutting blade and
adhesive cross-section at increasing thicknesses
against longitudinal motion. Furthermore, sharp
decreases were concluded at lower thicknesses by a
certain amount regardless of other parameters.
However, there has been a remarkable impact of
surface treatment especially for the samples of 5, 6
and 8 undergoing anodic oxidation treatment,
attaining the worst surface effect on scratch
resistance of adhesive. On the other hand,
implementation of sanding process has produced
quite better bonding characteristics comparing to the
others for all samples as it was stated in previous
works in literature [14]. Based on the variations of
specific energy values in Table 4, no significant
contribution of adhesive type was appeared although
it plays an important role in fracture mode which is
either adhesive or cohesive failure.In terms of
fracture energy approach, samples 2 and 7 having
0.6mm thickness and subjected to sanding process in
common have the optimum configurations with
similar scratch energies of 36kJ/m2
approximately.
The specific fracture energy for both brittle and
ductile fracture behavior were observed to be not
influenced due to the fact that brittle material
undergoes low displacement at high forces, whereas
ductile materials exhibits opposite response.
3.4 Statistical analysis
3.4.1. Analysis of signal-to-noise (S/N) ratio
Taguchi uses the S/N ratio as the quality
characteristic of choice. S/N ratio is considered as a
measurable value instead of standard deviation
because as the mean decreases, the standard deviation
also decreases and vice versa. In less technical terms,
signal-to-noise ratio compares the level of a desired
signal (such as music) to the level of background
noise. The higher the ratio, the less obtrusive the
background noise is. „„Signal-to-noise ratio‟‟ is
sometimes used informally to refer to the ratio of
useful information tofalse or irrelevant data in a
conversation or exchange. In other words, the

standard deviation cannot be minimized first and the
mean brought to the target [15]. Taguchi has
empirically found that the two stage optimization
procedure involving S/N ratios indeed gives the
parameter level combination, where the standard
deviation is minimum while keeping the mean on
target. The target mean value may change during the
process development. Two of the applications in
which the concepts of S/N ratio are useful are the
improvement of quality through variability reduction
and the improvement of measurement. The S/N ratio
characteristics can be divided into three categories
given by Equations (1) – (3), when the characteristic
is continuous [16]:
 Nominal is the best characteristic
𝑆
𝑁
= 10 log
𝑦
𝑆 𝑦
2(1)
 Smaller is the better characteristic
𝑆
𝑁
= −10 log
1
𝑛
𝑦2
(2)
 and larger the better characteristic
𝑆
𝑁
= − log
1
𝑛
𝑦2
(3)
Where 𝑦the average is observed data, 𝑆𝑦
2
the
variation of y, nthe number of observations, and y the
observed data or each type of the characteristic, with
S/N ratio, the better results when we consider surface
treatment, adhesive type, blade angle and thickness.
Factor levels that maximize the appropriate S/N ratio
are optimal. The goal of this research was to produce
maximum scratch force (F) and energy. Larger F and
energyvalues represent better adhesive resistance to
scratch. Therefore, a larger-the-better quality
characteristic was implemented and introduced in this
study. As mentioned earlier, there are three categories
of performance characteristics, i.e., the lower-the-
better, the higher-the-better, and the nominal-the-
better.
Figure 6.Specific scratch energy and forces
The Taguchi L16 orthogonal array was
employed to analyze experimental results of scratch
force, scratch energy, S/N ratio and failure modes
obtained from 8 experiments which are given in
Table 4.The level values obtained from MINITAB
17 software program[12]according to the Taguchi
design are given in Table 4 and Table 5.Table 5
shows the experimental results for scratch force,
scratch energy and the corresponding S/N ratio using
Equation(3).
Table4.Experimental results
Sample A B C D
Scratch
Force
(N)
Set 1
Scratch
Force
(N)
Set 2
S/N
ratio
ScratchEnergy
(KJ/m2
)
Set 1
Scratch
Energy
(KJ/m2
)
Set 2
S/N
ratio
Failure
mode
1
S1 Soft 0° 0.3
mm
388.7 384.6 51.8 8.6 9.7 19.1 Adhesive
2
S1 Rigid 15° 0.6
mm
1020.0 976.0 59.9 28.3 35.7 29.9 Cohesive
3
S2+S1 Soft 0° 0.6
mm
993.4 1000.7 59.9 27.6 33.2 29.6 Adhesive
4
S2+S1 Rigid 15° 0.3
mm
495.3 500.4 53.9 20.4 17.3 25.4 Cohesive
5 S2+S3 Soft 15° 0.3 492.0 435.5 53.2 14.0 9.6 21.0 Adhesive
9.7
35.7
27.6
20.4
9.6
12.8
35.6
11.8385
976 993
500
436
359
918
293
0
200
400
600
800
1000
1200
0
5
10
15
20
25
30
35
40
1 2 3 4 5 6 7 8
ScratchForce[N]
ScracthEnergy[kJ/m2]
Sample No Scratch energy Scratch Force

mm
6
S2+ S3 Rigid 0° 0.6
mm
372.1 359.3 51.25 7.1 12.8 18.85 Adhesive
7
S2+S1+S3 Soft 15° 0.6
mm
918.0 910.2 59.22 35.6 27.4 29.74 *
Hybrid
8
S2+S1+S3 Rigid 0° 0.3
mm
292.5 276.9 49.08 11.8 9.4 20.36 Cohesive
*Hybrid: Adhesive plus cohesive failure mode
Table 5.Response table mean signal to noise ratios for scratch force
Scratch Force Scratch Energy
Level A B C D Level A B C D
1 55.86 56.05 53.01 52.01 1 24.54 24.87 21.98 21.49
2 56.96 53.56 56.60 57.61 2 27.48 23.64 26.52 27.02
3 52.26 - - - 3 19.93 - - -
4 54.15 - - - 4 25.05 - - -
Delta 4.69 2.49 3.59 5.60 Delta 7.55 1.23 4.54 5.53
Rank 2 4 3 1 Rank 1 4 3 2
Total mean S/N ratio= 54.81 Total mean S/N ratio= 24.25
Accordingly, Figure 7shows that the second level of A factor (surface treatment), the first level of B factor
(adhesive type) and the second level of C factor (blade angle) and the second level of D factor (thickness) are
higher as both left and right side.
Figure 7.Mean of S/N ratios versus factor levels for scratch force (at left side) and scratch energy (at right side)
3.4.2. Analysis of variance for scratch force and
scratch energy analysis
The analysis of variance (ANOVA) is a
powerful technique in Taguchi method that explores
the percent contribution of factors affecting the
response. The strategy of ANOVA is to extract the
variations that each factor cause relative to the total
variation observed in the results. The results of the
ANOVA for scratch force and scratch energy with
surface treatment (A) adhesive type (B), blade angle
(C), thickness (D) and interaction (E) parameters are
shown in Tables 6and 7. This analysis was carried
out for a significance level of α = 0.05, i.e. for a
confidence level of 95%. Tables 6 and 7 show the P-
values, that is, the realized significance levels,
associated with the F-tests for each source of
variation. The sources with a P-value less than 0.05
are considered to have a statistically significant
contribution to the performance measures. The
other/error term, in the last row of ANOVA table,
contains thus the information about three sources of
variability of the results including uncontrollable
factors, factors that are not considered in the
experiments, and the experimental error [11]. It
should be emphasized that the interpretation of
ANOVA table is valid just in the range of
considered levels for the factors. That‟s why the

determination of levels is of great importance in any
experimental design approach.
The ANOVA table (see Table 6 and 7) of the
experimental results gives the relative importance of
all the variables. The main factors influencing the
scratch force are thickness 52.4%. The second factor
is surface treatment also significant contributionat
about 19.9%. The other factors; blade angle and
adhesive type have contribution, 13.7% and 7.3%
respectively. On the other hand, the interaction
between adhesive type-angle-thickness parameters
was determined as 6.5%.In case of analysis of
variance analysis for scratch energy, the trend of
contribution is similar to scratch force. Thickness
43.9% is main factor effect on scratch energy.
The other parameters; surface treatment 25.6%,
blade angle 17.8% and adhesive type 2.0% whereas
the interaction between adhesive type-angle-
thickness parameters were shown 3.7% given in
Table 7.The F-ratio in ANOVA table is a reliable
criterion for ranking the factors with respect to their
influence. A higher value of the calculated F-ratio
for a factor means a greater influence of that factor
on the experiment outcome. Moreover, if the percent
contribution of a factor would be equal to or less
than 10% of that of the most affecting factor, this
factor can be pooled with error terms [11].
According to Table 6 and Table 7, P value is scratch
force and scratch energy at the reliability level of
95%,because the results are lower than 0.05.
Table 6.Results of the analysis of variance analysis for scratch force
Source DF Seq SS Contribution Adj SS Adj MS F-Value P-Value
Surface Treatment3 255750 19.87% 339493 113164 314.31 0.000
Adhesive type1 94660 7.35% 178411 178411 495.530.000
BladeAngle 1176114 13.68% 8448 8448 23.46 0.001
Thickness 1 674051 52.36% 141155 141155 392.05 0.000
Adhesive type*Angle*Thickness 1 83913 6.52% 8391383913 233.07 0.000
Error 8 2880 0.22% 2880 360
Total 15 1287369 100.00%
The optimum conditions to attain scratch
force/displacement can be determined from
maximum points in main effect. Applying the
optimum condition, the contribution of each factor on
improvement of response can be found using Taguchi
approach [11]. A prediction for scratch force with
regarding factors and their levels was performed in
the MINITAB 17 Software program [12]. This
prediction based on S/N ratio‟s highest values is in
the parameter level chosen as (A2, B1, C2 and D2).
As a result of this prediction, the scratch force and
scratch energy are calculated 1062.04 N and 37.00
KJ/m2
, respectively.
Table 7.Results of the analysis of variance analysis for scratch energy
Source DF Seq SS Contribution Adj SS Adj MS F-Value P-Value
Surface Treatment 3 416.25 25.63% 474.83 158.28 11.31 0.003
Adhesive type 1 32.93 2.03% 91.19 91.19 6.52 0.034
BladeAngle 1 289.29 17.81% 42.71 42.71 3.05 0.019
Thickness 1 713.32 43.92% 179.39 179.39 12.82 0.007
Adhesive type*Angle*Thickness 1 60.32 3.71% 60.32 60.32 4.31 0.072
Error 8 111.92 6.89% 111.92 13.99
Total 15 1624.03 100.00%
IV. CONCLUSION
In this paper, an optimization process was
implemented in order to analyze influence of
different surface treatment processes, geometrical
parameters and material types on the scratch
resistance of two different adhesives onto one
substrate of S235 zinc plated mild steel. For
experimental procedure, eight different
configurations were prepared and subjected to scratch
tests including some surface examinations via SEM
and EDX analysis. The failure modes,mean of scratch
forces at stable region, specific scratch energy and
statistical calculations based on application of the
parameter design of the Taguchi method were carried
out. Consequently, a variance of analysis (ANOVA)
was introduced to estimate contribution of each
design parameter on the scratch resistance of

adhesive bondline in terms of force and energy. The
following conclusions can be drawn based on the
experimental results of this study are;
 Taguchi method of experimental design has been
carried out for optimizing scratch force response
parameters, evaluated with L16 orthogonal array.
This design method is suitable to predict the
scratch force as described in this paper.
 It is found that the parameter design of the
Taguchi method provides a simple, systematic,
and efficient methodology for the optimization
of the scratch force parameters.
 The experimental results demonstrate that
thickness of adhesive layer and surface treatment
are main parameters influencing scratch force
(52.4% and 19.9%) and scratch energy (44.0 %
and 25.6%), respectively.
 Although the adhesive type has an ignorable
effect on scratch energy, it was observed to be
very sensitive to changing of peak scratch forces.
 Sanding process was observed to give better
bonding ability as a result of mechanical
interlocking for almost all samples. However,
anodic oxidation process has a degradation
behavior on bonding ability and is not suggested
as an effective surface treatment.
 Blade angle was also concluded as a notable
parameter in the evaluation of scratch force and
energy by an amount of 13.8% and 17.8%,
respectively.
 SEM-EDX evaluations show that it is necessary
to apply a surface treatment on adherend to
prevent weak layer at interface adhesion due to
adherend surface corrosion.
Further study could consider more factors (e.g.
curing conditions, wetting angle of adherend, primer
application etc.) in the research to see how the factors
would affect scratch force and scratch energy. Also,
further study could consider the outcomes of Taguchi
parameter design when it is implemented as a part of
management decision-making processes.In
experiments the fracture mode is either adhesive or
cohesive. Further investigations are necessary to
determine the dependence of the traction–
deformation relation on the thickness of the adhesive
layer, shear deformation rate, type of adhesive etc.
REFERENCES
[1.] Cailler, M., Lee, G. H. Scratch Adhesion
Test of Magnetron-Sputtered Copper
Coatings on Aluminium Substrates: Effects
of the SurffacePreparition. Thin Solid Films,
168(2), 1989,193-205.
[2.] Knox, E. M., Cowling, M. J. A rapid
durability test method for adhesives. Int J
Adhesion and Adhesives, 20(3),2000, 201-
208.
[3.] Xie, Y., Hawtorne, H. M. Effect of contact
geometry on the failure modes of thin
coatings in the scratch adhesion test. Surface
and Coating Technology, 155 (2-3), 2002,
121-129.
[4.] Wirasate, S., Boerio, F. J. Effect of
Adhesion, Film Thickness, and Substrate
Hardness on the Scratch Behavior of
Poly(carbonate) Films. The Journal of
Adhesion, 81(5), 2005, 509-528.
[5.] Jiang, H., Browning, R., Liu, P., Chang, T.
A., Sue, H.-J. Determination of epoxy
coating wet-adhesive strength using a
standardized ASTM/ISO scratch test.
Journal of Coatings Technology and
Research, 8(2),2010, 255-263.
[6.] Moghbelli, E., Banyay, R., Sue, H.-J. Effect
of moisture exposure on scratch resistance
of PMMA. Tribology International,69,2014,
46-51.
[7.] da Silva, L. F. M., Carbas, R. J. C.,
Critchlow, G. W., Figueiredo, M. A. V.,
Brown, K. Effect of material, geometry,
surface treatment and environment on the
shear strength of single lap joints. Int J
AdhesAdhes., 29(6):2009, 621-632.
[8.] da Silva, L. F. M., Ferreira, N. M. A. J.,
Richter-Trummer, V., Marques, E. A. S.
Effect of grooves on the strength of
adhesively bonded joints. Int J
AdhesAdhes.30(8):2010, 735-743.
[9.] Özdemir, A., Svanda, P. The
Characterization of Zinc Plated Surfaces
After Aged Adhesives. University of
Pardubice Jan Perner Transport Faculty-
Serie B, 19, 2015, 137-144.
[10.] Rudawska, A. Selected aspects of the effect
of mechanical treatment on surface
roughness and adhesive joint strength of
steel sheets. Int J AdhesAdhes. 50, 2014,
235-243.
[11.] Majdzadeh-Ardakani, K., Navarchian, A.
H., Sadeghi, F. Optimization of mechanical
properties of thermoplastic starch/clay
nanocomposites. Carbohydrate
Polymers.79(3), 2010, 547-554.
[12.] Author. Minitab 17,
http://www.minitab.com/en-us/. 2014.
[13.] Ebnesajjad, S. Adhesives Technology
Handbook.(William Andrew Inc.,
Norwich,NY, USA, 2008).
[14.] Sancaktar, E., Gomatam, R. A study on the
effects of surface roughness on the strength
of single lap joints. Journal of Adhesion
Science and Technology,15(1), 2001, 97-
117.

[15.] Asiltürk, İ., Akkuş, H. Determining the
effect of cutting parameters on surface
roughness in hard turning using the Taguchi
method. Measurement, 44(9), 2011, 1697–
1704.
[16.] Shetty, R., Pai, B. R., Rao, S. S., Nayak, R.
Taguchi‟s Technique in Machining of Metal
Matrix Composites. J of the BrazSoc of
MechSci&Eng.31(1),2009, 12-20.

Analyzing Adhesion of Epoxy/Steel Interlayer in Scratch Test

More Related Content

What's hot (20)

Viewers also liked (20)

Similar to Analyzing Adhesion of Epoxy/Steel Interlayer in Scratch Test (20)

Recently uploaded (20)

Analyzing Adhesion of Epoxy/Steel Interlayer in Scratch Test