J Mater Sci (2010) 45:4390–4400

DOI 10.1007/s10853-010-4429-0

EUROMAT 2009

The influence of salt fog exposure on the fatigue performance

of Alclad 6xxx aluminum alloys laser beam welded joints
A. T. Kermanidis • A. D. Zervaki •
G. N. Haidemenopoulos • Sp. G. Pantelakis

Received: 14 November 2009 / Accepted: 18 March 2010 / Published online: 3 April 2010
Ó Springer Science+Business Media, LLC 2010

Abstract Laser welding is increasingly used for the above leads to inferior fatigue performance, a fact that
fabrication of lightweight and cost-effective integral stiff- should be taken under consideration in the design and
ened panels in modern civil aircraft. As these structures age maintenance of these structures.
in service, the issue of the effect of corrosion on their
damage tolerance requires attention. In this work, labora-
tory data on the influence of salt fog corrosion on the Introduction
fatigue behavior of cladded 6156 T4 aluminum alloy laser
welded specimens are presented. The experimental inves- Laser welding is a new, promising technology in aluminum
tigation was performed on 6156 T4 laser butt welded airframes towards weight saving and cost efficient design.
sheets. Prior to fatigue testing the welded joints were Although laser welded parts are increasingly used in
exposed to laboratory salt fog corrosion exposure for modern aircraft, the technology is not considered yet for
720 h. The results showed that the clad layer offers suffi- highly loaded parts, which require superior damage toler-
cient corrosion protection both on base metal and the weld. ance characteristics. One of the important issues for con-
Fatigue testing was followed by standard metallographic sideration of laser welding in demanding damage tolerant
analysis in order to identify fatigue crack initiation sites. applications is the thorough understanding of the welded
Crack initiation is located in all welded samples near the joints corrosion behavior, as it is a serious threat to its
weld reinforcement which induces a significant stress fatigue resistance, especially in aging aircraft. This is
concentration. Localized corrosion attack of the clad layer, particularly important when one considers interactions of
in the form of pitting corrosion, creates an additional stress corrosion with: (i) weld defects, (ii) stress concentrations
concentration which accelerates crack initiation leading to imposed by geometrical discontinuities in the welded joint,
shorter fatigue life relative to the uncorroded samples. The (iii) tensile residual stress fields in the vicinity of the weld.
potency of small corrosion pits to act as stress concentra- The laboratory and field data on the issues stated above are
tion sites has been assessed analytically. The above results scarce, particularly for laser welded airframe alloys used in
indicate that despite the general corrosion protection civil aircraft. This work aims to contribute with valuable
offered by the clad layer, the localized attack described laboratory data to a better understanding on the influence of
corrosion on the fatigue performance of the laser welded
6156 T4 airframe alloy.
A. T. Kermanidis  A. D. Zervaki (&)  Corrosion behavior of welded joints can be influenced
G. N. Haidemenopoulos
by the dissimilar microstructures at the weld region
Department of Mechanical Engineering, University of Thessaly,
38334 Volos, Greece including the weld metal, heat affected zone, partially
e-mail: [email protected] melted zone, and base metal. The authors of the present
paper presented evidence [1] that dissimilar microstruc-
Sp. G. Pantelakis
tures can be the cause of localized corrosion in the case of
Laboratory of Technology and Strength of Materials (LTSM),
Department of Mechanical Engineering and Aeronautics, 2139 aluminum laser beam welds. Corrosion properties of
University of Patras, 26500 Rion, Patras, Greece laser hybrid welds in AA 6061 have been reported in [2].

123
J Mater Sci (2010) 45:4390–4400 4391

Corrosion attack is hindered by the use of cladding or investigated and supported with metallographic and frac-
anodizing, which can offer sufficient corrosion protection tographic examination. Fatigue crack initiation sites and
even in aggressive laboratory corrosive environments [3, fracture path were identified with assistance of fractogra-
4]. It is also well known that cladding of Alclad products phy. The role of stress concentrations arising from the weld
electrochemically protects the core at exposed edges and at reinforcement and localized corrosion of the clad layer was
abraded or corroded areas [5]. Scratch tests performed in evaluated. Finally, the elastic stress concentration factors
[6] confirmed that surrounding cladding sufficiently pro- and associated local stresses at the weld toe and around
tects the core material at the areas of scratches made on corrosion pits responsible for crack initiation have been
cladded aluminum specimens, prior to their exposure to salt calculated analytically and used to explain the obtained
spray fog. Furthermore, the works in [3, 6] provided evi- fatigue results.
dence that only locally cladded specimens have been suf-
ficiently protected against corrosion also at their areas
which did not provide a clad layer. This observation is Experimental
important for understanding the corrosion susceptibility of
the laser welded joints. Still, in long term operation the The butt-welded sheets were received from Gesellschaft
protective cladding can deteriorate due to cyclic mechan- zur Förderung der Kernenergie in Schiffbau und Schiffs-
ical loads, environmental attack, or combination of the technik (GKSS) research centre (Germany) with dimen-
former [7, 8]. Experimental investigations exist, which sions of 270 9 315 mm. The thickness of the sheets was
show evidence of the corrosion susceptibility of the clad 2.8 mm. The chemical composition of the base metal was
layer following exposure to laboratory corrosion [4], and (wt%) Si 1, Mg 0.9, Cu 0.9, Fe 0.1, Mn 0.55, Cr 0.125, and
clad layer failure due to cracking under applied fatigue Zn 0.4. Prior to welding the sheets were tempered to T4
loads [7, 9, 10]. A protective layer, damaged either by condition which includes solution heat treatment and nat-
corrosion or fatigue, can assist the further development of ural aging to substantially stable condition. A diode-
corrosion, or promote the degradation of the substrate pumped 3.3 kW Nd:YAG laser DY033 from Rofin-Sinar
fatigue properties, thus leading to the condition of an aged was employed for the welding experiments. Laser power
structure. Damage tolerance behavior of aged aircraft was adjusted at 3000 W, and the focal point position was
materials has been a subject of investigation by many on the top of the sheet. Spot size in focus was 0.4 mm. The
researchers and attention has been focused on the fatigue welding speed was 1.8 m/min. Filler wire 4047 (12.25% Si,
behavior of pre-corroded aluminum alloys [11–15]. On the 0.6% Fe, 0.05% Cu, 0.25% Mn, 0.05% Mg, 0.1 %Zn, 0.1%
other hand, experimental results on the fatigue behavior of Ti) of 1.2 mm diameter was used at a feed rate of 2.5 m/min.
aged, welded aluminum-joints are very limited. The Surface protection was achieved by supplying 25 l/min He,
authors of this study presented experimental results in [1] while Ar was used for root protection at 15 l/min.
indicating that prior corrosion exposure leads to fatigue Machining of the fatigue specimens was performed
resistance degradation in bare 2139 aluminum alloy laser according to the ASTM E 466 specification [16]. All
welded samples. The reason has been localized stress specimens were cut in the longitudinal (L) orientation
concentration at corrosion pits, which acts in synergy with relative to the rolling direction. Fatigue tests were per-
the stress concentration induced by the notch at the weld formed with a stress ratio R = 0.1 and a frequency of
toe and results in premature crack initiation and reduction 25 Hz on corroded and uncorroded welded specimens for
of fatigue life. The corrosion behavior observed was comparison.
strongly related to the temper condition affecting the cor- Prior to fatigue testing the specimens were exposed to
rosion susceptibility of the 2139 alloy. salt fog environment in a salt spray chamber according to
In the current investigation, the influence of existing the ASTM B117 standard [17]. A continuous corrosion
corrosion damage on the fatigue performance of cladded exposure in the salt fog for 720 h was applied in order to
6156 T4 aluminum laser welded specimens was examined. cause sufficient corrosion damage for the fatigue perfor-
First, the corrosion susceptibility of the welded specimens mance investigation. The sodium chloride (NaCl) concen-
was evaluated after exposure to salt fog environment and tration used for the accelerated corrosion exposure was 5%
the protective role of the clad layer was assessed. Next, the mass. Throughout the test it was ensured that inside the
influence of corrosion exposure on fatigue behavior of chamber the temperature range was maintained at
6156 aluminum welded joints was investigated. For this 35 ± 1 °C, while the pH was maintained between 6.5 and
purpose, evaluation of the fatigue results was performed in 7.2. After exposure, the specimens were cleaned according
terms of stress versus number of cycles (S–N) fatigue to ASTM G1 specification [18].
curves. The effect of localized corrosion damage at the clad Standard metallographic techniques (including grinding
layer on the fatigue life of corroded welded joints was and polishing) were applied for the characterization of

123
4392 J Mater Sci (2010) 45:4390–4400

microstructures in the base material, heat affected zone,

120
and weld metal. Etching was performed by using Keller’s
reagent. Microhardness measurements in all weld sections
were performed using a Vickers indenter and a load of 110

Microhardness, HV0.2
HAZ
0.2 kg.
100

BM
Results WM
90
PMZ
Materials 80

The microstructure of the clad 6156 alloy is shown in the

70
macrograph of Fig. 1. The structure consists of Al-rich 0 1 2 3 4 5 6 7
grains elongated towards the rolling direction, associated Distance from weld axis, mm
with rounded Mg2Si particles. Clad layer has an average
Fig. 2 Microhardness profile across the weld bead in 6156-T4 alloy
thickness of 60 lm.
A typical microhardness profile is shown in Fig. 2,
where the distinct weld zones are indicated, i.e., weld metal estimated at 5.22 and 11.95 lm, respectively. The pit depth
(WM), partially melted zone (PMZ), heat affected zone values are average values of several measurements per-
(HAZ), and the unaffected material (BM). A continuous formed on the metallographic specimens. Pitting corrosion
hardness drop is observed as approaching the WM from the has been also found at the base metal, near the interface
BM. This is associated with the dissolution and coarsening with the weld reinforcement (Fig. 3c). The role of the clad
of the strengthening precipitates in the HAZ [19], while layer observed in terms of corrosion behavior is similar to
within the PMZ the hardness drop is associated with dis- the operation of the local cladding pattern, first described in
solution of the strengthening phases and the formation of [3] as a potential alternative, cost efficient method to pro-
eutectic constituents [1]. vide corrosion protection in aluminum alloys.

Corrosion behavior Fatigue results

After exposure in the salt fog environment, metallographic Taking into consideration that machining of specimens
examination of the corroded welded samples was per- from the rolled welded sheet causes a relief of the weld
formed. Images obtained with the use of optical micros- residual stresses [20], the fatigue results in the following
copy revealed that even after the relatively long exposure section are discussed with neglecting any effects caused by
time of 720 h in NaCl fog, the weld metal was free of residual stresses. In addition, the welded samples were
corrosion attack (Fig. 3a). The clad layer acted as a sacri- tested in fatigue as received, without mechanical removal
ficial anode offering sufficient corrosion protection to the of the weld reinforcement in order to investigate on the role
weld metal. This is supported by evidence of small-scale, of stress concentrations induced by geometrical factors
pitting corrosion damage, found on the clad layer as shown such as the weld toe and root reinforcement on the fatigue
in Fig. 3b. The average and maximum pit depth has been behavior of welded samples.

Fig. 1 Typical microstructures of clad 6156 T4 a base metal b weld region

123
J Mater Sci (2010) 45:4390–4400 4393

Fig. 3 Micrographs of weld and base metal area after 720 h exposure to salt fog a corrosion free weld metal b corrosion pits at the clad layer, at
the base metal area, c corrosion pits at the interface base metal–weld reinforcement

Constant amplitude cyclic tests were performed with a

stress ratio of R = 0.1 and a frequency of 25 Hz with the
use of a servohydraulic 100 kN fatigue testing machine.
Two fatigue specimens have been tested at each stress
level. The fatigue results are displayed in the form of S–N
experimental data fitted to a 4-parameter Weibull distri-
bution. The experimental results show a harmful effect of
corrosion on the fatigue behavior of welded specimens.
The fatigue behavior given in Fig. 4 in terms of S–N curves
is degraded for the corroded joints, which exhibit shorter
fatigue lives compared to the uncorroded welded speci-
mens for all stress levels examined. The S–N Weibull fit-
ting curves lead to a fixed fatigue limit at 107 cycles of 100
and 80 MPa for the uncorroded and corroded welded
joints, respectively. It corresponds to a fatigue limit
reduction of 20% for the aged welded joints. The average Fig. 4 Fatigue results (S–N curves) of uncorroded and pre-corroded
% reduction in fatigue life for the corroded specimens is 6156 T4 welded joints
65% obtained at fatigue stress levels of 100–150 MPa.
In both uncorroded and corroded welded joints, fatigue concentration plays a principle role, leading to controlled
fracture occurred at the boundary of weld reinforcement fatigue crack initiation in this area. The considerable
with base metal as shown in Fig. 5. The weld toe stress shortened fatigue lives of the corroded specimens

123
4394 J Mater Sci (2010) 45:4390–4400

Fig. 5 Fractured specimens indicating fatigue fracture at the weld boundary regions for a corroded and b uncorroded welded joints

presented in Fig. 4 indicate that despite the induced stress Secondary microcracks initiating at the weld boundary
concentration at the weld toe, corrosion introduces an were also observed. As shown in Fig. 7, they propagate in a
additional effect which further degrades fatigue crack ini- depth exceeding the clad layer thickness and then are
tiation resistance. Herein, the role of pitting corrosion arrested. Final failure is dominated by the principle crack,
seems to be critical for the fatigue behavior of corroded which initiates at a different location of the weld boundary
welded joints. Evidence on this observation is provided in a and propagates until fracture. While in the uncorroded
following section of this paper. samples, secondary cracks formed only at the boundary of
To support the obtained experimental results and ratio- weld reinforcement, in the case of corroded samples sec-
nalize the effect of corrosion on the fatigue behavior, fur- ondary cracks were also observed far from the stress con-
ther evaluation of the main factors influencing the cyclic centration region (Fig. 8). The clad layer material is
behavior has been performed. The evaluation performed AA1300, with 80 MPa yield strength and 70 GPa modulus
and described in the next paragraphs consists of two parts. of elasticity. The modulus of the clad layer is similar to that
First, fractographic investigation combined with metal- of the base material 6156. Hence the fatigue stresses applied
lography on sections of the welded joint was performed in at the clad layer and base metal are equal, reaching and
order to identify the exact fatigue crack initiation sites and exceeding, in some cases, the tensile strength of the clad
propagation of fatigue damage. Next, the effects of stress layer. Also, it should be expected that after 720 h exposure
concentrations at the weld reinforcement and at corrosion to salt fog the small corrosion pits observed at the clad layer
pits have been estimated and their role on fatigue crack surface give the aluminum layer a more brittle behavior
initiation has been assessed. under tensile loads. The above conditions may explain the
preferential cracking observed in the corroded samples in
opposition to the uncorroded samples at remote surface
Fractographic investigation regions with regard to the weld toe boundary. However, the
observed cracking at these locations does not prove to be
From stereoscopic examination, the fatigue section, as well essential for fatigue behavior, since it’s effect is over-
as, the fast fracture area on specimen fractured surfaces whelmed by the effect of stress concentration at the weld toe.
were identified. In most cases fatigue section has the typ- Corrosion pits at the clad layer have an additional
ical semi-elliptical shape with the center being the fatigue influence on fatigue crack initiation occurring at the weld
initiation site (Figs. 6b, 10b). In order to determine the reinforcement boundary. Fatigue cracks in the corroded
crack initiation site, cross sectioning was performed specimens initiate at the bottom of corrosion pits at the
exactly at the center of the fatigue section area. Fracto- PMZ as depicted in Fig. 9. Then, as in the case of uncor-
graphic observations in both uncorroded and pre-corroded roded samples, they propagate perpendicular to the applied
samples revealed that cracks initiate at the weld toe or weld load until final fracture. Again, inspection of the fracture
root of the specimen, at the interface of the partially melted surfaces revealed the typical appearance of a fatigue sec-
zone (PMZ) with the weld metal, and propagate perpen- tion initiated at the weld toe, followed by a final fast
dicular to the stress axis (mode I) until final fracture fracture area, as shown in Fig. 10.
(Fig. 6a, c). Inspection of the fracture surfaces revealed the In the previous analysis, the experimental results reveal
typical appearance of a fatigue section initiated at the weld the harmful influence of corrosion on the fatigue perfor-
toe or weld root, followed by a final fast fracture area, as mance of welded joints. Taking under consideration, that:
shown in Fig. 6b. (i) fatigue crack initiation occurs always in the PMZ at the

123
J Mater Sci (2010) 45:4390–4400 4395

Fig. 6 a Location of fatigue crack initiation and propagation, b fatigue section followed by brittle fracture, c crack propagation at the PMZ–weld
metal interface

weld metal–base metal interface, (ii) fatigue crack initia- facilitates crack initiation at the boundary of weld toe and
tion is generated around corrosion pits in the corroded base metal. To calculate the local stress induced by the
samples, the observed fatigue behavior may be explained presence of the weld toe/root, the elastic stress concentra-
under the viewpoint of a combined, pronounced stress tion factor due to the notch effect introduced by the weld
concentration effect caused by the co-existence of weld toe toe has been calculated analytically. For the estimation of
and corrosion pits at the same location. In order to support the elastic stress concentration factor at the weld toe, the
the experimental findings, the local stress at the weld analytical solutions by Lehrke et al. [21, 22] and Ushi-
reinforcement boundary has been estimated and compared rokawa and Nakayama [23] have been implemented. For a
to the localized stress around the corrosion pit. For this butt weld in tension, Lehrke presented the following
purpose, the elastic stress concentration factors have been equation for the elastic stress concentration factor at the
calculated for the distinguished effects and are presented in weld toe:
the following section. 12m  m
4 2 tan h2

t
Kt1 ¼ ð1Þ
2h þ sinð2hÞ 4q
Estimation of local stress at crack initiation with W = t. This condition is in agreement with the weld
location joint configuration in the present study as shown in
Table 1.
Estimation of local stress at weld toe The parameters W, h, q, and t displayed in Fig. 11 are
given in Table 1 for the welded joint configuration under
The stress concentration created by the existence of weld investigation. For a butt weld under combined tension and
reinforcement is dominant in the fatigue behavior and bending, Ushirokawa and Nakayama [23] proposed the

123
4396 J Mater Sci (2010) 45:4390–4400

Fig. 7 Secondary cracks initiate at the weld root boundary

Fig. 8 Secondary cracking observed (i) far from weld root boundary, (ii) at weld root boundary

following equation for the calculation of the elastic stress lack of penetration has been considered hence the value
concentration factor at the weld toe: C(a/t) = 1 was taken.
By neglecting bending and lack of penetration then
Kt2 ¼ ½1 þ f ðhÞfgðqÞ  1gCða=tÞ ð2Þ gb ðqÞ ¼ 0 and at ¼ 1 In this manner:
with 
1
0:65
n qffiffiffiffio gðqÞ ¼ gt ðqÞ ¼ 1 þ bt ðh=qÞ ð4Þ
W
1  exp 0:90ðp  hÞ 2h 2:8ðW=tÞ  2
f ðhÞ ¼ qffiffiffiffio ; ð3Þ
with
n
W
1  exp 0:90ðp=2Þ 2h
W ¼ t þ 2h þ 0:6hp ð5Þ
In Eq. 2 gðqÞ ¼ at gt ðqÞ þ ab gb ðqÞ; and C(a/t) is the
parameter for lack of penetration. In the present analysis no The value of bt for the butt joint configuration is bt = 2 [23].

123
J Mater Sci (2010) 45:4390–4400 4397

The parameters W, h, h, q, and t displayed in Fig. 12 are With the use of Eqs. 1 and 2 and the parameters given in
given in Table 1 for the welded joint configuration under Table 1 the values of stress concentration factors have been
investigation. calculated Kt1 = 1.62 and Kt2 = 1.54, respectively. Both
equations result in similar elastic stress concentration
values at the weld toe of the butt joint.

Estimation of local stress around corrosion pit

In the corroded welded joints premature initiation of fati-

gue cracks leading to reduction of fatigue life was attrib-
uted to the presence of corrosion pits at the clad layer
resulting to a supplementary local stress concentration at
the bottom the pit adjacent to the weld reinforcement.
To estimate the value of the elastic stress concentration
factor at the bottom of a corrosion pit, the pit has been
simulated as a small surface semi-circular notch with
radius of curvature r. In the case of a surface, semi-circular
notch in tension the elastic stress concentration factor can
be derived with the use of the equation suggested by Cole
Fig. 9 Crack initiation at the bottom of corrosion pit at the PMZ et al. [24]

Fig. 10 a Location of fatigue crack initiation and propagation, b fatigue section followed by brittle fracture, c crack propagation at the PMZ

123
4398 J Mater Sci (2010) 45:4390–4400

Table 1 Calculated values of elastic stress concentration factors with Eqs. 1 and 2
Parameter Thickness Radius Angle h Angle h Parameter Reinforcement
t (mm) q (mm) (Lehrke) (Ushirokawa) m [16] width W (mm)

Lehrke 2.8 0.5 24° 0.2145 2.8

Kt1 1.62
Ushirokawa 2.8 0.5 156° 2.8
Kt2 1.54

corrosion pit can be estimated with superposition of the

solutions for the elastic stress intensity factors Kt and Ktn
θ ρ
Ktc ¼ Kt Ktn ð7Þ
P P
t w with
rlocalpit ¼ rmacro Ktc ð8Þ
By implementing Eq. 7 the elastic stress concentration
Fig. 11 Parameters in Eq. 1 for a butt weld in tension factor results in the value Ktc = 4.8. This value is signifi-
cant higher than the value calculated for the uncorroded
specimens of Kt = 1.6.

ρ h Fatigue notch sensitivity

w/2
The fatigue notch factor due to corrosion pit can be esti-
t mated by using the fatigue strength values of rKuncorr =
100 MPa and rKcorr = 80 MPa for uncorroded and cor-
roded joints, respectively. The fatigue strength values
θ result in a fatigue notch factor due to corrosion of a value
of Kf = rJuncor/rKcorr = 1.25. It shows that the effect of
Fig. 12 Parameters in Eq. 2 for a butt weld in tension and bending corrosion although important is much less severe than the
one predicted with the elastic stress concentration at the
corrosion pit of Ktc = 4.8. Using the values Kf = 1.25 and
r Ktn = 3.03 the fatigue notch sensitivity due to the presence
of corrosion pit at the clad layer can be estimated as
t P
P H Kf  1
q¼ ð9Þ
d Ktn  1
Using the above values the fatigue notch sensitivity
Fig. 13 Parameters used in Eq. 6 for the calculation of Ktn
results in q = 0.125. Considering that the full notch
effect is described with the condition q = 1 and the no
Ktn ¼ 3:065  8:871ðt=HÞ þ 14:036ðt=HÞ2  7:21ðt=HÞ3 notch effect with the condition q = 0, the small value of
0.125 obtained, indicates that the actual notch effect
ð6Þ
introduced by the presence of the pit at the clad layer is
The parameters t, H, d, and r used in Eq. 6 are displayed in important, but not as severe as predicted in the analysis
Fig. 13. In Eq. 6 the welded joint configuration parameters performed in ‘‘Estimation of local stress around corrosion
H = 2.8 and t = 20 lm have been used. The notch radius pit’’ section.
of curvature has been estimated according to the maximum
pit depth determined experimentally in ‘‘Results’’ section
(r = 11 lm). Discussion of results
With Eq. 6 the resulting elastic stress concentration
factor around the corrosion pit has been calculated The results presented in this study show that the clad layer
Ktn = 3.03. The local stress, rlocalpit, at the bottom of the may offer satisfactory protection of the weld metal even in

123
J Mater Sci (2010) 45:4390–4400 4399

the case of unprotected weld region. The cladding works as Conclusions

a sacrificial anode maintaining the weld metal free of
corrosion attack even after long laboratory corrosion In this work, the influence of salt fog corrosion on the
exposure. The experimental fatigue results demonstrate fatigue behavior of cladded 6156 T4 aluminum alloy laser
that despite the clad layer protection, small corrosion pits welded joints was investigated. The main conclusion of the
created during corrosion exposure at the clad surface, may work is that despite the general corrosion protection
act as fatigue notches leading to accelerated crack initiation offered by the clad layer, the localized attack of the clad
compared to the uncorroded samples. layer, in the form of pitting corrosion, leads to inferior
The stress concentration imposed by the geometrical fatigue performance.
discontinuity at the weld toe/weld root has a predominant In all cases, the weld reinforcement acts as a significant
effect on fatigue behavior and assists fatigue crack initi- stress concentration site for fatigue crack initiation.
ation in both uncorroded and corroded welded joints. This
is supported by fractographic evidence showing fatigue Acknowledgements This work has been supported by EU Wel-Air
program under contract AST3-CT-2003-502832. The help of
crack initiation at the weld reinforcement boundary for Mrs. Polina Taiganidou with metallography work is gratefully
both uncorroded and corroded specimens. In the case of acknowledged.
corroded welded joints, the local stress concentration
created at the weld boundary is enhanced at the bottom of
the pit due to the presence of the pit. The higher stress References
concentration produced at the weld boundary at the bot-
tom of the pit in the corroded joints is responsible for a 1. Kermanidis AT, Zervaki AD, Haidemenopoulos GN, Pantelakis
smaller crack initiation period and reduction of the fatigue SpG (2010) Mater Des 31:42
life. 2. Zhang Da-Q, Li J, Joo H, Lee K (2009) Corros Sci 51:1399
3. Petroyiannis PV, Pantelakis SpG, Haidemenopoulos GN (2005)
The local stress concentration has been calculated ana- Theor Appl Fract Mech 44:70
lytically. The results indicate a local high stress concen- 4. Velterop L (18–22 May 2003) Phosphoric sulphuric acid ano-
tration of 4.8 at the bottom of the pit. This value is dising: an alternative for chromic acid anodising in aerospace
considerably higher than the estimated stress concentration applications? NLR Report,NLR-TP-2003-210. National Aero-
space Laboratory NLR, Structures Technology Department,
of 1.63 due to the weld reinforcement alone. It explains the Aluminum Surface Science and Technology, in Bonn, Germany
accelerated crack initiation in the corroded joints leading to 5. Davis JR (ed) (1993) ASM speciality handbook: aluminum and
shortened fatigue lives compared to the uncorroded sam- aluminum alloys. ASM International, OH, USA
ples. The actual fatigue life degradation in the corroded 6. Simulation Based Corrosion Management (SICOM) EU funded
Project, FP6 Contract No. 030804, 2007–2010
samples although important, is not as severe as predicted 7. Karlashov AV, Tokarev VP, Batov AP (1965) Fiziko-chimi-
by the analysis, leading to a calculated fatigue notch sen- cheskaya Mekhanika Materialov 1(6):707
sitivity of 0.125. 8. Genkin JM (1996) PhD thesis, Department of Materials Science
It should be noted that in the present work the corro- and Engineering, Massachusetts Institute of Technology
9. Duquesnay DL, Underhill PR, Britt HJ (2005) Fatigue Fract Eng
sion damage influence on fatigue performance is exam- Mater Struct 28:381
ined from the viewpoint of localized, pitting corrosion 10. Merati A (2005) Int J Fatigue 27:33
effect. Hydrogen embrittlement has also been recognized 11. Bray GH et al (1997) In: Van Der Sluys WA, Piascik RS, Za-
as a corrosion degradation mechanism for aluminum wierucha R (eds) Effects of the environment on the initiation of
crack growth ASTM STP 1298. American Society for Testing
alloys [25, 26]. In the present paper, the authors have not and Materials, PA, USA, pp 89–103
considered the role of hydrogen embrittlement, since it is 12. Chubb JP, Morad TA, Hockenhull BS, Bristow JW (1995) Int J
a diffusion-controlled process, expected to occur and Fatigue 17(1):49
influence crack growth in a local scale. Fatigue life, 13. Kermanidis AT, Petroyiannis PV, Pantelakis SpG (2005) Theor
Appl Fract Mech 43:121
which is predominantly dependent on crack initiation 14. Wang QY, Kawagoishi N, Chen Q (2003) Scripta Mater 49:711
stage, is not expected to be significantly influenced by 15. Jones K, Hoeppner DW (2006) Corros Sci 48:3109
hydrogen embrittlement, but mostly by the notch effect 16. ASTM E466-82 (1994) Annual book of ASTM standards, met-
induced by weld reinforcement acting synergistically with als—test methods and analytical procedures, Section 3, vol
03.01, Metals-mechanical testing; elevated low-temperature tests;
small pits at the clad layer. Furthermore, it should be metallography. ASTM, PA, USA
reminded that in the present analysis the effects of 17. ASTM B117-94 (1995) Annual book of ASTM standards, met-
residual stresses at the weld region have not been taken als—test methods and analytical procedures, section 3, vol 03.02,
into consideration due to the machining process, which is wear and erosion; metal corrosion. ASTM, PA, USA
18. ASTM G1-90 (2000) Standard practice for preparing, cleaning,
expected to cause a substantial relief. Yet, even small and evaluating corrosion test specimens. American Society for
values of residual stresses may have an effect on the Testing and Materials, PA, USA
calculated fatigue lives. 19. Zervaki AD, Haidemenopoulos GN (2007) Weld J 86:211

123
4400 J Mater Sci (2010) 45:4390–4400

20. Masubuchi K (1980) Analysis of welded structures. Pergamon, 24. Cole AG, Brown AFC (1958) J R Aeronaut Soc 62:597
Elmsford, NY 25. Petroyiannis PV, Kermanidis AlTh, Kamoutsi E, Pantelakis SpG,
21. Brandt U, Lehrke H-P, Sonsino CM, Radaj D (1999) Anwendung des Bontozoglou V, Haidemenopoulos GN (2005) Fatigue Fract Eng
Kergrundkonzeptes für die schwingfeste Bemessung von Schweiß- Mater Struct 28:565
verbindungen aus Aluminumknetlegierungen. Fraunhofer-Institut 26. Pantelakis SpG, Haidemenopoulos GN (2002) In: Proceedings of
fuer Betriebsfestigkeit (LBF), Darmstadt (Final-Report) the 4th international conference on new challenges in mesome-
22. Lehrke H-P (1999) Berechnung von formzahlen fuer Schweiß- chanics, Aalborg University, Denmark
verbindungen Konstruktion 51(1/2):47
23. Ushirokawa O, Nakayama E (1983) Stress concentration factor at
welded joints. Ishikawaj ima-Harima Gihou (Technical Report)
23(4) (in Japanese)

123