Microstructural characterization and corrosion behavior of multipass friction stir processed AA2219 aluminium alloy

Abstract

Multipass friction stir processing of AA 2219-T87 aluminium alloy to a depth of 2 mm in a 5 mm plate resulted in fine α-Al grains, reduction and dissolution of both eutectic phase (CuAl2) and the strengthening precipitates (CuAl2). Anodic polarization and electrochemical impedance tests in 3.5% NaCl showed an improved corrosion resistance of the processed alloy, which increased with the number of passes. Salt spray and immersion tests also showed improved resistance to corrosion. The increased resistance to corrosion is attributed to the dissolution of CuAl2 particles, which was established by XRD and DSC studies.

Full text

Microstructural characterization and corrosion behavior of multipass friction
stir processed AA2219 aluminium alloy
K. Surekha , B.S. Murty ⁎, K. Prasad Rao
Department of Metallurgical and Materials Engineering, Indian Institute of Technology Madras, Chennai, 600 036, India
Received 20 August 2007; accepted in revised form 1 February 2008
Available online 11 February 2008
Abstract
Multipass friction stir processing of AA 2219-T87 aluminium alloy to a depth of 2 mm in a 5 mm plate resulted in fine α-Al grains, reduction
and dissolution of both eutectic phase (CuAl
2
) and the strengthening precipitates (CuAl
2
). Anodic polarization and electrochemical impedance
tests in 3.5% NaCl showed an improved corrosion resistance of the processed alloy, which increased with the number of passes. Salt spray and
immersion tests also showed improved resistance to corrosion. The increased resistance to corrosion is attributed to the dissolution of CuAl
2
particles, which was established by XRD and DSC studies.
© 2008 Elsevier B.V. All rights reserved.
Keywords: Friction stir processing; AA2219 Aluminium alloy; Corrosion
1. Introduction
AA2219 (Al–6.5%Cu) alloy is a widely used age hardenable
alloy in aerospace and defence applications. It is a high strength
alloy with excellent weldability. However, its only disadvantage
is its poor corrosion resistance owing to the galvanic coupling
between the noble CuAl
2
precipitate and the matrix. If there
could be some means by which CuAl
2
could be taken into
solution or removed, the corrosion resistance of the alloy can be
improved, though this might lower the mechanical properties to
some extent. It is generally accepted that laser surface melting
(LSM) can be used for improving the corrosion resistance of
metallic alloys as a result of homogenization/refinement of
microstructure, dissolution/redistribution of precipitates or inclu-
sions and phase transformation. However, LSM needs costly
equipment and the operator should be well protected with all
accessories to avoid the hazardous effects of the laser. Similar
changes can be induced in an alloy by friction stir processing
(FSP), which is a cost-effective, environment and user friendly
technique. The present study is undertaken to study the effect of
FSP on corrosion behaviour of AA2219 alloy.
FSP is an emerging surface-engineering technology, which
uses the principles of friction stir welding [1,2] to process materials
in a variety of ways besides joining them. Mishra et al. [2]
developed FSP as a generic tool for microstructural modification
based on the basic principles of friction stir welding (FSW). FSP
can locally eliminate casting defects and refine microstructures,
thereby improving strength, ductility, resistance to corrosion,
formability and other properties. FSP is carried out by rotating and
plunging a specially designed cylindrical, shouldered tool with a
small diameter pin into the plate that is clamped firmly to the bed.
Frictional heat causes the metal to soften and allows the tool to
traverse along the plate. The FSP generates three distinct micro-
structural zones: the nugget, the thermo mechanically affected
zone (TMAZ) and the heat-affected zone (HAZ). The nugget is the
region through which the tool piece pin passes, and thus ex-
periences large deformation and high temperatures. It generally
consists of fine equiaxed grains due to recrystallisation. The
TMAZ adjacent to the nugget is the region where the metal is
plastically deformed as well as heated to a temperature, which is
not sufficient to cause recrystallisation. The HAZ experiences only
heating effect, with no mechanical deformation.
A
vailable online at www.sciencedirect.com
Surface & Coatings Technology 202 (2008) 4057 –4068
www.elsevier.com/locate/surfcoat
⁎Corresponding author. Tel.: +91 44 2257 4754; fax: +91 44 2257 4752.
E-mail addresses: surekhakrishnan@yahoo.co.in (K. Surekha),
murty@iitm.ac.in (B.S. Murty).
0257-8972/$ - see front matter © 2008 Elsevier B.V. All rights reserved.
doi:10.1016/j.surfcoat.2008.02.001

The corrosion properties of various zones of FSW have been
examined by a number of investigators. Corrosion attack in the
nugget has been found for AA2024-T351, AA5456-H116 and
AA7010-T7651 alloys [3]. Studies have shown attack to be pre-
dominantly in heat affected zone (HAZ), for example in AA2024-
T351 [4], AA7075-T651 [5], AA7075-T6 [6] alloys and for the
AA7050-T7651 alloy, corrosion occurred in the interface between
nugget and thermomechanically affected zone (TMAZ) [7].
Frankel and Xia [8] investigated pitting and stress corrosion
cracking behavior of FSWAA5454 alloy and compared them with
those of basealloy and gas tungsten arc welded (GTAW)samples.
In FSW, the pits formed in HAZ and in GTAW in the fusion zone.
FSW samples also showed pitting resistance greater than the
basemetal (BM) and GTAW welds. Corral et al. [9] investigated
the effect of FSW on the corrosion behavior of AA2024-T4 and
Al–Li alloys (AA2195) and showed that the diffusion-limiting
current densities and corrosion potentials of both AA2024 and
AA2195 FSW welds were nearly identical to those of the base
alloys for a 0.6 M NaCl solution. Zucchi et al. [10] reported that
the friction stir welded AA5083 alloy exhibited a higher corrosion
resistance in EXCO solution (4 M NaCl–0.5 M KNO
3
–0.1 M
HNO
3
) and a lower pitting tendency than the base alloy.
Meletis et al. [11] investigated stress corrosion cracking be-
havior of FSW AA7075-T6, AA2219-T87 and AA2195-T87
alloys and reported that SCC resistance of the welds is better than
the base metal. Pao et al. [12] studied corrosion fatigue crack
growth of friction stir welded AA7050 alloy and reported that in
both air and 3.5% NaCl solution, fatigue crack growth rates in FSW
HAZ are significantly lower than those in the base metal and the
weld. Lumsden et al. [5,13] and Paglia et al. [14] demonstrated that
the HAZ regions of friction stirred welds of AA7075, AA7010,
AA2024 and AA7050 alloys were more susceptible to intergra-
nular attack than the base alloy, TMAZ and nugget regions.
Jariyaboon et al. [15] reported the effect of welding parameters
during FSW, especially rotation speed and traverse speed, on
corrosion behaviour. All the corrosion studies reported till date
have been on FSW alloys. Though the principles of FSW and FSP
are same, the effect of FSP, a novel surface engineering technique,
on corrosion behaviour has not been studied yet. There have been a
few studies on the microstructure evolution, improvement in fa-
tigue resistance and superplasticity by multipass FSP [16–18].
However, no work has been reported on the effect of multipassing
on corrosion behaviour. The present study is focused on this aspect.
2. Experimental procedure
The material used in this work is AA2219-T87 alloy with the
nominal composition in wt.% of Cu–6.1, Mn–0.25, Zr–0.16,
V–0.09, Ti–0.05 and rest Al. The AA2219-T87 plates (250 ×
150 × 5 mm in size) were friction stir processed, with an
indigenously developed machine (3000 rpm, 15 HP and 25 kN)
at a constant axial force of 12 kN with a non-consumable threaded
tool made up of high-speed tool steel. The basic principle of FSP
is described in the previous section. Multipassing was done such
that one bead was laid exactly over the other. Up to three passes
were given at three rotation speeds of 800 (slow –S), 1200
(medium –M), 1600 (fast –F) rpm and two welding speeds (0.37
(slow –S) and 0.76 (fast –F) mm/s). The depth of the processed
region was 2 mm in a 5 mm thick plate. Single pass FSP was
carried out at a rotation speed of 800 rpm and traverse speed of
0.37 mm/s are named SS1, two passes as SS2 and three passes are
termed as SS3. Similarly, all other joints were named according to
the parameters used for the process and their nomenclature is
shown in Table 1. FF and FS3 samples are not shown in the table
as it was impossible to get a crack free bead with these parameters.
The parameter optimization and the selection of tool profile in the
present work was implemented taking cue from the work of
Elangovan et al. [19] on friction stir welding. The present work
being FSP, the effective volume of the material which acts as a
heat sink differs from that of welding. The microstructural ana-
lysis of the friction stir processed samples was carried out by
Table 1
Parameters used for friction stir processing
Samples Rotation speed (rpm) Travel speed (mm/s) Number of passes
SS 800 0.37 SS1, SS2, SS3
SF 800 0.76 SF1, SF2, SF3
MS 1200 0.37 MS1, MS2, MS3
MF 1200 0.76 MF1, MF2, MF3
FS 1600 0.37 FS1, FS2
Fig. 1. Optical micrographs of (a) BM and (b) nugget region in MS3 sample.
4058 K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

Leica optical microscope and FEI scanning electron microscope
(SEM). The samples were etched with Keller's reagent to reveal
the grain boundaries. Microhardness measurements were carried
out using Matsuzawa Vicker's micro hardness tester at 25 g load
for 30 s.
Corrosion studies have been carried out using potentiody-
namic polarization tests (ASTM G3) and Electrochemical Im-
pedance tests (EIS). Software based PAR basic electrochemical
system is used for the tests. A flat cell was used for all the
experiments and the 5 mm wide nugget is used as working
electrode, carbon is used as auxiliary electrode and saturated
calomel electrode as reference electrode. 3.5% NaCl was used as
the electrolyte. Potentiodynamic polarization tests were carried
out a scan rate of 0.166 mV/s with an initial potential of
−0.25 mV. The E
pit
value signifies the breakage of the passive
film and hence the corrosion current increases drastically with the
applied voltage after E
pit
. For the electrochemical impedance
tests, the samples were immersed in the electrolyte for 30 min
before the test. The samples were exposed (0.16 cm
2
)suchthat
only the nugget is subjected to the corrosion tests and the rest of
the areas were masked. EIS measurements were carried out in the
frequency range of 10 mHz to 100 kHz.
Immersion corrosion tests were performed on etched sam-
ples according to ASTM standard G110. The samples were
immersed in a solution of 57 g/l (0.98 M) NaCl and 10 ml/l
H
2
O
2
(0.09 M) for 6 h and the extent of corrosion attack was
observed in SEM. Salt Spray corrosion (ASTM B117) tests
Fig. 2. (a) and (b) SEM micrographs of base metal at different magnifications.
Fig. 3. SEM micrographs of nugget regions in (a) MS1, (b) MS2 and (c) MS3
samples.
4059K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

were carried out and weight loss was measured. SEM was used to
know the extent of corrosion. The top surface of the sample alone
was exposed to the fog and the rest of the sides were masked.
Grain size and particle size measurements were carried out
using an Image analyzer attached to an optical microscope. X-
ray diffraction (XRD) and differential scanning calorimetry
(DSC) measurements were carried out by Philips x-ray diffrac-
tometer and Netzsch differential scanning calorimeter, respec-
tively, to find the amount of CuAl
2
dissolved during FSP.
3. Results and discussion
The corrosion behaviour of the nugget region of the FSP
samples obtained at various rotational speeds and traverse speeds
(SS, MS andFS, SF and MF) followed a similar trend. The traverse
speed did not have any significant influence on the corrosion
behaviour, similar to the earlier observations by Jariyaboon et al.
[15]. In the present paper, the results of multipassing with one set of
rotational speed and traverse speed (MS) are discussed. Inter-
metallic phases formed during casting (eutectic solidification) and
those formed during aging are two types of second phase particles
in precipitation hardenable aluminium alloys. Both these second
phase particles influence corrosion behavior of the alloy and hence
the effect of multipassing on these particles is discussed below.
3.1. Microstructural characterization
Fig. 1 shows the optical micrographs of the nugget region of
the MS3 sample along with the base metal. Significant grain
refinement can be noticed in the alloy on FSP in comparison to
the base metal. Since the second phase particles were not dis-
cernible by optical microscopy, SEM studies were carried out to
know the dissolution of intermetallic particles formed during
solidification. Fig. 2(a)–(b) shows SEM micrographs of the
base metal at different magnification, which indicate that the as
cast grain size is very large in this alloy. Fig. 3(a)–(c) shows the
nugget region of MS samples on multipassing, which clearly
demonstrates that the grain size of the α-Al has significantly
decreased after FSP. In high strength Al alloys the nugget zone
contains both soluble and insoluble second phase particles
[20,21]. In AA2219 alloy, CuAl
2
precipitates are soluble and
Al
3
Zr and dispersoids formed with Ti and V are insoluble. The
second phase particles in Figs. 2 and 3 are identified as CuAl
2
particles by energy dispersive x-ray (EDX) microanalysis.
The histograms of grain sizes of different MS samples are
shown along with that of BM in Fig. 4.Table 2 gives the
average grain size of α-Al and second phase particle size values.
The average grain and particle sizes were calculated considering
Fig. 4. Average grain size of (a) BM, (b) MS1, (c) MS2 and (d) MS3 samples.
Table 2
Average grain and particle size in the nugget region of different friction stir
processed samples along with the base metal
Alloy condition Average grain size (μm) Average particle size (μm)
BM 67.4 20.9
MS1 6.2 5.2
MS2 6.7 4.5
MS3 7.0 3.5
4060 K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

at least hundred grains and particles. The BM shows a large
average grain size of 67.4 μm while the friction stir processed
sample showed an average grain size of 6.2 μm in the first pass
itself. With subsequent passes the average grain size showed a
marginal increase with MS2 and MS3 samples showing 6.7 μm
and 7 μm, respectively. The formation of these fine grains during
FSP can be attributed to dynamic recrystallization [22,23].Hassan
et al. [24] have reported that a low heat input during FSW results
in an exceptionally fine grain structure along with dissolution of
the precipitates. When the FSP is carried out with higher heat
inputs, the grains in the nugget are coarser. On multipassing, the
heat input is slightly increased and hence the small increase in
grain size on multipassing can be attributed to this factor. Fig. 5
shows a significant reduction in the average particle size of CuAl
2
on multipassing as observed earlier by Hassan et al. [24].The
average particle size of CuAl
2
of BM reduced from 20.9 to 4.9 μm
on the first pass and to 4.5 μm and 3.5 μm on second and third
passes. Thus, the present results confirmed that FSP reduces the
size of intermetallic particles, formed during solidification of the
alloy, by fracture of the particles.
The decrease in the number of Cu rich particles in the pro-
cessed alloys is further confirmed by EDX line scan studies. Fig. 6
shows the SEM images of MS samples along with the base metal
and the variation of copper concentration at various passes. It can
be observed that the copper concentration of the particles is very
high (~55 wt.%) in the base metal and it has been lowered
significantly by single pass (~15 wt.%). Further decrease in
concentration of copper is noticed in MS2 (~10 wt.%) and MS3
samples (~5 wt.%). This is due to refinement of the particle size,
which makes the matrix contribute significantly to the line scan
analysis as the number of passes increase.
TEM analysis was carried out to show the dissolution of
strengthening precipitates (CuAl
2
) since they cannot be resolved
by SEM studies. Fig. 7 shows the TEM images of MS samples
along with the base metal. TEM studies also showed that the size
and volume fraction of the precipitates decreases with increase in
the number of passes. Both eutectic CuAl
2
and the strengthening
precipitates influence corrosion and hence it has been established
by SEM and TEM studies that size and volume fraction of both
the CuAl
2
of the eutectic mixture and the strengthening pre-
cipitates decreased with multipass FSP.
The amount of second phase (both eutectic and strengthening
precipitates) dissolved during FSP has been calculated from the
XRD and DSC studies. Fig. 8 shows the XRD patterns of MS
samples along with the base metal. The XRD pattern is magnified
to clearly show the CuAl
2
peaks, which have much lower
intensity than Al peaks. The figure clearly shows that the intensity
of the CuAl
2
(JCPDS file number 25-0012) peaks decrease with
multipassing, which indicates dissolution of the precipitate par-
ticles. Similar dissolution of particles on heavy deformation has
been reported earlier [25,26]. The ratio of the area under the most
intense CuAl
2
XRD peak and the sum of the areas under most
intense CuAl
2
and Al peaks (JCPDS file number 04-0787) is
found for all FSP parameters and the base metal. Let the value
obtained be ‘X’. One minus ‘X’of the processed metal divided by
the ‘X’of the base metal gives the fraction of CuAl
2
dissolved at a
particular FSP parameter. It was found that the dissolution of
CuAl
2
is more at higher rotation speeds and with multipassing.
Fig. 5. Average CuAl
2
particle size in (a) BM, (b) MS1, (c) MS2 and (d) MS3 samples.
4061K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

Tab le 3 shows the amount of CuAl
2
dissolved on multipassing
based on the XRD calculations.
The amount of CuAl
2
dissolved was estimated from DSC
traces by considering the area of the peak corresponding to
CuAl
2
dissolution. For a particular FSP parameter, the
difference between the area under the CuAl
2
peak of the base
metal and the processed metal divided by the area under CuAl
2
peak of the base metal gives the fraction of CuAl
2
dissolved at a
particular FSP condition. Fig. 9 shows the DSC traces of the MS
samples. The dissolution is higher with higher rotation speeds
and multipassing, which confirmed the XRD results. However,
the DSC calculations showed a lower amount of CuAl
2
dissolution than XRD calculations for any given condition.
For example, the amount of CuAl
2
dissolved in MS1, MS2 and
MS3 samples were found as 42.8, 57.1, and 71.4%, respectively
by XRD measurements and as 23.1, 29.2, 36.9% respectively
Fig. 6. SEM-EDX line scan analysis of MS samples along with the BM.
4062 K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

by DSC measurements. The amount of CuAl
2
dissolved at other
parameters is shown in Table 2. The DSC results are more
reliable than the XRD results as the intensity of the XRD peak
of the precipitate not only depends on its weight fraction but
also its structure factor and a number of corrections to it.
3.2. Hardness
Dissolution of strengthening precipitates impairs the mecha-
nical properties. To have an insight into the mechanical pro-
perties, hardness measurements were carried out. The average
hardness values of MS samples on multipassing are shown in
Fig. 10. In all parameters, the nugget showed a lower hardness
compared to the base metal. This suggests dissolution of
Fig. 7. TEM images of (a) BM and nugget regions in (b) MS1, (c) MS2 and (d) MS3 samples.
Fig. 8. XRD patterns of MS samples on multipassing along with that of base metal.
Table 3
XRD and DSC results showing the amount of CuAl
2
dissolved during FSP
Process
parameter
Intensity of
CuAl
2
Intensity
of Al
ICuA12
IA1þICuA12
% of CuAl
2
dissolved
(XRD result)
% of CuAl
2
dissolved
(DSC result)
BM 124.37 1519.0 0.07 ––
MS1 82.02 1604.0 0.04 42.8 23.1
MS2 74.54 1851.8 0.03 57.1 29.2
MS3 70.75 2821.4 0.02 71.4 36.9
FS1 209.3 8243.9 0.02 71.4 27.5
FS2 24.6 1713.7 0.01 85.7 34.8
SS1 42.6 807.2 0.05 28.6 17.4
SS2 209.1 4825.6 0.04 42.8 24.7
SS3 95.6 2746.5 0.03 57.1 29.6
4063K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

precipitate particles, which was confirmed by XRD and DSC
results. Sato et al. [27] have examined the hardness profiles
associated with the microstructure in an FSW precipitation
hardenend AA6063-T5 alloy and observed similar results as in
the present case.
It was found that the area of the softened zone (nugget region)
and the hardness in the nugget region increased with increase in
thenumberofpasses.Theincreaseinareaofsoftenedzoneis
attributed to the increase in the heat input and the increase in
hardness with multipassing may be attributed to the decrease in
particle size of insoluble dispersoids. FSP creates a softened
region around the weld center in a number of precipitation-
hardened aluminium alloys. It was suggested that such a softening
is caused by coarsening and dissolution of strengthening pre-
cipitates during the FSW [28,29].
3.3. Corrosion behaviour
Potentiodynamic polarization tests were carried out to find the
pitting corrosion resistance. Anodic polarization curves were
obtained by exposing the nugget area alone to 3.5% NaCl solution,
whose pH is maintained at 10 by adding KOH. Fig. 11 shows the
potentiodynamic polarization curves of MS samples on multi-
passing. It was found that with three passes the pitting potential
was nobler (−466 mV) compared to single pass (−498 mV), two
passes (−484 mV) and the base metal (−587 mV). I
corr
gives the
direct measure of corrosion rate. The I
corr
value of the base metal is
869 µA and it is 77, 31 and 6 µA for MS1, MS2 and MS3 samples,
respectively. Table 4 lists the E
pit
and I
corr
values of MS samples
along with the base metal. From these results it can be seen that the
corrosion rate decreased drastically with FSP and further im-
provement in corrosion resistance is observed with increase in
number of passes.
Electrochemical Impedance Spectroscopy (EIS) results for
base material and the nugget exposed to 3.5% NaCl solution for
30 min are plotted in Fig. 12. The low frequency impedance
indicates the corrosion resistance of the surface. It was found that
the nugget of MS3 exhibits higher electrochemical corrosion
resistance (18.3 kΩcm
2
) than the base material (484.6 Ωcm
2
).
It was also found that the electrochemical resistance increased
Fig. 9. DSC traces of MS samples on multipassing.
Fig. 10. Average hardness value of MS samples on multipassing along with that
of the base metal.
Fig. 11. Potentiodynamic polarization curves of MS samples on multipassing.
Table 4
Corrosion values of FSP samples after pitting, impedance and salt spray tests
Alloy
condition
Corrosion rate
in mpy
E
pit
(mV) I
corr
(μA) Z(kΩcm
2
)
BM 15.7 −587 869.6 0.484
MS1 4.5 −498 77.8 9.6
MS2 4.0 −484 31.5 14.2
MS3 3.8 −466 6.3 18.3
Fig. 12. EIS curves for MS samples on multipassing.
4064 K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

with increase in the number of passes. The impedance values of
MS1 and MS2 samples were 9.6 and 14.3 kΩcm
2
, respectively.
Table 4 lists the impedance values of the MS samples along with
the base metal.
The galvanic coupling between the Al matrix and the CuAl
2
precipitate is the main reason for corrosion of AA2219 alloy.
Dissolution of the CuAl
2
precipitates decreases the sites for
galvanic coupling and hence increases the corrosion resistance.
Dissolution of the precipitates was confirmed by both XRD and
DSC techniques. From Fig. 13 it is evident that the corrosion
resistance increases with increase in number of passes. Fig. 13
shows the influence of CuAl
2
dissolved on the impedance
values. The amount of dissolution increased with the increase in
number of passes and hence the corrosion resistance also fol-
lowed the same trend.
Salt spray tests were carried out in 5% NaCl solution for
100 h to assess the uniform corrosion resistance. Weight loss
measurements were made to find the corrosion rate. Pits of very
small diameter were observed on the surface after corrosion test
and a continuous decrease in thickness over the entire surface
area of the metal was observed throughout the corrosion test.
Fig. 14 shows the SEM images of the MS samples after salt
spray test along with the base metal. Tab le 4 shows the cor-
rosion rate of the processed alloys after salt spray test. It is seen
that the base metal has corroded very severely and corrosion
products are seen throughout the surface. But in MS1 only a few
pits are seen. It can further be seen that the density and size of
pits in MS2 and MS3 are lower in comparison to MS1. This
Fig. 13. Influence of amount of CuAl
2
dissolved on impedance values.
Fig. 14. SEM images after salt spray test of MS samples on multipassing.
4065K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

increased corrosion rate of the base metal is due to the large
number of CuAl
2
sites available for the galvanic coupling
between the Al matrix and the CuAl
2
particles, which is clearly
revealed by the particle size. The corrosion rate in the base
metal is 15.7 mpy where as it is 4.5, 4.0 and 3.8 mpy in MS1,
MS2 and MS3 samples, respectively. From these results, it
can be inferred that with the increase in number of passes the
corrosion resistance increases. The corrosion product formed is
known to be Al(OH)
3
. EDX microanalysis of the corroded
products was carried out. However, as this technique can not
reveal the presence of hydrogen, the presence of Al(OH)
3
could
not be proved at the moment. Attempts are on to carry our FTIR
studies to prove the presence of Al(OH)
3
as a part of future
work.
To investigate the susceptibility to intergranular attack, 6 h
immersion test was carried out in a solution containing 57 g/l
NaCl and 10 ml/1 H
2
O
2
(30 vol.%). Fig. 15 shows the SEM
images of the samples after immersion test. The results are
consistent with the impedance, potentiodynamic and salt spray
tests. The number of passes is the primary factor in determining
the rate of attack. The corrosion attack is less in MS3 sample
compared to MS1, MS2 samples and the BM. It is seen the base
metal has corroded intergranularly whereas only a few pits were
seen in the processed alloys. Intergranular corrosion will occur
only when the following three conditions are simultaneously
met [30]:
1. Presence of a corrosive medium,
2. Difference in potential in the order of 100 mV between the
intermetallics and the matrix,
3. Continuous network of the intermetallics at the grain
boundaries such that intergranular cracks can propagate.
All the above three conditions are met in the case of base
metal. Hence, the base metal corroded intergranularly. The
corrosion products are identified as Cu
2
O in this case based on
the EDX microanalysis results as shown in Fig. 16. Copper
enrichment (54–58 wt.%) in the corrosion deposit was
observed, which was due to the selective dissolution of copper
rich intermetallic followed by redeposition of copper on the
surface in the form of oxide. In the case of FSP alloys, con-
tinuous network of CuAl
2
at grain boundaries is not observed
due to the breakage and dissolution of the intermetallic par-
ticles. Hence, there was no copper enrichment in the corrosion
Fig. 15. SEM images after G110 corrosion of MS samples on multipassing.
4066 K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

products and only a few pits were observed. The propagation of
intergranular corrosion starts at pits. This indicates that it needs
some more time for the intergranular cracks to propagate.
Hence, the processed alloys have better corrosion resistance
compared to the base metal. Fig. 17 shows the EDX analysis of
the corrosion products of MS1, which clearly indicates that the
corrosion product is Al
2
O
3
in this case.
4. Conclusions
FSP is a novel surface modification technique and in the
present work multipass FSP is carried out to improve the cor-
rosion resistance of AA 2219 aluminium alloy. The following
conclusions can be arrived at based on the present work.
1. The effectiveness of FSP in improving the corrosion
resistance has been demonstrated on the AA2219 alloy. All
the processed alloys, irrespective of the processing para-
meter studied showed superior corrosion resistance com-
paredtothebasemetal.
2. Number of passes during FSP has a significant effect on
the corrosion resistance. Three pass and two pass FSP alloy
showed better corrosion resistance in comparison to the
single pass processed alloy.
3. Dissolution of the CuAl
2
particles during FSP reduces the
number of sites available for galvanic coupling and hence a
reduction in the corrosion rate is observed in the processed
alloys.
Acknowledgement
The authors gratefully acknowledge the help and support ren-
dered in FSP by Prof. V. Balasubramanian of the Department of
Manufacturing Engineering, Annamalai University, Chidambaram.
References
[1] W.M. Thomas, E.D.Nicholas, J.C. Needham, M.G.Murch, P Smith and C.G.
Dawes, Friction Stir Butt Welding, G.B. Patent Application No. 9125978.8.
[2] R.S. Mishra, Z.Y. Ma, Mater. Sci. Eng. R50 (2005) 1.
[3] B.J. Connolly, A.J. Davenport, M. Jariyaboon, C. Padovani, R. Ambat,
S.W. Williams, D.A. Price, in: A. Wescott, C.J. Goodfellow, C.M. Lee
(Eds.), Proc. 5th International Friction Stir Welding Symposium, Metz,
France, 2004.
[4] S. Williams, R. Ambat, D. Price, M. Jariyaboon, A. Davenport, A.
Wescott, Mat. Sci. Forum 426 (2003) 2855.
[5] J.B. Lumsden, M.W. Mahoney, G. Pollock, C.G. Rhodes, Corrosion 55
(1999) 1127.
[6] C.S. Paglia, M.C. Carroll, B.C. Pitts, A.P. Reynolds, R.G. Buchheit, Mat.
Sci. Forum 396 (2002) 1677.
[7] J.B. Lumsden, M.W. Mahoney, C.G. Rhodes, G.A. Pollock, Corrosion 59
(2003) 212.
[8] G.S. Frankel, Z. Xia, Corrosion 55 (1999) 139.
[9] J. Corral, E.A. Trillo, Y. Li, L.E. Murr, J. Mater. Sci. Lett. 19 (2000) 2117.
[10] F. Zucchi, G. Trabanelli, V. Grassi, Mater. Corros. 52 (2001) 853.
Fig. 16. (a) SEM micrograph and (b) EDX spectrum and quantitative analysis of
the corrosion product labeled as ‘x’in (a) of BM after immersion test.
Fig. 17. (a) SEM micrograph and (b) EDX spectrum and quantitative analysis of
the corrosion product labeled as ‘x’in (a) of MS1 sample after immersion test.
4067K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068

[11] E.I. Meletis, P. Gupta, F. Nave, in: K.V. Jata, M.W. Mahoney, R.S. Mishra,
S.L. Semiatin, T. Lienert (Eds.), Proc. Friction Stir Welding and Processing
II, TMS, Warrendale, PA, USA, 2003, p. 107.
[12] P.S. Pao, S.J. Gill, C.R. Feng, K.K. Sankaran, Scr. Mater. 45 (2001) 605.
[13] J. Lumsden, G. Pollock, M. Mahoney, in: K.V. Jata, M.W. Mahoney, R.S.
Mishra, S.L. Semiatin, T. Lienert (Eds.), Friction Stir Welding and Pro-
cessing II, TMS, Warrendale, PA, USA, 2003, p. 99.
[14] C.S. Paglia, L.M. Ungaro, B.C. Pitts, M.C. Carroll, A.P. Reynolds, R.G.
Buchheit, in: K.V. Jata, M.W. Mahoney, R.S. Mishra, S.L. Semiatin, T.
Lienert (Eds.), Friction Stir Welding and Processing II, TMS, Warrendale,
PA, USA, 2003, p. 65.
[15] M. Jariyaboon, A.J. Davenport, R. Ambat, B.J. Connolly, S.W. Williams,
D.A. Price, Corros. Sci. 49 (2007) 877.
[16] Z.Y. Ma, S.R. Sharma, R.S. Mishra, Scr. Mater. 54 (2006) 1623.
[17] K. Nakata, Y.G. Kim, H. Fujii,T. Tsumura, T. Komazaki, Mater. Sci. Eng., A,
437 (2006) 274.
[18] B. Johannes, R.S. Mishra, Mater. Sci. Eng., A, 464 (2007) 255.
[19] K. Elangovan, V. Balasubramanian, Mater. Sci. Eng., A 459 (2007) 7.
[20] F.J. Humphreys, M. Hatherly, Recrystallization and related annealing phe-
nomenon, Pergamon, 1995, p. 314.
[21] M.W. Mahoney, C.G. Rhodes, J.G. Flintoff, W.H.S. Bingel, Metall. Mater.
Trans. 29A (1998) 1955.
[22] A.A. Hassan, P.B. Prangnell, A.F. Norman, D.A. Price, S.W. Williams, Sci.
Technol. Weld. Join. 8 (2003) 257.
[23] K.V. Jata, S.L. Semiatin, Scr. Mater. 43 (2000) 743.
[24] Kh.A.A. Hassan, A.F. Norman, D.A. Price, P.B. Prangnell, Acta Mater. 51
(2003) 1923.
[25] M. Muruyama, Z. Horita, K. Hono, Acta Mater. 49 (2000) 21.
[26] K. Venkateswarlu, M. Chakraborty, B.S. Murty, Mater. Sci. Eng. A364
(2004) 75.
[27] Y.S. Sato, H. Kokawa, M. Enmoto, S. Jogan, Metall. Mater. Trans. 30A
(1999) 2429.
[28] G. Liu, L.E. Murr, C.S. Niou, J.C. McClure, F.R. Vega, Scr. Mater. 37
(1997) 355.
[29] Y.S. Sato, H. Kokawa, M. Enmoto, S. Jogan, T. Hashimoto, Metall. Mater.
Trans., A 30 (1999) 3125.
[30] C. Vargel, M. Jacques, M.P. Schmidt, Corrosion of Aluminium, Elsevier,
2004, p. 125.
4068 K. Surekha et al. / Surface & Coatings Technology 202 (2008) 4057–4068