(Gilho Kang)
1*
(Wonjong Choi)
1
Copyright © 2021 The Korean Institute of Metals and Materials
Key words(Korean)
adhesive, bonding strength, anodizing, surface roughness
1. Introduction
Bond strength and durability depend on the type of interaction between the adhesive
and the adherend. The type of interaction depends upon the chemical makeup of the
adhesives and adherends and also the topography of the adherend surface [1-3]. Adhesion can be improved by mechanically removing the weak surface layer, increasing
the substrate surface roughness, surface area and wettability, and improving the energy
dissipation mechanism of the adhesive. There is no universal adhesion model that accurately
describes the bonding properties of various bonded joints, the various combinations
of attachments as well as adhesives exposed to various environments. Almost all adhesion
models can be grouped into five major adhesion theories – the mechanical interlocking
theory, chemical/adsorption theory, diffusion theory, electrical theory and weak boundary
layer theory [4,5].
Mechanical interlocking is a major contributing factor in the adhesion of polymers,
metals and other materials [6-8]. The adherend surface is mechanically or chemically treated to obtain the desired
porosity and rough adherend surface morphology. Chemical treatments involve polishing,
etching, anodizing and other special treatments such as plasma and laser techniques
[9-18]. Studies have shown that a two-stage polarization process can increase surface porosity
and roughness [19]. Mechanical treatments usually involve scrubbing and polishing with sandpaper and
grit blasting. Adhesion via mechanical interlocking requires that the adhesive penetrates
the geometric irregularities of the adherend surface. Hennemann and Brockmann provided
evidence of penetration of thermoset adhesives into the pores of anodized oxide [20]. The degree to which the polymer penetrates into the oxide pores depends on the
wettability of the surface and the shape of the pores. The chemical/adsorption properties
between the adhesive and adherend are also important for adhesion. These include the
energy of adhesion in terms of the surface energy of the materials. Joannie W. Chin
et al. measured the surface energy and chemistry of a toughened BMI composite where
the surface was pretreated via solvent wiping, peeling, plying, grit-blistering and
oxygen plasma. They subsequently facilitated a comparison using double lap shear and
wedge tests [21]. Kunio Uehara and Mitsuru Sakurai studied bonding strength in relation to the surface
roughness of the joining parts. They concluded that an optimum surface roughness (3~6
μm) existed in determining the tensile strength of the adhesion, and the curve trend
showing the relationship between bond strength and surface roughness could be explained
by considering three factors: (1) the strength based on the adhesion theory (2) the
surface area effect (3) the notch effect due to surface roughness [22].
The purpose of this study is to clarify the impact of the surface treatment of an
aluminum adherend, from the aspects of micro-roughness and surface energy. The surface
treatment of the aluminum involved solvent wiping, Forest Products Laboratory (FPL)
etching, phosphoric acid anodizing (PPA) and chromic acid anodizing (CAA). Single
lap-shear and fatigue tests were performed to compare the adhesion properties for
each pretreated surface.
2. Experiment
2.1 Materials
Cytec FM 300M epoxy-based film adhesive, widely used to bond metal to metal, metal
to composite, and metal to honeycomb, was used. FM 300M adhesive film has a high level
of elongation and toughness along with high ultimate shear strength, which is particularly
suitable for redistributing the high shear stress concentration of graphite epoxy-to-metal
bonds and can accommodate the low interlaminar shear strength of composites [23]. It has excellent fatigue resistance in these joints. Tricot carriers, tightly woven
in properly designed and machined joints, provide a certain level of electrical insulation
between the metal and graphite composites to reduce galvanic corrosion. 1.6-mm thick
bare Al 2024-T3 (Alcore, Inc.) was used for the adherend.
2.2 Surface pretreatment
A summary of the surface treatment is presented in Table 1. The aluminum surface cleaning consisted of ultrasonic cleaning for 5 min in an acetone
solution followed by rinsing with distilled water and then drying at 60 °C.
FPL etching is a common chemical method that provides a rough surface for bonding.
It uses a mixture of sodium dichromate and sulfuric acid, and can be used as a standalone
pretreatment for bonding or in conjunction with an anodizing process. The specification
covering FPL etching is ASTM D 2674-72. The FPL etching procedure is as follows. The
aluminum surface was degreased in both vapor and hot liquid 1,1,1-trichloroethane
for 4 min at 88 °C. Alkaline cleaning was carried out for 13 min at 50 °C and was
then rinsed for 4 min at room temperature. The FPL etching process was carried out
for 13 min at 63 °C and was then rinsed for 5 min at 23 °C. Finally, it was dried
at 60 °C. In addition to etching, some coupons were also anodized in phosphoric acid
for 23min at 23 °C, 15 V and 1500A.
Another anodizing pretreatment was carried out in a chromic acid solution for 35 min
at 34 °C, 22V and 2000A. After the chromic acid anodizing treatment, it was sealed
in a dilute chromate solution for 25 min at 90 °C.
2.3 Surface energy measurement
Surface free energy (γ) generally consists of a dispersion component (γD) and a polar component (γP), which arise from two types of intermolecular forces, dispersion and polar, as shown
by
The surface tension leading to the formation of the angle is pictured in Fig 1.
This relationship was described by Young in 1805, resulting in the following well
known expression:
πe is the spreading pressure, which turns out to be negligible in liquids where the
contact angle of the polymer surface is not zero. The interfacial tension between
the solid and the liquid was obtained using the geometric-mean method as
Combining the above two equations and neglecting the spreading pressure gives [24].
We calculated the surface energy of the aluminum using the Young-Depre relationship
from two liquids for which the surface energy is known. We used water and diiodomethane.
For the water, the values of
γ
S
V
D
and
γ
S
V
P
used in the calculation were 51 and 21.8. For the diiodomethane, the values of
γ
S
V
D
and
γ
S
V
P
were 50.42 and 0.38 [25]. A Sigma 70 (KSV surface tensiometer) was employed to carry this out.
The work of adhesion, WA, in turn is defined as the energy per unit area of the interaction between the liquid
and solid:
Eliminating YSL from the combined Young equation enabled the Young-Dupre relationship to be determined,
as follows
According to Owens and Wendt's geometric approach, the work of adhesion, WA, between the solid and liquid is equal to the sum of the dispersive and polar interactions
[26,27].
The roughness correction factor had to be calculated because the difference in surface
roughness can change the measured contact angle and affect subsequent surface energy
calculations [21]. Samples and a smooth slide-glass plate were coated with approximately 50 nm of
gold in a sputter coater. The slide-glass plate served as a reference smooth surface.
The contact angles of 10 drops of liquid were measured on the gold-coated samples
as well as the gold-coated slide-glass plate. A roughness correction factor, RC was calculated via
where θ is the contact angle of a liquid with a known surface energy. These roughness correction
factors (RC) were used to correct the contact angle values of various pretreatment substrate
surfaces.
2.4 Surface roughness measurement
Surface analysis was performed to confirm the changes introduced to the Al substrate
by the various pretreatments. An SE1700α laser scanning microscope (LSM) (Carl Zeiss
Alpha 7) and SPM-400 atomic force microscope (AFM) (Seiko Instruments) were used to study the surface roughness and topography, respectively.
The scanning length differed for each measuring instrument because the resolution
of each system was very different. Several areas on each sample were scanned at widths
of 0.3 μm by 0.3 μm and 1 μm by 1 μm using AFM. From the LSM, scanning widths of 6 μm by 6 μm, 60 μm by 60 μm and 1000 μm by 1000
μm were used. For the SE1700α, the scanning length of the surface was 4 mm. We compared
the roughness of each different surface with the scanning length.
The root mean square (RMS) (Ra), average roughness (Rz) and Sratio were measured and compared using the AFM. The average roughness was the area between the roughness profile and its mean line,
or the integral of the absolute value of the roughness profile height over the evaluation
length.
The RMS roughness of a surface was calculated from another integral of the roughness
profile.
Rz (ISO) is a parameter that averages the height of the 5 highest peaks and the depth
of the 5 deepest valleys over the evaluation length. Sratio was calculated by
AM: measured surface area from AFM
AF: measured flat area
2.5 Bonding properties
Single lap-shear (SLS) test coupon profiles of the rectangular cross-section with
a thickness of 1.6 mm and width of 25.4 mm were cut at 101.6 mm, and lap shear joints
were produced with a 12.7-mm bond overlap according to ASTM D1002. The specimens that
were bonded with the film adhesive were completely cured and bonded at 175 °C for
60 min in an autoclave at 276 kPa.
The specimens were immersed in water at 70 °C for 60 days, and some specimens were
tested under salt spray conditions (5% NaCl, 200 hr.) using a salt spray tester (SUGA,
Model ST90). The lap-shear strength was subsequently measured at a test speed of 1.3
mm/min.
The specimen’s geometry of fatigue tests was identical to the SLS test. Fatigue tests
were performed on an Instron servohydraulic machine in constant amplitude load. All
fatigue tests were conducted at room temperature with a stress ratio of 0.5, a maximum
load of 500 kgf and a tension of 30 Hz.
3. RESULTS AND DISCUSSION
3.1 Surface free energy
The surface energy and the work of adhesion (WA) for each pretreated aluminum surface are shown in Figs 2 and 3, respectively. The polarity component of the cleaned aluminum surface was almost
zero, but after FPL etching and anodizing, it rose to about 20 mJ/m2. The PAA treated surface had the highest surface energy and work of adhesion. However,
this was a result of not considering the effects of surface roughness. The roughness
correction factor for each surface condition was measured relative to the slide-glass
plate. The roughness factor (Rc) was taken to be 1.0 for the slide-glass plate. The
roughness factor of each treated aluminum surface is shown in Table 2. Following the cleaning and CAA treatment, the roughness factor was about the same
as that of the glass plate, and the PAA treated surface showed the highest roughness
factor. The surface energy and work of adhesion, after taking into account the correction
factor, are shown in Figs 4~5, respectively. In this case, the CAA treated surface had the highest polarity, surface
energy and work of adhesion.
3.2 Surface roughness
The surface roughness is an important parameter in adhesive bonding. The average roughness
(Ra) of each pretreated aluminum surface along the scanned length is shown in Fig 6. The smaller the scanning length, the larger the difference in surface roughness.
The difference in surface roughness was noticeable when the scanning length was 0.3
μm. The scanning image from the AFM is shown in Fig 7. The anodized surface had a regular cell structure, while the cleaned surface had
an irregular shape. From this image, the cell size of the PAA treated surface was
about 35 to 45 nm. The cell structure of the CAA treated surface was finer because
this surface was finally sealed. Sealing is usually completed by reacting the CAA
surface with hot water. Hydrous oxides produced from this process fill the pores and
make an impermeable anodized layer that is stable in a wide range of atmospheric and
environmental conditions.
The Ra, Sratio, RMS and Rz values from the AFM are shown in Figs 8 and 9. The roughness values increased in the following order of surface treatment: Cleaning
< CAA < FPL < PAA.
3.3 Bonding properties
The results of the single lap shear test are shown in Fig 10. All coupons failed in cohesive failure mode, but the shear strength increased in
the following order: Cleaning<FPL <CAA<PAA. The shear strengths after the salt spray
test (5% NaCl, 200hr) are shown in Fig 11. Only the cleaned coupons failed under the perfect adhesive failure mode, and they
had nearly zero strength. The strength of the PAA and CAA coupons showed a slight
decline and failed with the cohesive failure mode. The shear strength after hot-water
immersion and hot-water/ice/thaw cycles are given in Figs 12 and 13. The cleaned coupons failed under the partial cohesive failure mode, and the PAA
and CAA coupons failed under the cohesive failure mode. After immersion in hot water
and hot water/ice/thawing, the shear strength of the cleaned coupon decreased by about
5 MPa, and the PAA coupon decreased by about 2 MPa. However, the strength of the CAA
coupon did not change. The fatigue test results are shown in Fig 14. The fatigue life of the CAA coupons increased 800 times compared to the cleaned
coupons, and the fatigue life of the PAA coupons increased 5,000 times compared to
the cleaned coupons.
4. Conclusion
The experimental results on the influence of surface energy and topography in aluminum-to-aluminum
bonding are as follows.
(1) The surface energy and work of adhesion of the aluminum alloy considering the
correction factor were higher in the order of Cleaning <FPL <PAA <CAA. The polarity
of the cleaned surface was almost zero, but after anodizing (CAA), the polarity had
increased significantly.
(2) The macroscopic roughness of the surfaces of Cleaning, PAA and CAA was similar,
but the microscopic roughness (scanning length=0.3 μm) increased in the order of Cleaning<CAA
<FPL <PAA. The microscopic surface roughness of the PAA-treated sample was the highest.
(3) The lap-shear strength increased in the order of Cleaning<FPL <CAA <PAA. It can
be seen that the bonding strength is most strongly influenced by the microscopic roughness
of the adherend surface. This is because the larger the roughness, the larger the
surface area and the larger the bonding area. The CAA treated surface showed good
bond retention performance in harsh environments(hot water, salt spray, hot-water/ice/thawing
cycle). This seems to have occurred because the CAA-treated surface energy and work
of adhesion were the highest, taking into account the roughness correction factor,
and also had good corrosion resistance.
(4) The fatigue test most clearly showed the different effect of each surface pretreatment
on adhesion properties.
In conclusion, the microscopic roughness and surface energy of the aluminum alloy
were the main factors influencing aluminum adhesion properties and durability.