You are on page 1of 20

This article was downloaded by: [Northeastern University]

On: 21 November 2014, At: 18:02


Publisher: Taylor & Francis
Informa Ltd Registered in England and Wales Registered Number: 1072954 Registered
office: Mortimer House, 37-41 Mortimer Street, London W1T 3JH, UK

Journal of Adhesion Science and


Technology
Publication details, including instructions for authors and
subscription information:
http://www.tandfonline.com/loi/tast20

A study of the effect of surface pre-


treatment on the adhesion of coatings
a a a
Anna Guzanov , Janette Brezinov , Dagmar Draganovsk &
b
Frantiek Ja
a
Faculty of Mechanical Engineering, Department of Technology
and Materials, Technical University of Koice, Msiarska 74, 040 01
Koice, Slovak Republic
b
Renojava, s.r.o., Bulharsk 26, 080 01 Preov, Slovak Republic
Published online: 29 May 2014.

To cite this article: Anna Guzanov, Janette Brezinov, Dagmar Draganovsk & Frantiek Ja (2014)
A study of the effect of surface pre-treatment on the adhesion of coatings, Journal of Adhesion
Science and Technology, 28:17, 1754-1771, DOI: 10.1080/01694243.2014.920762

To link to this article: http://dx.doi.org/10.1080/01694243.2014.920762

PLEASE SCROLL DOWN FOR ARTICLE

Taylor & Francis makes every effort to ensure the accuracy of all the information (the
Content) contained in the publications on our platform. However, Taylor & Francis,
our agents, and our licensors make no representations or warranties whatsoever as to
the accuracy, completeness, or suitability for any purpose of the Content. Any opinions
and views expressed in this publication are the opinions and views of the authors,
and are not the views of or endorsed by Taylor & Francis. The accuracy of the Content
should not be relied upon and should be independently verified with primary sources
of information. Taylor and Francis shall not be liable for any losses, actions, claims,
proceedings, demands, costs, expenses, damages, and other liabilities whatsoever or
howsoever caused arising directly or indirectly in connection with, in relation to or arising
out of the use of the Content.

This article may be used for research, teaching, and private study purposes. Any
substantial or systematic reproduction, redistribution, reselling, loan, sub-licensing,
systematic supply, or distribution in any form to anyone is expressly forbidden. Terms &
Conditions of access and use can be found at http://www.tandfonline.com/page/terms-
and-conditions
Downloaded by [Northeastern University] at 18:02 21 November 2014
Journal of Adhesion Science and Technology, 2014
Vol. 28, No. 17, 17541771, http://dx.doi.org/10.1080/01694243.2014.920762

A study of the effect of surface pre-treatment on the


adhesion of coatings
Anna Guzanova*, Janette Brezinova, Dagmar Draganovska and Frantiek Jab
a
Faculty of Mechanical Engineering, Department of Technology and Materials,
Technical University of Koice, Msiarska 74, 040 01 Koice, Slovak Republic; bRenojava, s.r.o.,
Bulharsk 26, 080 01 Preov, Slovak Republic
Downloaded by [Northeastern University] at 18:02 21 November 2014

(Received 26 June 2013; nal version received 22 April 2014; accepted 25 April 2014)

This paper deals with the study of effects of mechanical surface preparation on the
adhesion of coating with high content of zinc dust. Five kinds of mechanically
blasted surfaces were studied. The following were used as abrasives: steel shot, steel
grit, brown corundum oxide and zirblast. The last surface type was modied by
MBX Blaster technology (mechanical bristle blasting). The surfaces topography was
quantied by a roughness prolometer. The shape and size of the incurred inequali-
ties on the modied surfaces were studied using 3D microscope images of the sur-
face. The purity of the surfaces after pre-treatment was evaluated by impurity glued
on the tape and measuring the reection of light from the surface. Fractal analysis
was used to evaluate the diversity of inequalities on the prepared surfaces. Cross-
sections were also taken of the prepared surfaces. The prepared surfaces were coated
with zinc-lled coating. The adhesion of the coating to the substrate was evaluated
by a pull-off test after curing the coating (as sprayed), as well as after exposure to
severe corrosive environments. The best adhesion of the coating was found for the
coating applied to the substrate which had been pre-treated with brown corundum
and steel grit.
Keywords: zinc-lled coating; surface preparation; adhesion; surface purity; fractal
analysis

Introduction
The issue of clean surfaces and their impact on the quality of coating systems, or on
the strength of bonded joints is well known and frequently discussed in the professional
community. The presence of a few monolayers of contaminants on the surface leads to
a reduction in the joint strength when under load and breach in adhesion is caused, as
showed by Watts in [1].
Anderson et al. [2] and other researchers [39] examined the sensitivity of various
bonded-joints and coatings to contamination of the adherents surface, by simulating
surface contamination using a specied amount of silicone. The measurement of surface
cleanliness is a signicant problem in many industrial and technological processes.
Typical surface analysis techniques often used in research of adhesion and interfaces
were X-ray photoelectron spectroscopy, time-of-ight secondary ion mass spectrometry,
Auger electron spectroscopy, together with its imaging counterpart scanning Auger
electron microscopy, etc. All these techniques were working under ultra-high vacuum

*Corresponding author. Email: anna.guzanova@tuke.sk

2014 Taylor & Francis


Journal of Adhesion Science and Technology 1755

conditions. The existing methods are based on laboratory procedures that are not per-
formed in real time, cannot be automated and usually are restricted to a small portion
of the sample. Bilmes et al. [10] designed a new method for real time measurement of
the amount of surface dirt or residual contamination on a surface based on the ablation
of the surface dirt lm by means of a short laser pulse, and the subsequent measure-
ment of the emitted sound. The intensity of the sound is proportional to the amount of
surface dirt and provides a direct measurement of the cleanliness of the surface.
A specic problem is the contamination of surfaces prepared for bonding. Different
types of impurities occur on metal surfaces, which are intended for mechanical cleaning
(Figure 1).
On the surface area often present products of chemical reactions between material
and environment (corrosion products, scale), foreign impurities, coming from the pro-
duction history of the substrate material (metal forming lubricants) or water, dust, ash,
Downloaded by [Northeastern University] at 18:02 21 November 2014

gas that are bonded to the surface by adsorption, when atoms, ions or molecules from
a gas, liquid or dissolved solid are adhered to a surface by adhesion forces (preserving
agents wax, oil, silicone and paints). Pocius states layers of adsorbed gases, polar
organics, non-polar organics, adsorbed water and metal oxides on unprepared metal
surface.[11]
Impurities coming from the surface layer (oxide substrate material particles, dust
and old paint), as well as from the cleaning process itself (broken abrasive particles)
are present on mechanically prepared surfaces.
The surface of the adherend has contaminants and the adhesive has a nite
viscosity. Therefore, pore penetration is not complete, leaving voids at the interface.
Pocius stated, the presence of voids as well as cohesively weak contaminants decreas-
ing the strength of the adhesive bond.[11] Since adhesives and coatings are based on a
similar binder matter, it can be generalised these statements also for paint adhesion.
The adhesion of adhesives/coatings is inuenced by the following key factors:

 presence of impurities;
 presence of voids imperfect contact between the adhesive and the substrate/
coating;
 size of the surface roughening. The size of the contact area and adhesion has a
direct relationship;

Figure 1. Types of impurities on metal surfaces.


1756 A. Guzanov et al.

 shape of the surface roughness especially important are inequalities which are
suitable for the mechanical anchoring of coating.

Zinga paint was chosen to verify the impact of purity and surface morphology on
the adhesion of the coating. Zinga contains electrolytic zinc with a purity of 99.995%,
as well as being free of lead and cadmium. The zinc content in the resin binder base is
so high, that when cured, the coating consists of 96% pure zinc. In the corrosive envi-
ronment are formed zinc salts on the surface of the paint layer, which produce a barrier
to further corrosion, together with specic resin binder. Zinga can be applied to the res-
toration of damaged hot-dip galvanised surfaces. Surfaces treated with Zinga paint can
be repeatedly renovated by Zinga without the risk of separation of the individual layers,
because in a very short time they are able to integrate perfectly with each other. Zinga
can be used as a primer under other coats with different binder bases (duplex, triplex
Downloaded by [Northeastern University] at 18:02 21 November 2014

coating systems), thereby increasing the life of the coating system by 2.5 times. Zinga
has excellent adhesion to steel, it is very exible, so it resists cracking as a result of
the thermal dilation of the substrate, mechanical damage, abrasion or impact. The
cathodic protection provided by Zinga coating acts also when any local mechanical
damage of the coating occurs. The properties of the Zinga coating allow to design this
exceptional paint to the most aggressive corrosive atmospheres (C4, C5-I, C5-M).
This paper deals with the assessment of the impact of surface cleanliness, and the
obtained size of the cleaned surface area after blasting on coating adhesion to the
surface according to the type of abrasive treatment. Adhesion of the zinc-lled coating
Zinga is studied immediately after its application and after exposure to various
corrosive environments.

Materials and methods


Material of the substrate
The substrate for mechanical pre-treatment and the coating application was made of
steel S235JRG2 (EN 10025A1). The tensile strength is 363441 MPa, yield strength
min 235 MPa. The chemical composition of the steel substrate, as indicated by manu-
facturer,[12] is shown in Table 1.

Surface preparation
The substrate was pre-treated by:

(1) Air blast cleaning using the following four types of blasting abrasives:
(a) Steel shot (SS)
(b) Steel grit (SG)
(c) Brown corundum (ZB)
(d) Zirblast (ZB)

Table 1. Chemical composition of the steel substrate (in wt. %) [12].

Cmax Mnmax Almin Smax Pmax Nmin Fe


0.17 1.400 0.020 0.045 0.045 0.009 balance
Journal of Adhesion Science and Technology 1757

The chemical composition of the blasting abrasives used for air blast cleaning is
listed in Table 2.
The parameters of the air blasting are summarised in Table 3.

(1) MBX Blaster technology (MBX). It involves the use of specially shaped rotat-
ing wire bristles that are anchored into the rotating hub of an electrically or
pneumatically driven tool, see Figure 2.

The bristle blasting process (MBX) removes corrosion and generates an anchor pro-
le by using a specially designed rotary bristle tool. This tool consists of ground wire
bristle tips that are bent forward and dynamically tuned to a hand-held power tool
which operates at approximately 2500 rpm. The mechanical principle of the bristle
blasting tool is summarised as follows: Bristle tips are designed to strike the corroded
Downloaded by [Northeastern University] at 18:02 21 November 2014

surface with kinetic energy that is equivalent to standard processes that use grit blast
media. Immediately after the bristle tips strike the corroded steel surface, they retract
(i.e. rebound) from the surface, which results in both corrosion removal and a micro-
indentation that exposes a fresh surface. Consequently, surfaces that have been treated
by the bristle blasting tool have a texture and visual cleanliness that mimics those
obtained by traditional grit blasting processes.
The parameters of the bristle blasting technology can be dened only by the bristle
tool speed, mentioned above. In general, the face of the tool hub is oriented perpendic-
ular to the treated surface during use. During corrosion removal, the bristle tips are
brought into direct contact with the corroded surface using minimal applied force, and
the rotating tool is gradually moved along the feed direction, that is, either to the left
or right of the user. Thus, the appropriate pressure and feed rate of the tool is obtained
by direct experimentation and by visually inspecting the trial-tested region to ensure
that the desired cleaning standard/requirement has been reached.[13]

Cleanliness of the surface


A 100 mm long piece of self-adhesive tape was applied on the mechanically cleaned
surface. The tape was evenly pressed and left on the surface for about 10 s. Next, the
tape was peeled from the surface, glued to a clean white sheet of paper and was com-
pared with the purity evaluation range (ISO 8502-3). A more accurate determination of

Table 2. The chemical composition and grain size of the blasting abrasives.

Grain size
Abrasive Chemical composition (%) (mm)
Steel shot C Mn Si P S 0.9
0.75 1.20 0.60 1.1 0.60 1.1 max. 0.04 max. 0.04

Steel grit C Mn Si P S 0.9


min. 0.75 min. 0.40 max. 0.70 max. 0.04 max. 0.04

Brown corundum Al2O3 SiO2 Fe2O3 TiO2 CaO Na2O 0.9


>95.0 <1.00 <0.30 2.8 0.30 0.12

Zirblast ZrO2 SiO2 others 0.8


67 30 3
1758 A. Guzanov et al.

Table 3. Blasting parameters.


Blasting pressure MPa 0.4
Blasting nozzle diameter mm 9
Stand-off distance mm 200
Blasting time s 15
Downloaded by [Northeastern University] at 18:02 21 November 2014

Figure 2. (a) Initial contact of bristle tips with accelerator bar and (b) acceleration of bristle tip
towards the target surface.[13]

surface cleanliness was done by reectance measurement of the tape applied to the
paper. Reectivity was expressed in %, and consider a reectance of the white clean
paper sheet 100%. The apparatus for measuring the reectivity was calibrated to the
same type of clean self-adhesive tape glued to the white paper.

Micro-geometry of the surfaces


The surface roughness was evaluated according to ISO 4287 using the stylus prolome-
ter Surftest SJ 301, Mitutoyo (Japan), expressed by parameters Ra arithmetical
mean deviation of the assessed prole and Rz maximum height of the prole
(Table 4).
The 3D imaging of surfaces was obtained using a stereomicroscope. The actual sur-
face area of the treated surfaces was found using software Matlab and Surfer. Prepared
surfaces were evaluated also using a fractal analysis through software HarFA
(Harmonic and Fractal Image Analyser),[14] as explained in the following.

Fractal analysis of surfaces


Detailed evaluation of the segmentation process of the surfaces was given in [15].
The assumption for the evaluation of the degree of surface irregularity from the
surface images, is the fact that the grey shade of each pixel in the digital image
represents the real height, at a particular place on the prole of a real surface. The

Table 4. The selected parameters of roughness measurement.


Sampling length c 2.5 mm
Number of sampling lengths N 5
Measured prole R (system of the median line)
Filter Gauss
Evaluation length ln 12.5 mm
Journal of Adhesion Science and Technology 1759

surface images were at rst thresholded and next processed with a software for the
determination of fractal dimensions,[14] by using the box counting method. The fractal
dimension D is a value which can vary within the limits 1 < D < 2.[1618] When the
value D is closer to 2, the evaluated surface is more irregular. The box counting
method consists of counting the squares needed to cover the tested blackwhite object
completely.[1921] According to Refs. [14,22], there was used a modication of the
box counting method. The squares which are completely black (NB), the squares which
cover the border of the black and white area of the image were counted separately
(NBW). Finally, the squares which contained just the white part of the evaluated image
were counted separately, i.e. those which are completely white (NW) (Figure 3).
Using this modication three fractal dimensions DB, DBW and DW were obtained,
which characterise the fractal properties of the black part of the evaluated image DB,
the black-and-white border of the evaluated image DBW (this information seems to be
Downloaded by [Northeastern University] at 18:02 21 November 2014

the most interesting, because surface irregularity is proportional to the size of the
black-white border) and the fractal properties of the white part of the evaluated image
DW. If such measurements for various sizes of the squares will be carry on, the individ-
ual fractal dimensions as the slope of the following functions can be established:
NB r DB log1=r kB (1)

NBW r DBW log1=r kBW (2)

NW r DW log1=r kW (3)
where r represents the size of the squares and k is the fractal measure.
For purpose of the research will be considered only DBW, which characterises the
fractal properties of the evaluated surface the best. However, the rst step before
calculating the fractal dimension was to convert each image in the grey scale into a

Figure 3. Counting of black (NB), black-and-white (NBW) and white (NW) squares.
1760 A. Guzanov et al.
Downloaded by [Northeastern University] at 18:02 21 November 2014

Figure 4. (a) Fractal spectrum computed from the evaluated image, Ip = 109; (b) image of the
surface in gray scale; and (c) image of the surface after threshloding.
Journal of Adhesion Science and Technology 1761

black-and-white image, because fractal dimension can only be determined from black-
and-white images. Conversion of the grey scale image into a black-and-white image
can be done by thresholding. Thresholding is the conversion of a grey-scale image into
a black-and-white image using the so-called threshold intensity of the shade Ip. A new
black-and-white image is formed as follows: all pixels of the original image with
intensity Ii < Ip (where Ii is the intensity of a particular pixel) are shown as black,
while pixels with Ii > Ip are shown as white.
It is necessary to determine which grey scale contains the most information on the
irregular character of a given surface and, thus, may be considered as threshold inten-
sity Ip. Therefore, using the software mentioned above, the fractal spectrum was
calculated, i.e. the fractal dimension was computed for each shade of grey (256), from
0 (black) to 255 (white). The shade of grey for which the fractal dimension DBW has
its maximum value, is considered as the threshold intensity Ip. Figure 4 shows the
Downloaded by [Northeastern University] at 18:02 21 November 2014

fractal spectrum computed from the evaluated image where Ip = 109.


Consequently, the parameter Ip was determined for each evaluated-surface image.
After determining Ip, the thresholding of each surface image was performed and the
fractal dimension DBW for the black-and-white border of each image was determined.
Since the blasted surfaces have highly random characteristics, 20 different locations on
the surface were chosen to evaluate the fractal dimension for a single specimen. It was
assumed from the results obtained before [15] and from Refs. [14,1618] that the sur-
faces with the highest fractal dimension DBW are the most rugged and should show the
highest adhesion of the coating. Now, it was necessary to examine these assumptions
experimentally.

Figure 5. Schematic of failure for a pull-off test of the coatings. K = substrate, Z = paint Zinga,
L = epoxy adhesive, P = button, and F = tensile adhesion force.
1762 A. Guzanov et al.

Evaluation of the actual surface areas


The pre-treated surfaces were photographed and transformed into shades of grey. Using
Matlab, a spatial display of the surfaces was made, where each shade of grey is a
surface height prole. The vertical distances of the surface were imported into the
software Surfer, with that the surface area itself was calculated. The surface was
expressed only in pixels, in order to compare the size of the surfaces after various pre-
treatments.

Coating
The pre-treated surface was coated with Zinga by pneumatic spray. The dry lm thick-
ness was 65 5 m as measured by digital coating gauge.
Downloaded by [Northeastern University] at 18:02 21 November 2014

Testing
Adhesion of all paints was tested by pull-off test, EN 24624, using a tensile button-to-
plate conguration (Figure 5). Tensile adhesion strength and locus of failure was
investigated.
The coating was evaluated also using the cross-cut test, ISO 2409. The pull-off test
and the cross-cut test were performed on the coating (as-sprayed), as well as on the
coatings that had been exposed to different exposure atmospheres. Table 5 provides the
types of exposure atmospheres and exposure modes.
Table 6 provides an overview of the exposure of particular pre-treatments.

Table 5. Exposure conditions of the coatings.

Time of
Mode Exposure environment Conditions of exposure exposure
1 Humid atmosphere containing sulphur 100% RH, 40 2 C; SO2 21 days
dioxide accelerated corrosion test contamination
2 Sulphur dioxide accelerated corrosion test Wet period: 100% RH, 40 Wet period:
with condensation of water vapour (wet 2 C, SO2 contamination; 4 days
period)/UV radiation test (dry period)
Dry period: room Dry period:
temperature, UV radiation, 3 days
wave length 340 nm
Total duration
of exposure:
28 days
3 Indoor environment/environment with low Dry low temperature: 8 C 12 h
temperature
Room temperature 12 h
Total duration
of exposure:
28 days
4 Organic spreading material Solmag S Immersion in a 5% solution 21 days
of Solmag S in distilled
water
5 Solution of road salt Immersion in a solution of 21 days
5% NaCl in distilled water
Journal of Adhesion Science and Technology 1763

Table 6. Overview of the exposure of particular pre-treatments.

Pre-treatment Mode of exposure


Shot 1
Grit 1
Brown corundum 1, 2, 3, 4, 5
Zirblast 1
MBX blaster 1, 4, 5

Results and discussion


The 3D appearances of the pre-treated surfaces are shown in Figure 6.
With the application of the selected pre-treatment methods, three types of surfaces
arose. The rst type consists of many intersecting spherical canopies as a result of the
Downloaded by [Northeastern University] at 18:02 21 November 2014

blasting process and abrasives (steel shot and zirblast). The resulting surfaces are rela-
tively smooth (especially with the surface treated by zirblast) and dont contain a lot of
places for the mechanical anchoring of coating. The second type of surface is made up
of randomly spaced notches as a result of sharp angular abrasive blasting (by steel grit
and corundum). Notches on the surface treated by steel grit are deeper and less numer-
ous, by corundum are numerous and shallow, although the grain size of both abrasives
used were the same. The different depth of the notches was attributed to the fact that
the density of corundum is lower than that of the steel grit. The different density of the
notches was caused by the fact that there is a higher amount of abrasive grains of
corundum per unit than for the same weight of a unit of steel grit. The third type of
surface preparation technology was the MBX Blaster. It has a specic type of inequal-
ity, i.e. spade-like notches. These notches actually do not cover the entire surface con-
sistently but are distributed in a relatively uniform spacing. These basic types of
surfaces can be seen in Figure 7.
The results of the pull-off test could conrm these observations.
Roughness of the individual surfaces found by prolometer is listed in Table 7 as
an arithmetic mean from 10 measurements.
The fractal dimension DBW was calculated for the evaluated surfaces. The results
are summarised in Table 8.
The larger the value of DBW is, the more fragmented the surface. In Table 8, the
prepared surfaces are ranked according to segmentation.
Their appearance after thresholding by shading the borders is shown in Figure 8.

Results of the evaluation of surface cleanliness


Table 9 shows the results of the light reectivity from the tape which had been
removed from each cleaned surface, as well as from the substrate before cleaning.
The results let suggest that all types of surface preparation technology, except the
MBX Blaster, had a higher level of purity compared to the untreated steel surface. The
MBX Blaster technology loosened contaminants from the surface, but didnt remove
them from the surface of the material. The presence of these contaminants is likely to
adversely affect the adhesion of the coating applied.
The relative size of the actual nished surfaces vs. the ideal geometric surface
(expressed in pixels using image analysis) is shown in Figure 9.
1764 A. Guzanov et al.
Downloaded by [Northeastern University] at 18:02 21 November 2014

Figure 6. The appearance of the prepared surfaces: (a) SG, (b) SS, (c) BC, (d) ZB and
(e) MBX.
Journal of Adhesion Science and Technology 1765
Downloaded by [Northeastern University] at 18:02 21 November 2014

Figure 7. Metallographic sections of the prepared surfaces.

Figure 9 shows that all types of pre-treatment contributed towards increasing the
actual surface area compared to the geometric surface area. The largest real surface area
was found after surface preparation by zirblast, followed by SG, SS, MBX Blaster and
BC. The nominal geometric surface area in Figure 9 is represented by a solid black
horizontal line. This line represents the assessed area of 220 220 pixels. The surface
area before the pre-treatment was not actually perfectly smooth; it was larger than
48,000 pixels. Therefore, the increase of surface area compared to a real non-treated
surface was not as large as seen in the gure.
1766 A. Guzanov et al.

Table 7. Roughness of the prepared surfaces.

Surface preparation Ra (m) Rz (m)


Steel shot 8.68 0.48 50.75 4.23
Steel grit 9.1 0.54 55.69 6.58
Corundum 6.18 0.41 40.31 4.52
Zirblast 8.17 0.63 52.49 5.25
MBX blaster 3.35 0.87 21.64 8.40

Table 8. Fractal dimension of the evaluated surfaces.

IP DBW
Downloaded by [Northeastern University] at 18:02 21 November 2014

BC 52 1.832 0.032
MBX 55 1.820 0.030
SS 41 1.815 0.029
SG 47 1.792 0.034
ZB 52 1.751 0.021

Figure 8. The resulting surface images after thresholding.

Table 9. Light reectivity of cleaned surfaces.

Surface Light reectivity (%)


ZB 97.899.2
SG 96.597.8
BC 94.896.5
SS 93.193.9
Base material S235 83.587.1
MBX Blaster 58.970.9
Journal of Adhesion Science and Technology 1767

Results of the cross-cut test


The results of the cross-cut test and appearance of the grids are shown in Table 10.
The results point the fact that the pre-treatment by abrasive zirblast and the MBX
Blaster technology is critical for ensuring the coating adhesion. The cross-cut test
results were not sensitive to environmental exposure, but rather to the purity and also
the nature of the emerging unevenness.

Result of pull-off tests


The results of pull-off tests are summarised in Table 11.
The adhesion of the coating applied on substrates pre-treated by round blasting
abrasives (SS and ZB) was relatively low. The coating was not good mechanical
anchored, namely when used ZB see Table 10. The coating showed peeling in the
Downloaded by [Northeastern University] at 18:02 21 November 2014

range of 80100% of the coated area. The adhesion strength of the coating was higher
on the substrate pre-treated with SS than with ZB.
Adhesion of the coating on substrates pre-treated by angular blasting abrasives (SG
and BC) was higher than with round abrasives. In most cases, cohesive failure in coat-
ing or between coating/adhesive occured; this showed good mechanical anchoring of
the coating.
The coating applied to a substrate treated by MBX Blaster technology had relatively
good adhesion (as sprayed), but prevailed adhesive failure of the substrate/paint. The
coating after exposure in environments 4 and 5 failed adhesively on 4060% of the
area.
Table 12 summarises results of individual preparation technology according to
results obtained by particular criteria.
Each surface pre-treatment was evaluated by score from 1 to 5. Value 1 was
assigned to the pre-treatment technology, which gave the best results obtained in partic-
ular category listed in Table 12 (substrate surface characterised by high arithmetic aver-
age of the absolute values of the roughness prole Ra, high fractal dimension DBW,
good surface cleanlines, large surface area, good adhesion assessed by cross-cut and
pull-off test), the worst result has been assigned with the value 5. Appropriate pre-
treatment technology is the technology, which achieves the lowest overall score (pre-
treatment using steel grit and brown corrundum) (Table 12).

Figure 9. Actual surface area of treated surfaces.


1768 A. Guzanov et al.

Table 10. Results of the cross-cut test.


Downloaded by [Northeastern University] at 18:02 21 November 2014

Discussion
From the results mentioned above, it is possible to make the following conclusions
concerning the different methods of surface preparation:

 Preparation by steel shot: This abrasive has a relatively high roughening effect,
but because of its shape, it doesnt create a sufciently dissected prole for
anchoring of the coating. The surface contamination, although weakened, still
remained, loosely adhered to the surface, which aggravates the coating adhesion.
The actual surface area of the prepared surfaces was relatively large in all prepa-
ration methods, which ensures equal default conditions for the coating to have a
sufciently large contact area and good conditions for adhesion. Therefore, the
resulting adhesion of the coating is due to the impact of other factors, such as
Journal of Adhesion Science and Technology 1769

Table 11. Results of the pull-off test.

Surface Mode of Adhesion


preparation exposure (MPa) Failure mode
Steel shot As sprayed 0.39 20% coh paint
80% adh substrate/paint
1 0.16 10% adh substrate/paint 90% coh paint/adhesive
Steel grit As sprayed 0.67 90% coh paint
10% coh adhesive/button
1 0.17 90% coh paint
10% coh adhesive/button
Corundum As sprayed 0.74 30% adh substrate/paint
70% coh paint/adhesive
1 0.44 90% coh paint/adhesive, 10% coh in paint
2 0.29 5% coh in paint, 95% coh paint/adhesive
Downloaded by [Northeastern University] at 18:02 21 November 2014

3 0.62 100% coh in paint


4 0.33 100% coh in paint
5 0.34 100% coh paint/adhesive
ZB As sprayed 0.23 100% adh substrate/paint
1 0.13 85% adh substrate/paint
15% coh paint/adhesive
MBX blaster As sprayed 0.42 80% adh substrate/paint 20% coh paint

1 0.24 30% adh substrate/paint, 20% coh in paint, 50%


coh paint/adhesive
4 0.31 40% adh substrate/paint, 60% coh in paint
5 0.37 60% adh substrate/paint, 40% coh paint/adhesive
adh = adhesive, coh = cohesive.

Table 12. Results of preparation technology according to particular criteria.

Attribute
Actual size Cross- Type of failure after
Abrasives Ra DBW cleanliness of surface cut test Adhesion pull-off test Score
SG 1 4 2 2 1 2 2 14
SS 4 3 4 4 1 4 4 24
BC 2 1 3 3 1 1 1 12
ZB 3 5 1 1 5 5 5 25
MBX 5 2 5 5 4 3 3 27

surface cleanliness, topography, height of inequalities, etc. The cross-cut test


showed good adhesion of the coating, but in stricter conditions of perpendicular
pulling-off adhesion showed insufcient the coating showed peeling over a rel-
atively large area (see Table 11).
 Preparation by steel grit: This has the highest roughening effect of all evaluated
methods of surface preparation. Prepared surface belongs to the relatively less
dissected surfaces. The surface was relatively clean, and the applied coating
showed a very good adhesion in the cross-cut test as well as in the pull-off test.
 Preparation by brown corundum: This, due to its low density, has a relatively
small roughening effect compared to other methods of surface preparation, which
was reected by the smallest increase in surface area, but the resulting surface is
1770 A. Guzanov et al.

rated as the most dissected. Surface cleanliness is average, due to the known
increased dustiness of non-metallic mineral abrasives. The applied coating
showed the best adhesion in the cross-cut test, as well as in the pull-off test.
 Preparation by zirblast: Due to its average density, it had an average roughening
effect. Due to its circular shape creates at least dissected, but the clearest and
largest surface, but like SS it did not create good conditions for anchoring the
coating, which results in the worst adhesion coating in the cross-cut test as well
as in the pull-off test.
 Preparation by MBX Blaster: despite the lowest roughening effect, it created a
dissected surface with good conditions for anchoring the coating, but the rela-
tively high content of impurities on the prepared surface prevents to achieve suf-
cient adhesion. This was reected in the relatively low coating adhesion. It
should be noted that unlike other methods of surface preparation, the results of
Downloaded by [Northeastern University] at 18:02 21 November 2014

this technology depend strongly on the technology implementation by the particu-


lar operator. It is difcult to ensure repeatable identical conditions of the treat-
ment. The quality of the prepared surface also depends on the wear of the bristle
edges. Therefore, the unfavourable results of this study only apply to that particu-
lar operator of the implementation and cannot be generalised. Similarly, this does
not impair the strengths of this technology and its suitability for small cleaning
jobs in the eld, where it is much easier to work with the simple, small device
than with big blasting equipment.
 Based on presented results, it can be concluded that among all evaluated blasting
abrasives brown corundum and steel grit can be recommended for blasting before
application of zinc-lled coatings.

Conclusion
Experimental results point out that the highest roughening effect of all evaluated blast-
ing abrasives showed steel grit. Surface blasting with steel grit showed good purity
without secondary pollution and subsequently applied coating reached satisfactory
adhesion indicated by cross-cut test and pull-off test. Among evaluated non-metallic
blasting abrasives, brown corundum conrmed its ability to create highly irregular sur-
faces. This is important for optimal anchoring of the coating and its adhesion.

 The impact of the exposure environment can be evaluated as follows: all the
environments in which the coating came into contact with high humidity (con-
densing chamber, condensing chamber combined with UV radiation, immersion
in solutions of road salt), showed corrosion of the zinc contained in the coating.
The zinc corrosion led to change in volume of zinc particles in the coating. It
induced change in the stress conditions in the coating. This resulted in a reduc-
tion of the proportion of adhesive failure of the coating, compared to the as
sprayed state in all methods of preparation.
 Coating applied on surface, pre-treated by brown corundum was studied at the
environment with low temperatures (8 C). It led to expansion changes at the
interface of the substrate/coating, but did not result in reduced the coating adhe-
sion. The coating showed excellent ductility and adaptation to these changes. In
all used exposure atmospheres no corrosion of the substrate occurred; the zinc
contained in the coating had completely protected the steel by a cathodic
Journal of Adhesion Science and Technology 1771

protection mechanism. The coating Zinga applied on surface blasted using brown
corundum is therefore suitable for all structures exposed to the outside atmo-
sphere, such as bridges, tanks, crash barriers, wagons, lampposts, poles and other
structures near the road that may come into contact with de-icing materials.

Acknowledgement
This contribution is the result of the Grant Scientic Project VEGA No. 1/0600/13 and KEGA
059TUKE-4/2012.

References
[1] Watts JF. Microbeam analysis applied to adhesion, surfaces and interfaces. Microchim. Acta.
2009;164:379385.
Downloaded by [Northeastern University] at 18:02 21 November 2014

[2] Anderson GL, Stanley SD, Young GL, Brown RA, Evans KB, Wurth LA. The effects of
silicone contamination on bond performance of various bond systems. J. Adhes.
2010;86:11591177.
[3] Anderson GP, Devries KL. Predicting bond strength. J. Adhes. 1987;23:289302.
[4] Foister RT. Adhesive bonding to galvanized steel: II. substrate chemistry, morphology and
bond failure analysis. J. Adhes. 1987;24:279313.
[5] Williams O, Webb DP, Liu C, Firth P. Evaluation of contamination of ceramic surfaces and
its effect on epoxy bleed. Int. J. Adhes. Adhes. 2012;32:6169.
[6] Jeenjitkaew Ch, Luklinska Z, Guild F. Morphology and surface chemistry of kissing bonds
in adhesive joints produced by surface contamination. Int. J. Adhes. Adhes. 2010;30:
643653.
[7] Baldan A. Adhesion phenomena in bonded joints. Int. J. Adhes. Adhes. 2012;38:95116.
[8] Luke Y, Kaichang K, Tiezheng J, Guozheng L, Enhua H. Application of a new modied
epoxy adhesive for bonding uorine rubber to metal. J. Adhes. Sci. Technol. 2007;21:1
4831496.
[9] Adoberg E, Podgurski V, Peetsalu P, Lind L, Mikli V, Hvizdos P, Kulu P. The effect of sur-
face pre-treatment and coating post-treatment to the properties of TiN coatings. Est. J. Eng.
2012;18:185192.
[10] Bilmes GM, Orzi DJO, Martnez OE, Lencina A. A real time method for surface cleanliness
measurement. Appl. Phys. B. 2006;82:643648.
[11] Pocius AV. Adhesion and adhesives technology. Munich: Hanser; 2002.
[12] Production program of U.S. Steel [Internet]. Kosice: U.S. Steel; [cited 2012 Nov 22].
Available from: http://www.usske.sk/
[13] MONTI green surface preparation systems [Internet]. Germany: Werkzeuge GmbH;
[cited 2013 Feb 10]. Available from: http://www.monti.de/en
[14] Buchnicek M, Nezadal M, Zmeskal O. Image Science Lectures, HarFA Harmonic and
Fractal Image Analyser [Internet]. Software product; [cited 2012 Nov 21]. Available from:
http://www.fch.vutbr.cz/lectures/imagesci/
[15] Kniewald D, Guzanova A, Brezinova J. Utilization of fractal analysis in strength prediction
of adhesively-bonded joints. J. Adhes. Sci. Technol. 2008;22:113.
[16] Mandelbrot BB. Fractal geometry of nature. New York, NY: Freeman; 1983.
[17] Mandelbrot BB. Self-afne fractals and fractal dimension. Phys. Scr. 1985;32:257260.
[18] Feder J. Fractals. New York, NY: Plenum Press; 1988.
[19] Amada S, Satoh A. Fractal analysis of surfaces roughened by grit blasting. J. Adhes. Sci.
Technol. 2000;14:2741.
[20] Amada S, Hirose T. Inuence of grit blasting pre-treatment on the adhesion strength of plasma
sprayed coatings: fractal analysis of roughness. Surf. Coat. Technol. 1998;102:132137.
[21] Amada S, Yamada H. Introduction of fractal dimension to adhesive strength evaluation of
plasma-sprayed coatings. Surf. Coat. Technol. 1996;78:5055.
[22] Zmeskal O, Nezadal M, Buchnicek M. Fractal analysis of printed structures. J. Imag. Sci.
Tech. 2002;46:453456.

You might also like