Beruflich Dokumente
Kultur Dokumente
ALLOYS
DISSERTATION
Presented in Partial Fulfillment of the Requirements for
the Degree of Doctor of Philosophy in the Graduate School of
The Ohio State University
By
Xinyan Zhao, M.S.
*****
The Ohio State University
2006
Dissertation Committee:
Dr. Gerald S. Frankel, Adviser
Approved by
____________________________
Adviser
Graduate Program in
Materials Science and Engineering
ABSTRACT
Subsequent
quantitatively. Optical microscopy and analytical TEM were used to investigate the
effect of microstructure and local chemistry at grain boundary on EFC susceptibility.
Alloys with more elongated grain shape are more susceptible to EFC and a high Zn
content in the grain boundary precipitate free zone relates to a high susceptibility.
The effects of RH, temper and applied stress on EFC kinetics of AA7178 were
investigated by ESH tests. The critical RH for EFC propagation in AA7178 was found
to be about 56% and the EFC kinetics increased with RH. The effect of temper in ESH
test was consistent with published results from EXCO tests, but they provide a
quantitative description of the temper effect. The effects of applied compressive and
tensile stresses on EFC kinetics were studied using a four-point bending jig.
Compression accelerated EFC significantly and tension reduced kinetics. An equation
describing the effects of RH, stress, and time on EFC kinetics was developed based on
the ESH results using Eyring model.
In situ X-ray radiography was used to characterize intergranular and exfoliation
corrosion in high strength Al alloys. The samples were either exposed to sodium
chloride solution (NaCl) at a controlled potential or to high humidity after an
electrochemical pretreatment in NaCl solution. Intergranular corrosion (IGC) growth
rates in solution for samples encased in epoxy and exposed to solution were distributed
over a range with an upper limit equal to the rates determined by the foil penetration
technique. Unencased AA2024-T3 samples that were held at a controlled potential in
solution exhibited IGC and then exfoliation corrosion. AA2024 and AA7178 samples
encased in epoxy developed sharp intergranular fissures during exposure to 96% RH
following an electrochemical pretreatment.
iii
In situ X-ray
iv
ACKNOWLEDGMENTS
I would like to express my sincere thanks to my adviser, Dr. Gerald Frankel, for
his guidance, inspiring motivation, and helps during the supervision of this project. I
have been fortunate to study under his tutelage. Not only his knowledge, but also his
passion towards aircraft corrosion has been inspired me a lot through 5 years of
graduate study.
Next, I would like to thank Dr. Rudy Buchheit, for his participation on my
academic advisory committee and for comments and suggests on my research. I am
also grateful for a lot of good comments and suggestions given by Dr. Michael Mills
and Dr. Stanislav Rokhlin, both on my advisory committee.
I would like to
acknowledge funding from the Aging Aircraft Division of ASC in support of the
Aeronautical Enterprise Structures Strategy with a contract through S&K Technologies.
My special thanks go to Dr. Bahman Zoofan from Welding department for his
kind help and corporation with the use of X-ray radiography facility. Dr. Bahman
Zoofan has been spending many hours working with me and contributes a great amount
of work to this effort. I also want to thank Mr. Andrew Bonifas for his input to this
project.
He helped and designed the four point bending apparatus for studying
mechanical effect on exfoliation corrosion. I also wish to thank Mr. Jim Suzel at S&K
vi
Technology for providing the AA7178 samples and Dr. William Abbott of Battelle for
performing the environmental exposure testings for these samples.
I would like to thank all members of Fontana Corrosion Center: Dr. Weilong
Zhang, Dr. Wenping Zhang, Dr. Qingjiang Meng, Dr. Xiaodong Liu, Dr. Tsai-Shang
Huang, Ms. Hong Guan, and many current FCC members. I would also like to thank
Mr. Henk Colijn, Mr. Cameron Begg and Dr. Lisa Hommel, who helped and trained me
how to use TEM, FIB, SEM and XPS. I would like to thank Ms. Dena Bruedigam, Ms.
Chris Putnam for their office help. I would like to acknowledge Mr. Steve Bright, Mr.
Ken Kushner for their helps in preparing my experimental samples. I also wish to
thank Dr. Suliman Dregia, Mr. Mark Cooper, Ms. Mei Wang, and Ms. Wendy Ranney
for their kind help with my academic problems.
Finally, I would like to thank my parents for their unwavering support and
encouragement throughout my educational years.
vii
CURRICULUM VITA
July 1999B.S. Materials Physics,
University of Science and Technology
Beijing, China.
March 2003M.S. Materials Science and Engineering,
The Ohio State University.
2000-presentGraduate Research Associate,
Materials Science and Engineering,
The Ohio State University.
PUBLICATIONS
1.
2.
3.
4.
5.
6.
FIELDS OF STUDY
Major Field: Materials Science and Engineering
ix
TABLE OF CONTENTS
Page
ABSTRACT...................................................................................................................II
ACKNOWLEDGMENTS ............................................................................................ VI
VITA .......................................................................................................................VIII
LIST OF TABLES .....................................................................................................XIII
LIST OF FIGURES ....................................................................................................XIV
CHAPTERS:
1.
INTRODUCTION ..................................................................................................1
2.
LITERATURE REVIEW........................................................................................3
2.1 BASIC FORMS OF CORROSION OF ALUMINUM ALLOYS ........................................4
2.2 PITTING .............................................................................................................5
2.3 EXFOLIATION CORROSION ..................................................................................6
2.3.1 The Mechanism of Exfoliation Corrosion .................................................6
2.3.2 Factors That Influence EFC ....................................................................11
2.3.2.1 Effects of Alloying Elements on EFC..........................................11
2.3.2.2 Effects of The Grain Shape on Exfoliation Corrosion..................12
2.3.2.3 Effect of Heat Treatment on Susceptibility to Exfoliation
Corrosion ....................................................................................16
2.3.3 Environmental Aspects...........................................................................18
2.3.4 Exfoliation Corrosion Testing.................................................................19
2.3.4.1 Immersion Test ...........................................................................20
2.3.4.2 Salt Spray Tests ..........................................................................22
2.3.4.3 Electrochemical Impedance Spectroscopy...................................24
2.3.4.4 Deflection Technique ..................................................................25
2.4 INTERGRANULAR CORROSION ..........................................................................27
2.4.1 IGC Mechanisms....................................................................................27
2.4.1.1 Galvanic Couple Theory .............................................................27
2.4.1.2 Precipitate Free Zone Breakdown Model.....................................29
2.4.1.3 Anodic Dissolution of Grain Boundary Precipitates ....................32
x
4.
6.
BIBLIOGRAPHY .......................................................................................................182
xii
LIST OF TABLES
Table
Page
xiii
LIST OF FIGURES
Figure
Page
2.1
2.2
2.3
2.4
(a) Dimensions of a thin walled hemispherical shell used to calculate the excess
pressure in a blister. (b) Section through a blister diameter.[9] (Reprinted from
Corrosion Science, Vol. 22, M.J. Robinson, Mathematical Modelling of
Exfoliation Corrosion in High Strength Aluminum Alloys, Pages: 775-790,
Copyright (1982), with permission from Elsevier.).............................................63
2.5
xiv
2.6
(a) Corrosion product forces generated on T/2 specimens of 8090 plates. (b)
Corrosion product forces generated on T/4 specimens of 8090 plate.[8]
(Reprinted from Corrosion Science, Vol. 49, D.J. Kelly and M.J. Robinson,
Influence of Heat Treatment and Grain Shape on Exfoliation Corrosion of AlLi Alloy 8090, Pages: 787-795, Copyright (1993), with permission from
Elsevier.)............................................................................................................65
2.7
2.8
29
Potentiodynamic polarization scans for bulk Al2 CuMg in 0.5M NaCl solution
open to air, actively aerated, and actively deaerated by sparging with N2.[75]
(Reproduced by permission of The Electrochemical Society, Inc.) .....................67
2.10
2.11
Anodic part of cyclic polarization curve for Al3 Fe in deareated 0.1N NaOH
solution. The curve corresponds to the first cycle starting with an as polished
surface at 1.5V. In this particular run, the sweep was reversed at 0V.
Continuous curve: forward sweep. Dashed curve: reverse sweep. [91]
(Reproduced by permission of The Electrochemical Society, Inc.) .....................68
2.12
3.1
3.2
3.3
3.4
3.5
3.6
3.7
Metallographic sections of good plate and bad plate. Also given is the
terminology used for the different sections. (a) bad plate. (b) good plate......92
3.8
Metallographic sections of AA7178 wingskin plate. (a) Bad plate. (b) Good
plate. The sections are through-thickness montages, starting at the right of the
top image in each pair and then wrapping around to end at the left side of the
bottom image in each pair. .................................................................................93
3.9
Grain size distribution and grain aspect ratios through the thickness of (a)
good plate and (b) bad plate.........................................................................94
3.10
3.13
3.14
Nano-EDS line profile of solute depleted zone around grain boundary. The data
are reported as ratio of the x-ray intensities for each element relative to that of
Al. (a) Near up side of the good plate (shown in figure 3.12a). (b) Near down
side of the good plate (shown in figure 3.12b)...............................................101
4.1
4.2
Current response of two AA7178 samples during potentiostatic pretreatment at 710 mV SCE in 1 M NaCl. Both samples had the same exposed area. .............123
4.3
4.4
4.5
4.6
4.7
4.8
SEM micrograph of FIB cross section of IGC attack in the same AA7178
sample as in figure 4.7......................................................................................126
4.9
4.10
AA7178 wingskin slices after pretreatment and then 13 days in 30% constant
humidity. The rolling direction is vertical. .......................................................128
4.11
4.12
4.13
4.14
4.15
4.16
4.17
4.18
4.19
4.20
4.21
4.22
5.1
5.2
5.3
5.4
5.5
5.6
Plot of the depth of various specific sites for L oriented AA2024-T3 sample as a
function of time. Also shown is the curve for L oriented foil penetration
samples. ...........................................................................................................156
5.7
5.8
5.9
5.10
Plot of the depth of various specific sites for L oriented AA7178 wingskin
sample as a function of time. ............................................................................160
5.11
5.12
5.13
5.14
Samples were put in 96%RH after above electrochemical treatment. Sharp IGC
continued along L direction at high rate. (a) 0 day in 96%RH (after 7 hours in 1
M NaCl, -710mV SCE). (b) 6 days in 96%RH. (c) 30 days in 96%RH. ............164
xviii
5.15
Samples were put in 96%RH after above electrochemical treatment. Sharp IGC
continued along T direction at high rate. (a) 0 day in 96%RH (after 7 hours in 1
M NaCl, -710mV SCE). (b) 6 days in 96%RH. (c) 48 days in 96%RH. ............165
A. 1
A. 2
A. 3
A. 4
2D contour plot showing stress field around intergranular crack tip simulated by
ABAQUS.........................................................................................................181
xix
CHAPTER 1
INTRODUCTION
between these forms of localized corrosion, a good understanding of SCC and IGC will
help us to investigate the mechanism of EFC and develop a good methodology to
quantify the corrosion process.
Chapters 3 through 5 cover the details of the technical findings of this
dissertation. They are written as stand alone papers, and will be submitted individually
for publication. Chapter 3 describes the development of a novel approach for studying
EFC kinetics. Electrochemical methods such as cyclic polarization, galvanostatic and
potentiostatic polarization together with materials characterization techniques such as
optical metallography, SEM/EDS, TEM, FIB, and STEM/Nano-EDS were utilized in this
1
study. The role of grain aspect ratio and the distribution of Zn at the grain boundary on
EFC susceptibility were studied and discussed.
Chapter 4 describes the effect of relative humidity, mechanical stress and heat
treatment on EFC kinetics. The ESH technique developed in chapter 3 was utilized for
the study of these factors. A four point bending jig was designed to investigate the effect
of mechanical stress on EFC. An empirical model was then developed utilizing an
Erying relationship.
Chapter 5 covers a study of intergranular corrosion and EFC as well as the
transition from IGC to EFC. In situ X-ray radiography was utilized to study intergranular
corrosion over a wide range of corrosion kinetics and visualize the process of exfoliation
corrosion in immersed condition and humidity exposure. The conditions required for
IGC and EFC to occur were discussed in this chapter.
Chapter 6 summarizes the findings of this work and states conclusions on the
kinetics of exfoliation corrosion. This chapter also covers suggestions for future work
and some unresolved issues.
A finite element simulation of EFC process utilizing commercial software
ABAQUS is covered in Appendix A. The simulation is based on a phenomenological
model of exfoliation corrosion in humidity.
CHAPTER 2
LITERATURE REVIEW
INTRODUCTION
High strength aluminum alloys such as AA2xxx and AA7xxx are extensively
used for aircraft structures. Localized corrosion is a major cause of their failure. Among
various types of localized attack that aluminum alloys undergo, exfoliation corrosion
(EFC) is an important form. EFC is thought to be a particular form of intergranular
corrosion (IGC) that can occur on the surface of wrought aluminum alloys with an
elongated grain structure.[1] During this corrosion process, internal stresses caused by
voluminous corrosion products force layers of uncorroded metal away from the body of
the material, giving rise to a layered appearance.[2] An example of EFC is shown in
figure 2.1. EFC can lead to metal failure by gradual disintegration and the presence of
exfoliation corrosion reduces the toughness of the alloy and enhances the fatigue crack
growth rate.[3-7] All of these will cause serious problems to aircraft and limit their
3
lifetime. EFC has been studied since the 1950s, yet crucial phenomena remain unclear.
Several papers suggested that the exfoliation corrosion results either from intergranular
corrosion or alternatively from a stress corrosion mechanism.[1, 8-10] Experiments
conducted by Robinsons group have shown a strong similarity between the mechanism
of stress corrosion cracking (SCC) and EFC mechanism.[1, 8-10] However, exactly how
this mechanism works in the case of EFC is still unknown.
develop by the formation of corrosion product play a crucial role in the kinetics of EFC
and make exfoliation corrosion more difficult to understand. Though many efforts have
been made to enhance the resistance of aluminum alloys to EFC, the lack of a
quantification method and limited laboratory tests for EFC have slowed down the
development of protection methods for EFC. The purpose of this chapter is to provide an
overview of previous efforts and accomplishments towards the understanding of the EFC
of aluminum alloys including the critical factors that influence exfoliation corrosion and
the relationship between EFC, IGC and SCC.
2.1
two groups. The kinetics of intergranular corrosion (IGC), stress corrosion cracking
(SCC) and exfoliation corrosion (EFC) are more dependent on the metallurgical structure
of the alloys and are so-called structure dependent corrosion.[2] They are major problems
in aircraft structures. The corrosion rate of this type of corrosion is strongly influenced
by grain shape.[1, 11, 12] Uniform corrosion and other forms of localized corrosion,
mainly pitting corrosion, are more dependent on local chemistry of the environment and
are included in the second group.
2.2
PITTING
Pitting is an important form of localized corrosion and it has been extensively
studied. Pitting can lead to IGC, EFC and SCC under certain circumstances, so it is
important to understand pitting corrosion, its initiation and propagation. According to
Frankel, pitting corrosion is defined as localized accelerated dissolution of metal that
occurs as a result of a breakdown of the otherwise protective passive film on the metal
surface.[13] The presence of specific anions such as Cl- or Br- is required for the
initiation of pitting corrosion.[2] Furthermore, the distribution of pitting of Al alloys is
influenced by the distribution of cathodic metallic constituents in the metal surface.
Pitting can develop into IGC, which will initiate EFC and SCC when interacting with
stress, eventually leading to catastrophic failure of the structure. A characteristic feature
of pitting corrosion is the pitting potential. The pitting potential is the potential above
which pits will initiate and below which they will not. The pitting potential is sometimes
referred to breakdown potential related to the localized breakdown of oxide film.[14]
Since each intermetallic phase has a unique breakdown potential in given electrolyte, the
potential measurement can be used to characterize the specific phases in alloys. A phase
is unattacked at potentials below its breakdown potential.[15]
2.3
EXFOLIATION CORROSION
2.3.1
Exfoliation corrosion is often observed in 2xxx (Al-Cu-Mg) and 7xxx (Al-Zn-CuMg) series high strength alloys.[16] This type of corrosion is characterized by lamellar
surface attack of alloys containing a highly directional grain structure. Robinson has
pointed out that exfoliation corrosion is generally intergranular in nature due to the
galvanic interaction between grain boundary precipitates and the adjacent matrix.[1] If
the precipitates are active to the matrix, they may be preferentially dissolved. If they are
noble to the matrix, then attack at the solute depleted zones around the precipitates is
probable. Intergranular corrosion can proceed along multiple intergranular paths parallel
to the surface. The hydrated aluminum oxide corrosion product has a higher volume than
the alloy matrix that converts into the corrosion product. Consequently, relative large
wedging stresses are produced at the IGC front and lift the surface grains, promoting
further attack. The corrosion product of aluminum alloys is assumed to be hydrated
alumina.[17, 18] During localized corrosion in chloride environments, hydrolysis of
aluminum ions can be described as following reaction:
AlCl3+3H2O=Al(OH)3+3HCl
The aluminum hydroxide gel is not stable and crystallizes with time to convert into the
rhombohedral monohydrate (Al2O3H2O or boehmite), the monoclinic trihydrate
(Al2O33H2O or bayerite) and another monoclinic trihydrate(hydrargilite).[19]
Generally speaking, three factors which are required for exfoliation to occur[10, 20]:
1. Highly directional microstructure.
They proposed that the stress corrosion cracking mechanism is probably the
predominant factor that determines the kinetics of exfoliation corrosion.[9] The role of
wedging forces on exfoliation corrosion of aluminum alloys was studied by using a
special jig with a load cell that transmitted the force from the propagating exfoliation
blister.[23] The wedging force generated by insoluble corrosion products in aluminum
alloy L95 (W)(Zn 5.8, Mg 2.5, Cu 1.6, Cr 0.15) developed during immersion in EXCO
solution was measured. It was found that there is an incubation period, during which the
corrodant penetrated the surface layer of wrought aluminum grains.[23] After the
incubation time, the force increased linearly until it attained a maximum value of 600-800
Newtons in EXCO when observations showed that the entire area of the surface had
suffered exfoliation corrosion. This process is shown in figure 2.2. The period before
the maximum value was reached was considered to be due to the build-up of compressive
stresses overcoming the propagation of exfoliation. The corrosion product pressure was
7
also calculated to be 2MN m-2 when the surface of the specimen was affected by
exfoliation. The highest wedging pressure within the corrosion product is 4.8MNm-2.
The stress intensity at the developing corrosion front in exfoliation corrosion was then
calculated by crack opening displacement measurement (2.9MN m-3/2) and it had a
comparable value with the KISCC (<=2.6MNm-3/2) for alloy L95 in EXCO derived from
double cantilever beam specimen measurements, indicating that SCC is a possible
mechanism.
Kelly also measured the corrosion product force of Al-Li alloy AA8090 using the
same experimental arrangement as in Robinsons experiment.[10, 23] The magnitude of
maximum corrosion product force generated in AA8090 with different aging time was
ranked in the following order: under aged > over aged > peak aged condition for the plate
and sheet material.
hardened Al alloys was dependent on the magnitude of the force generated by corrosion
product. Since the corrosion product force resulted in a net tensile stress at the crack tip
during SCC, it is then a question whether the corrosion product force can generate the
stress that is necessary to propagate an SCC crack. To answer this question, Kelly
compared the maximum corrosion product force with lowest KISCC of AA8090 at
different aging times. The comparable results supported the proposition that exfoliation
proceeded by a stress-assisted corrosion mechanism.[10] In conventional alloys, overaging increases the resistance to EFC. The increased susceptibility to IGC in over-aged
AA8090 could be attributed to the enhanced galvanic action promoted by the formation
of Al2CuMg and Al2 CuLi.
More recently, the corrosion product force was measured in 7xxx alloys (AA7150T651, AA7055-T7751, AA7449-T79) by McNaughtan et al.[24] The magnitude of
maximum force generated in EXCO solution is in the order of: AA7150-T651 (25-50%
recrystallized) > AA7150-T651 (80% recrystallized)> AA7449-T79> AA7055-T7751
indicating their decreasing susceptibility. The increase of corrosion product wedging in
AA7150-T651 (with 25-50% recrystallized) is faster than in AA7150-T651 (with 80%
recrystalized) indicating a relationship between the grain aspect ratio and wedging stress.
It was also found that first drying out and then re-wetting in EXCO solution increased the
corrosion product force. McNaughtan et al. explained that this is because the corrosion
product was a hydrated alumina gel.[17] On drying out, the corrosion product changed to
crystalline Al2O33H2O with higher mechanical strength and this gave rise to a further
increase in corrosion product force.
corrosion cracking and exfoliation corrosion, the stress corrosion resistance was
determined by measuring KISCC values using DCB specimens.
An inverse linear
relationship between the corrosion product forces and the KISCC values for stress
corrosion cracking indicates their common corrosion mechanism and low KISCC
corresponds to high susceptibility to SCC. This study also shows the strong dependence
of corrosion product force on the IGC rate as well as the susceptibility of the grain
boundaries to IGC, which is controlled by the composition and distribution of grain
boundary precipitates.
Considerable effort has been spent on proving that exfoliation propagates by an
SCC mechanism. The fundamental question as to what electrochemical interactions are
responsible for the generation of corrosion products has also been explored.
9
Some
investigators have suggested that the cause of exfoliation in 7xxx aluminum alloys is an
electrochemical potential developed between primary or secondary particles and the
surrounding AlZnMg solid solution.[20, 25, 26] Mattsson et al. studied the mechanism of
EFC in AlZn5Mg1 by investigating the phases occurring in the alloy in the naturally aged
and artificially aged conditions.[20] In naturally aged condition, GP zones are the main
phases. EFC is caused by galvanic corrosion cell consisting of cathodic -Al(Fe,Me)Si
particles along the extrusion or rolling direction and narrow zones of anodic Zn-, Mg rich
matrix. This narrow zone of Zn and Mg rich matrix will be anodically attacked leading
to EFC along boundary. The MgZn2 phase is much less noble and subject to dissolution,
and the matrix is low in Zn and Mg so that it is cathodically protected. Since MgZn2 is
evenly distributed in the artificially aged condition, the attack is general and the alloy is
not prone to EFC. Polarization of these phases showed that artificially aged material is
more noble than naturally aged materials, further confirming his model.
In summary, the exfoliation process is a synergic combination of chemical and
mechanical actions. Papers containing mechanistic studies on exfoliation corrosion show
that electrochemical mechanisms underlie intergranular corrosion and the wedging stress
generated from intergranular corrosion decides the exfoliation characteristics.
The
10
2.3.2
Al(Fe,Me)Si and hence increase the susceptibility to EFC. They suggested that Si and Cr
might also have a similar effect. An influence of chromium has been reported elsewhere,
the susceptibility being higher with increased chromium content.[30] In AlMgSi (6000)
series alloys, Mg, Fe and Cu at concentration levels about 0.28-0.29, 0.3-0.32 and 0.050.22, respectively, develop high susceptibility to EFC, while alloys with low level of
these elements did not show a tendency to EFC.[5] Robinson also mentioned that the
difference in copper/magnesium ratio between AA2024 and AA2014 results in the
formation of different grain boundary phases.[8] CuAl2 is predominant in alloy AA2014
and CuMgAl2 is predominant in AA2024. The preferential dissolution of CuMgAl2 in
AA2024 leads to fast rates of IGC and EFC.
It was assumed that Mn, Fe and Zr diffuse slowly and can maintain stable
concentration so that they play a primary role in creating exfoliation paths, while Zn and
Mg play a secondary role because they diffuse readily at normal homogenizing
temperatures.[20, 25, 26, 31, 32] Evans and Jeffrey found that only those alloys
11
containing both Fe and Mn in AA7004 develop a high sensitivity to EFC and the
corrosion path is at grain boundaries where there is depleted in Mn.[31] They proposed
that the Mn-rich and Mn-depleted regions form an electrochemical cell and these regions
are elongated during fabrication, resulting in exfoliation path. This is in agreement with
the report by Bassi et al. that Mn content was a primary factor in EFC.[25] They further
stated that Mn could replace Fe in compounds such as -AlFeSi so that the formation of
-AlFeSi will result in depletion of Mn in surrounding areas.
Posada et al. have stated that, in AA2024 aluminum body skins on aging aircraft,
structural differences might control the extent of exfoliation corrosion by control of
boundary precipitation and elemental depletion.[33] They carried out an analysis of
precipitation within the grains and grain boundaries and elemental depletion profiles
along grain boundaries. Mg was somewhat uniformly distributed and no Mg containing
precipitates were observed even though the previous studies showed that Mg and Zn
depletion zones were involved in the attack of 7xxx alloys.[34] They also did not observe
the systematic depletion or enrichment of Cu, Si, Mn, and Fe in grain boundary zones.
material, with the most advanced attack occurring on the mid-section of the plate
containing more grains with higher grain aspect ratio. The severity ranked in the order:
T/2 > T/4 >surface. More elongated grains result in larger surface blisters before fracture.
The depth of penetration was also in the same rank order, which contradicts their
previous finding that the depth of penetration is unrelated to the grain shape. It is
possible that AA2024 and AA2014 could have different grain boundary chemistry, which
has some effect on their susceptibility to IGC. The surface ratings of different sections
from the two materials are shown in figure 2.5. The influence of grain shape on the depth
of exfoliation corrosion was measured by both a four-point bend method[4] and optical
metallography.[37] The result is in agreement with the observations of the surface ratings.
AA2024-T351 with higher grain aspect ratio exfoliated faster than AA2014-T651.
Constant load stress tests were performed in EXCO solution for the same material to
investigate the SCC behavior. It was shown that the cracks propagated most rapidly on
the T/2 section where the grains were most elongated. The authors concluded that the
grain structure has a similar effect in both exfoliation and SCC, providing further
evidence for a link between the two forms of corrosion.
The same results were also obtained by corrosion tests on lithium aluminum alloys
AA8090.[37] The susceptibility of AA8090 to EFC was ranked as T/2 > T/4 >
Surface.[37] EXCO and MASTMAASIS tests were conducted on the AA8090 plates
with different degrees of recrystallization.
exfoliation ranked as fully partially unrecrystallized sheet > recrystallized sheet >
recrystallized sheet, which means the susceptibility to EFC decreases in this sequence.
The microstructure of the plate affects the corrosion product force. More elongated
14
grains introduced a longer and more tortuous corrosion path, slow development of the
corrosion product force and higher maximum value of corrosion product force. In a
rolled or extruded product, variations in exfoliation susceptibilities and attack rates can
be often attributed to differences in thickness of the recrystallized layer usually present
on the outside surface.[38]
Exfoliation corrosion of aluminum alloys HE10, (Al-Cu-Cd) and HE15 fully aged
materials were also extensively studied by Liddiard et al.[4] Extrusions were tested, with
recrystallized surface layers and elongated grains below the surface.
Atmospheric
exposure tests for 8 years and salt spray tests in the lab were performed. The specimen
was first corroded in a salt spray cabinet and then loaded in a deflection jig. The
approximate average material loss, or the change in the effective cross section was
obtained by measuring the deflection of the specimen under load. Three distinctive
periods for the loss of metal were observed. The first stage, the incubation period, was
explained by intergranular penetration of the recrystallized surface. The second period
involved rapid metal loss and corresponded to the attack of the elongated grains beneath
the recrystallized layer, with lifting of the metal. The third period showed an even faster
attack, explained by the authors as a consequence of the highly directional longer grains
found at the center of the extrusion.
An elongated grain structure is an important prerequisite for exfoliation.
Intergranular corrosion does occur in materials with equiaxed grains, but this tends to
result in pitting rather than the formation of exfoliation.
15
At slightly longer aging times, a precipitate free zone was seen, which
appeared sensitive to attack. Peak aging would seem to give rise to the copper containing
phases in the matrix as well as at the grain boundaries, so attack at this temper was seen
to consume matrix material as well as that at the grain boundaries. The underaged
material was most susceptible to intergraunular corrosion, because of the continuous
anodic zone formed.
16
Liddiard et al. also studied aging effect on the exfoliation corrosion of HE10 and
HE15.[4] Aging was observed to confer some resistance to exfoliation corrosion, because
of increased subgrain boundary precipitation for longer aging times, allowing less attack
at the grain boundaries and relieving the corrosive process there. The susceptibility can
be reduced by increasing aging time at 190C from 1h to 24h, though overaging can result
in a loss in mechanical properties. It was also proposed by Lifka that over-aging tends to
reduce an alloys susceptibility to EFC by virtue of larger precipitates giving a more
broken anodic path, and reducing the potential difference between the grain boundaries
and the matrix.[29]
In Li-bearing alloys, the relationship between heat treatment and susceptibility to
exfoliation is not as straightforward as in the case of AA7xxx series. Work done on AlLi alloys indicates that, in alloys containing Cu and Mg, the susceptibility to exfoliation
increases with over-aging.
exfoliation increases with increased aging time. They attributed this to the increased
precipitation of T1 (CuAl2Li) at the grain boundary.[42] The increased precipitation
enhances the galvanic action between the matrix and the active T1 (CuAl2Li) phase.[10]
This is supported by studies on bulk T1 phase indicating that it is active with respect to
the matrix.[43, 44]
Recent studies on AA8090 alloys by Kelly showed that corrosion product forces were
ranked in the order: under aged > over aged > peak aged, which in turn determine the
exfoliation susceptibility as shown in figure 2.6.[10] The authors explained that the high
susceptibility of over aged temper in AA8090 was attributed to the formation of anodic
17
phases (Al-Li, S phase, and T1 phase Al2 CuLi) that precipitated at grain boundaries
during the over aging.[45]
Several combinations of tempers were shown to increase the resistance to exfoliation
corrosion. Lifka et al. conducted corrosion tests in a laboratory accelerated environment
(acidified 5% NaCl intermittent spray at 49C) as well as seacoast and inland industrial
atmospheres on high strength AA2xxx and AA7xxx aluminum alloys with different
temper combinations.[29] They found several compositions and tempers that exhibit
virtual immunity to EFC and SCC: AA2024-T851, AA2219-T851 and AA7075-T7351.
AA2020-T7351, AA2024-T851, AA2219-T851, AA7001-T7551, AA7075-T7351 and
AA7178-T7651 were free of exfoliation in laboratory tests. For T4 and T3 tempers,
decreasing the plate thickness increased the degree of susceptibility. The 3.5% NaCl
alternate immersion test on these materials showed that the AA7075-T651 and AA7079T651 were susceptible to SCC. Similar results were obtained from seacoast atmosphere
exposure tests at Point Judith, Rhode Island. AA7xxx series alloys in T76 or T73
tempers exhibited relative resistance to EFC. T6 7xxx series were susceptible to EFC.
AA2024-T3 plates exhibited severe exfoliation, whereas less exfoliation corrosion was
observed for the artificially aged T8 condition.[46]
2.3.3
Environmental Aspects
by Carter and Campbell that the weather conditions during the initial period of exposure,
particularly as they affect the time of wetness, would have a critical influence on the
pattern of corrosion that develops on aluminum alloys.[49]
2.3.4
exfoliation corrosion. The common exfoliation corrosion tests in laboratory that have
been reported are: immersion test method, salt spray method, electrochemical corrosion
test and mechanical tests. The first two methods usually involve in the interaction
between alloys and acidic solution to determine the susceptibility of certain alloy. The
resulting surface is often compared to the environmental exposure data.
19
Exfoliation
performances of AA7075 and AA7178 alloys in atmospheric exposure and laboratoryaccelerated tests were compared.
21
tartrate and hydrogen peroxide was devised for the ASSET test. This test was found to
give a good correlation with natural exfoliation.
specimens using a salt spray cabinet. The cabinets enable a cycle of wetting and drying
to take place. These tests have been deemed successful in that the corrosion product does
not form away from the site of corrosion attack as with some constant immersion tests.
The corrosion product may therefore cause surface lifting of grains and blister formation.
The SWAAT test can distinguish between alloys not susceptible to exfoliation and those
having a degree of susceptibility under specialized conditions of exposure, and
susceptible alloys under normal exposure. A period of one-week duration was thought to
be sufficient to give optimum results for surface analysis. The standard wetting and
drying cycle was 30 minutes for spraying the specimens, and 90 minutes allowed for
drying.
A prolonged drying period was seen to increase the surface ratings for
exfoliation attack, whereas a longer wetting period had no effect on the exfoliation
ratings.
Cyclic acidified salt fog MASTMAASIS testing is another accelerated test method
developed by Lifka and Spowls.[21] The solution for MASTMAASIS is 5 percent NaCl
in distilled water buffered to pH 3 with acetic acid.[21] A commercial fog cabinet at 95F
is used to permit automatic cycling. One-spray cycle consists of a 45min spray period, a
22
2 hour purge period and a 3 hour soak period at high relative humidity due to the solution
maintained in the cabinet. Samples are suspended at an angle of 45o. The result then is
compared to standard photographs for determination of EFC extent.[35] The test period
depends on the susceptibility of alloys. A good correlation was also found between
results of outdoor exposure and MASTMAASIS for Al-Li alloys.
The MASTMAASIS test produces better results for long-term outdoor exposure. In
contrast to the simple, rapid EXCO test, however, cyclic acidified salt fog testing is
complex, requiring specialized apparatus and relatively long testing periods.
Comparison of accelerated and atmospheric exposure tests for corrosion of
aluminum alloys was conducted by Dedamborenea et al.[51] Two Al-Cu alloys
(AA2024-T4 and AA7075-T7351) and two Al-Li alloys (AA2091-T84 and AA8090T8171) were exposed for 2 years in a moderately aggressive marine environment, and
their corrosion behavior was compared with that in accelerated tests for intergranular
corrosion and exfoliation. The results in the aggressive EXCO solutions did not agree
well with those from outdoor exposure, even using the modified solution. The EXCO
test resulted in a much greater extent of attack than outdoor exposure for AA8090-T8171
and AA2091-T84. None of the solutions reproduced the pitting or intergranular attack
that developed during outdoor exposure.[51]
Braun also conducted laboratory exposure tests on these alloys.[52] His research
showed that the cyclic acidified salt fog test reproduced the outdoor corrosion data for the
Al-Li alloys AA8090-T81 and AA2091-T84 and marine exposure results reported for the
conventional alloys AA2024-T351 and AA7075-T7351.
indicated better exfoliation corrosion behavior of the alloys investigated except for
23
sufficient evidence of the depth of penetration or kinetics. To get the full information of
EFC, metallographic sectioning is often required.
corrosion.[54] Some other works have also used EIS to characterize the exfoliation
corrosion process of aluminum alloys.[55-57] Conde and Damborenea performed EIS
measurement on AA8090 in EXCO solution and found that impedance technique made it
possible to detect the different stages of localized attack of the alloy. One single time
constant in EIS spectra related to a corrosion process under activation control with hardly
any detachment of the surface grain. A depression in the capacitive arc and narrowing of
the phase angle indicated that the material was being attacked. Two time constants was a
consequence of the great amount of surface grains lost by the effect of the
delamination.[56] Similar work was done on Al-Li-Cu-X and Al-Mg-Li-X alloys in
EXCO solution. The earlier appearance of two-time constants for Al-Li-Mg-X indicates
that this alloy has higher susceptibility to EFC than AA2091 alloy, where two-time
constants became evident after 96hrs.[55] More recently, Cao et al. studied the
exfoliation corrosion of aluminum alloy AA7075 by EIS method.
equivalent circuit shown in figure 2.7 to analyze the EIS data.[54, 56] Similar to
Keddams findings, they concluded that the decrease of low frequency resistance and
increase of low frequency capacitance might originate from the appearance of pitting on
the surface of the corroding surface.[57] A decrease in Clf indicates the occurrence of
exfoliation corrosion. When EFC occurs, the low frequency inductive arc disappears and
the Nyquist plot is composed of two capacitive arcs.[57]
under four point loading condition is related to the thickness of the specimen by equation
/W=K/d3 where is the deflection, W is the load, d is thickness, K can be determined
initially from the sample dimensions.[4] By following the change in the deflection per
unit load, the changes in the effective cross-section of the specimen can be determined.
During the test, it is assumed that corrosion causes loss in one dimension only. Changes
in this dimension with time can then be followed. In Liddiards work, the specimens
were corroded in the unstressed condition, and the thickness change with time due to the
exfoliation corrosion can be determined by intermittent measurement. Robinson et al.
also used four point bending method to measure the mean depth of attack of the AA2024
and AA2014 specimens after salt fog cabinet or coastal exposure tests.[1] However, this
technique requires a complicated experimental setup.
thinning of the specimen during corrosion was met after long exposure times when the
entire specimen surface underwent exfoliation. It was less valid in the earlier stages of
the test when some areas of the surface were free of attack or where the corrosion was in
the form of localized pitting.
corrosion cracking as well as their relation with exfoliation corrosion will be reviewed.
26
2.4
INTERGRANULAR CORROSION
Intergranular corrosion is a selective attack of grain boundaries or closely adjacent
regions without appreciable attack of the grains themselves.[16] Most heat treatable high
strength aluminum alloys are susceptible to IGC to some extent due to the formation of
precipitates along grain boundaries during heat treatment. It has been reported that there
is a correlation between the susceptibility to IGC and exfoliation in some Al alloys.[58,
59] In this section, the mechanisms of IGC, the relationship between IGC and exfoliation
and the causes of susceptibility to IGC in high strength aluminum alloys will be reviewed.
2.4.1
IGC Mechanisms
The mechanisms of IGC are electrochemical.[38] Generally speaking, there are three
major accepted explanations of IGC in aluminum alloys. They are:
1.
2.
3.
27
between the grain boundary region, which includes precipitate free zone and the grain
boundary precipitates, and the matrix was only 5 mV. Though the potential difference is
small, the material failed along the grain boundaries with the application of stress.[60, 63]
Dix et al. explained this behavior based on the fact that the (MgZn2) phase is active to
28
the matrix and the preferential dissolution of the phase leads to susceptibility to
intergranular and IGSCC.[60, 63] Many experiments have shown that IGC and SCC are
not seen in solutions that do not contain Cl- ions. The galvanic couple theory fails to
explain the role of chloride ions and other halide ions on intergranular corrosion of Al
alloys.
precipitate Free Zone (PFZ) Breakdown Model, which explain the effect of chloride ion
on IGC behavior in Al-Cu alloys.[8] They proposed that the difference in breakdown
potential between the grain matrix and the solute depleted zones adjacent to the grain
boundaries produces the IGC susceptibility of aged Al alloys.
In studying the mechanism of IGC of Al-4%Cu aged alloys, solutionized Al4%Cu, Al2Cu and Al-0.2%Cu were used to simulate the behavior of the grain bodies,
intermetallic Al2Cu phase and copper depleted zone along the grain boundaries in aged
Al-4%Cu, respectively.[64] Anodic polarization was performed in deareated NaCl
solution with various concentrations to determine the breakdown potentials. The critical
potential of Al-4%Cu solid solution, Al2Cu intermetallic phase, Al-0.2%Cu and pure Al
is related to activity of NaCl. The breakdown potentials of the Al-4%Cu solid solution
alloy were similar to that of the intermetallic Al2 Cu. There were no differences in the
breakdown potentials of the Al-0.2%Cu and high purity aluminum. The breakdown
29
potential for the Cu rich phases were 100 mV higher than the values for lower Cu content
specimens.
Galvele and DeMicheli found a clear relationship between pitting and
intergranular corrosion. Al-4%Cu in its peak-aged condition exhibits two breakdown
potentials when anodically polarized in Cl- containing solution.[64-66] They explained
that the two breakdown potentials in this alloys are due to the copper depleted PFZ
present along the grain boundaries of the peak-aged Al-4%Cu. This Cu depleted zone
approximate to Al-0.2%Cu has a lower breakdown potential than the grain matrix and
grain boundary precipitate (Al2Cu).
breakdown potential but lower than the matrix breakdown potential, localized attack on
the PFZ results in IGC; when the applied potential is above the more noble breakdown
potential, it showed both IGC and pitting in the matrix. Consistent with this explanation,
the experiment showed that the over aged Al-4%Cu had only one breakdown potential
because the grain body is mainly a two-phase structure and the increase in anodic area
will favor generalized corrosion rather than IGC. They found that the IGC and pitting of
Al only occurred in the presence of certain anions such as Cl-, Br-, I-, ClO4 -.
Intergranular attack in AA7xxx (Al-Zn-Mg-Cu) series was studied by Maitra et
al.[15, 67]
They found that the alloys showed different localized corrosion behavior
solution for W temper showed only one breakdown potential at 800 mV SCE, which
corresponded to pitting corrosion in the single-phase solid solution matrix. The alloys
were susceptible to intergranular corrosion and exhibited two breakdown potentials in T6
temper. The more noble breakdown potential 725mV SCE corresponded to the pitting
30
of solid solution matrix (containing less Zn and Mg) and the breakdown potential
800mV SCE was assumed to relate to the pitting of the solute enriched grain boundary
region. IGC was observed between two breakdown potentials in T6 temper. T7 temper
exhibited decreased susceptibility to IGC and the polarization curve consisted of one
breakdown potential at 0.765mV SCE. Pitting was observed above this potential. They
mentioned that if the microstructure contains more phases with distinct breakdown
potentials, the polarization curves would consist of a number of plateaus. The phase will
dissolve if the potential is above its breakdown potential. The authors explained that the
difference is due to the segregation of different elements to the grain boundary during
tempering. During T6 tempering, Zn and Mg solute atoms diffuse to the grain boundary,
which results in lowering the breakdown potential of the grain boundary.[15] The reason
for only a single breakdown potential in the T7 temper is attributed to the fact that Cu is
removed from solid solution and incorporated in eta phase (Mg (Al, Cu, Zn)) resulting in
the shift of second breakdown potential to negative value while the solute enriched grain
boundary region becomes narrower and less continuous.
particles and the surrounding solid solution in a thin surface layer. The copper content in
these hardening particles controls the first breakdown potential. The second breakdown
potential was associated with combined intergranular and selective grain attack.
Guillaumin and Mankowski studied the corrosion behavior of AA2024-T351 in
chloride media and proposed that the first breakdown potential was related to the
dissolution of Al2CuMg particles or dealloying of Mg from the particles and the second
breakdown potential corresponded to the matrix breakdown potential. Both pitting and
intergranular corrosion developed above second breakdown potential.[69] In their study
on AA6065-T6 in 1 M NaCl solution, they found that intergranular corrosion and pitting
are dependent on each other. Pits first developed and then led to intergranular corrosion.
They proposed that the mechanism of AA6056-T6 alloys consists of preferential
dissolution of the anodic Cu and Si depleted zone along grain boundaries.[70]
3wt%.[59] In copper free AA7xxx (Al-Zn-Mg) series alloys, the anodic path is generally
considered to be the anodic zinc and magnesium bearing constituents at the grain
boundary.[71] Selective T1 phase dissolution for IGC in Al-Li-Cu alloy was first
proposed by Rinker when he studied the corrosion and stress corrosion behavior of
AA2020.[43] T1 phase (CuAl2 Li) is believed to be the most active phase present in the
subgrain boundaries and matrix.[43] Buchheit et al. used anodic dissolution theory to
explain the IGC of Al-Li-Cu alloys.[43, 44, 72] They conducted electrochemical
measurements on T1 phase and -Al-matrix to determine the mechanism of IGC. The
Cu-depleted zone in AA2090 alloy was modeled by pure Al. The anodic polarizations of
Al2CuLi and pure Al in deareated 0.6 M NaCl solution showed a larger dissolution rate at
Ecorr of Al2 CuLi phase (10-4A/cm2) than that of Cu depleted zone and -Al-matrix, which
were less than 10-6 A/cm2.
deareated 3.5wt% NaCl solution showed that the T1 phase was more active than alloy
matrix. They concluded that the intergranular corrosion and pitting were controlled by
selective dissolution of T1 phase, which resulted in the creation of an aggressive
occluded environment leading to the continued grain boundary attack.
In summary, there are essentially three different theories proposed to explain IGC
susceptibility of Al alloys from electrochemical point of view. It is generally accepted
that susceptibility to IGC results from the precipitation of intermetallic phases in the
vicinity of the grain boundaries.
precipitates and the adjacent PFZ zones vary from one alloy system to another and will
be reviewed in the following.
33
2.4.2
Alloy Microstructure
It is known that the susceptibility to IGC, EFC, and IGSCC of aluminum alloys is
strongly affected by alloy microstructure such as the dispersion of various intermetallic
particles and precipitate free zones.
2.4.2.1.1
Al2Cu ( phase) and Al2CuMg (S phase) are the most important intermetallic
particles in Al-Cu and Al-Cu-Mg alloys. The electrochemical behavior of these phases
has been studied in detail.[66, 74-78] It is well known that Al2Cu phase has a more noble
corrosion potential than the Al matrix.[64, 66] A study on the effect of solution aeration
on the OCP of Al2Cu showed that OCP ranges from 0.59 to 0.7 V for aerated and
deareated solutions with chloride concentrations ranging from 0.2 to 1M.[65] The
susceptibility to IGC of alloys is influenced primarily by copper concentration gradients
in the Al-Cu solid solution in the grain boundary regions.[38] Brown et al. claimed that
the retention of 5 percent copper in solid solution changes the electrode potential of
aluminum about 200 mV in the cathodic direction.[38, 62] Mazurkiewicz et al. studied
the electrochemical properties of Al2 Cu using potentiodynamic polarization on stationary
electrodes and rotating ring disk-electrodes to study the mechanism of dissolution of
Al2Cu.[79] The dissolution of Al2Cu leads to the plating of copper. They stated that
Al2Cu dissolution followed the following reactions:
Al2CuCuo +2Al3++6eAl2Cu2Al+3+Cu2++8eand Cu2++2eCuo
Potential dynamic polarization of Al2Cu in sulphate solutions showed that when
the anodic potential of copper dissolution was reached, the maximum current was
reached.[79] Above the current maximum, the anodic process is controlled by aluminum
hydroxide film in low and neutral pH solutions because copper dissolves at a relatively
high rate. In alkaline solution, it is controlled by aluminum hydroxide films and the
35
stability of copper before the potential of formation of CuO22- is reached. The anodic
polarization of Al2 Cu and Al in 1 M NaCl solution showed that the breakdown potential
of Al was 0.2V lower than that of Al2Cu. Guillaumin et al. concluded that the large
potential difference between Al and Al2Cu leads to the galvanic dissolution of aluminum
anodes at OCP.
Al2CuMg (S phase) is a key contributor to the localized corrosion in Al-Cu-Mg
alloys. Electrochemical measurements on bulk S phase showed that it has an open circuit
potential ranging from 925 to 935 mV SCE in 0.5 M NaCl solution independent of
solution aeration.[75] S phase is more active than the Cu depleted zone and -Al matrix
phase (Al-4Cu solid solution). It undergoes strong anodic polarization, which stimulates
dealloying of the particles and nonfaradaic liberation of Cu clusters, consequently
changing the galvanic relationship with the surrounding microstructure.[76] Buchheit et
al. investigated the dealloying of S phase.[76] They stated that Al2CuMg (S) particles are
initially electrochemically active with respect to the AA2024-T3 matrix. S phase is
dealloyed in aggressive solutions leaving Cu rich remnants that behave like noble sites.
X-ray microanalysis of corroded surface and X-ray line scans across an S phase exposed
to a pH 4.2 chloride solution provides the proof that Al and Mg are selectively dissolved
from S phase and leave Cu rich remnants behind. Two types of pit morphologies were
found in the matrix as a result of S phase dealloying. The first type of pits was observed
at the periphery of the remnants. The second type was observed far away from the
primary pits. For the second type, Buchheit proposes that the particle remnant from
selective dissolution decomposes into 10 to 100 nm Cu clusters, which detach from the
36
alloy surface and redistributed across the surface. These clusters act as local cathodes
and induce peripheral matrix pitting.
Further investigations using a rotating ring disk collection experiment and
stripping voltammetry provide the evidence for Cu ion formation by dissolution and
dealloying the Al2CuMg intermetallic compound.[80] The cyclic voltammogram for the
collection experiment conducted in aerated solution with Al2CuMg shows three peaks
that correspond to the oxidative formation of CuCl(s) from Cu, formation of Cu+ at
conditions far from equilibrium and the formation of Cu2+ due to oxidation of CuCl (s).
Metallic Cu was detected by X-ray microchemical analysis of the ring surface. Buchheit
et al. concluded that the Cu ion generation mechanism contributed to the poor corrosion
resistance and poor conversion coating characteristics of Al2CuMg bearing alloys. Figure
2.8 shows a schematic illustration of the mechanism for redistribution of Cu by
dissolution of Al2CuMg and Al2Cu intemallic particles. The particles are anodically
polarized by the surrounding Al matrix phase at an OCP of 0.93~-0.915V, which is
negative of E Cu / Cu 2 + . The Cu cluster remnants are captured by the corrosion product gel,
detach from the particle remnant and move away from the corrosion site due to solution
movement. The OCP of Cu in near neutral chloride solutions is positive to E Cu / Cu 2 + so
that the oxidation of the cluster is possible to generate Cu ions.
Figure 2.9 shows potentiodynamic polarization curves for Al2CuMg in air sparged
and N2 sparged 0.5 M NaCl solution.[75] The positive shift in OCP is due to the
increasing oxygen dissolved in the solution, which intensifies the Cu enrichment process
causing a greater shift in OCP.[75] On the return scan, the corrosion potential was
detected to return to 0.92 to 0.94V indicating that Cu enrichment is eliminated by
37
anodic polarization. The returning value of OCP also supports the previous study on
nonfaradaic liberation of Cu clusters from dealloyed Al2 CuMg in which the detached Cu
cluster did not directly contribute to measurable electroactivity.[76]
Other studies on AA2024-T3 showed that another type of intermetallic particle in
Al-Cu-X alloys is AlCu(Fe,Mn).[81, 82] Al-Cu(Fe, Mn) particles have a potential of
675mV SCE, which is 200mV more noble than the matrix of Al-Cu-Mg based alloys
having an OCP of 0.65V SCE in 0.1 M NaCl.
2.4.2.1.2
It was shown that MgZn2 has OCP of 1400mV SCE and a breakdown
potential of 1140mV SCE in deareated 0.5 M NaCl. The addition of 17 and 27 atom %
Cu results in an increasing corrosion potential by 250 and 300 mV and breakdown
potential by 150 and 210 mV, respectively, while the breakdown potential of eta phase is
relatively unaffected by Al addition up to 10 atom%.
38
provided the relevance to localized corrosion of eta phase. The low Cu containing
intermetallic compounds completely dissolved in chloride solution while high Cu
containing intermetallic compounds dealloyed, resulting in the enrichment of Cu, which
can have a different effect on the corrosion behavior of AA7xxx. Unlike the uniform
attack in neutral solution, the eta phase thin film analog was susceptible to more classic
localized breakdown in high pH due to the enhanced stability of Mg (OH)2 in high pH.
Ramgopal investigated the mechanism of IGC in AA7150 and explained the
higher resistance to IGC of T7 temper than T6 temper. He used analytical transmission
electron microscopy to study grain boundary phases of AA7150 and found that the
particles in T7 temper had much higher Cu concentration and lower Zn concentration
than those in T6 temper.[84] The composition of the matrix and PFZ for two tempers is
similar and has similar electrochemical behavior in the same solution. Electrochemical
tests on thin film compositional analogs of T6 and T7 grain boundary precipitates
indicated that the breakdown potential for both tempers is below the breakdown potential
of the alloy, indicating that grain boundary precipitate does not initiate IGC directly.
However, the dissolution of precipitates of two tempers can create different
microchemistries. Dissolution of T7 grain boundary precipitates Mg (Zn, Cu, Al) results
in higher content of Cu ions in solution, hence preventing the dissolution of PFZ and
resulting in a high resistance to IGC.
2.4.2.1.3
AlFeSi phases on the pitting[85-89] and exfoliation[5, 20, 90] of aluminum alloys has
been extensively studied. It is generally known that the presence of iron in such forms is
deleterious to the corrosion resistance of aluminum alloys. Nisancioglu investigated the
electrochemical properties of the phases Al3 Fe, -Al(Fe, Mn)Si and AlFeSi in NaOH
solutions.[91] An alkaline environment was chosen as the test medium because diffusion
layers of high pH are known to form around intermetallics acting as local cathodes on a
corroding alloy in aqueous solutions. Figure 2.10 shows the variation of corrosion
potential of Al3Fe and -Al(Fe, Mn)Si with time in deaerated 0.1N NaOH solution.[91]
The rapid shift in potential for Al3 Fe was also observed by Golubev and Ronzhin.[92]
According to their interpretation, aluminum undergoes selective dissolution at the more
negative potential, and the surface of Al3Fe becomes richer in Fe component, which
shifts corrosion potential from the initial value of 1270 mV SCE to 1180 mV SCE. Al(Fe, Mn)Si follows the same process at a much slower rate. At an applied potential of
1.2V SCE, the current density-time behavior of Al3Fe was unstable due to the
detachment of Fe rich flakes after dealloying. Similar treatment on -Al(Fe, Mn)Si
particles also resulted an enrichment of surface with Mn and Si in addition to Fe. Cyclic
polarization of Al3Fe shows several peaks, figure 2.11.
~ -1.2V was correlated with the selective dissolution of aluminum followed by inhibition
of aluminum dissolution by the formation of a passive FeOH monolayer.[88, 92, 93]
Other peaks at more noble potentials were related to oxidation of iron into successively
higher valence states.[91] The polarization curves of -Al(Fe, Mn)Si and AlFeSi show
two superimposed anodic peaks, which are also controlled by aluminum dissolution and
the formation of passive iron oxide film.
40
formation of precipitate nuclei.[95] The other mechanism is that the PFZ is formed
through solute depletion caused by the preferential precipitation in the grain
boundaries.[73, 96] Which mechanism is important depends on the heat treatment of the
material. The width of the PFZ depends on many factors such as alloy composition,
solution treatment temperature, quenching rate, aging temperature and period of natural
aging before artificial aging. Many investigators have studied the effect of cooling rate
during quenching on the resistance to intergranular corrosion and to stress corrosion
cracking.[97-99] Higher solution treatment temperature, faster quenching rate, and lower
aging temperature produce narrower PFZs.[38]
The existence of PFZs has been reported to cause high susceptibility to IGC in
many Al alloy systems, such as Al-4%Cu[64], AA2024 (Al-Cu-Mg)[69], AA2090 (AlLi-Cu)[100], AA6056-T6 (Al-Mg-Si-Cu)[70], AA6013-T6 (Al-Mg-Si-Cu)[70], and AlZn-Mg-Cu (7xxx)[15]. The anodic PFZs along grain boundaries in these alloys contribute
to their high susceptibility to intergranular corrosion.
2.5
will help us to better understand the phenomenon of exfoliation corrosion. The literature
review here will be confined to the two main proposed mechanisms of SCC. The effect
of materials microstructure and microchemistry on stress corrosion cracking will also be
reviewed.
2.5.1
Mechanisms of SCC
Critical environment.
2.
A susceptible alloy.
3.
The sequence of events involved in the SCC process is divided into three stages:[12]
1.
2.
3.
propagation rate as a function of crack tip stress intensity behavior. The magnitude of the
stress distribution at the crack tip is quantified by the stress intensity factor, K, for the
specific crack and loading geometry. No crack propagation is observed below KISCC,
which is determined by the alloy, its environment and metallurgical condition. The rate
of crack propagation exceeds the plateau velocity as the stress intensity level approaches
KIc.[12]
Though SCC has been studied for many years, the initiation of cracks is not well
understood.
Many papers provide the evidence that cracks can initiate at surface
43
path.
2.
model.[12] This model assumes that the stress acts to open the crack and rupture the
protective surface film. Some reports have evidence to support this model.[101-106]
However, this theory is not generally accepted as a mechanism of transgranular SCC.[12]
Most researchers proposed that the mechanism of SCC for AA2xxx alloys (both
Al-Cu and Al-Cu-Li alloys) is the anodic dissolution along the grain boundaries.[107] Izu,
Shiokawa, and Sato consider local anodic dissolution as the controlling factor in the SCC
of AA2017 alloys. They showed that the time to failure of stressed alloy was dependent
on the applied current density, not on the applied stress.[108] Gruhl stated that Al-Cu-Mg
alloys did not truly stress corrosion crack, but rather exhibit intergranular corrosion,
44
which could occur without stress but was accelerated under stress.[109] In the study of
AA2020 alloys under varying aging treatments, Rinker et al. proposed that the variation
in SCC was due to preferential corrosion of the T1 phases.[110] Meletis studied the SCC
behavior of Al-2.9Cu-2.2Li-0.12Zr in the peak-aged and over-aged conditions using
DCB specimens and TEM analysis. He found that peak-aged temper was the most
resistant to SCC and over-aged temper is more susceptible to SCC. Meletis explained
that preferential T1 precipitation in over-aged condition produced localized strain at the
grain boundary, which promoted either dissolution or hydrogen embrittlement along the
grain boundaries.[111]
Some researchers also considered anodic dissolution mechanism as the
mechanism of SCC in the AA7xxx alloys.[15, 112-115] Joshi et al. found that there were
greater amounts of Cu, Mg and Zn along grain boundaries in T6 temper than T73 temper,
which can explain their different corrosion behavior.[116] Sedriks et al. studied the
electrochemical behaviors of the intermetallic phase MgZn2 and Al-Zn-Mg alloys in
NaCl and AlCl3 solutions.[112] They proposed that a crack was formed by the lateral
coalescence of pits due to the dissolution of MgZn2 precipitates and propagated by
ductile fracture and active dissolution of MgZn2 particles. The process is repeated when
the crack tip reaches the next MgZn2 particle. However, Poulose et al. have a different
point of view regarding the role of MgZn2. They found that the SCC crack velocity was
inversely proportional to the volume of MgZn2. Hence they proposed that the grain
boundary precipitates acted as sacrificial anodes to retard IGSCC.[117]
45
transport hydrogen deep into the alloy and cause brittle intergranular fracture.[129] Grugl
suggested that hydrogen transport was by the grain boundaries.[109] He proposed that
increasing concentration of Zn in solid solution would decrease the solubility of hydrogen
in the grain boundaries and hence increased the time to failure. On the other hand, tensile
stress normal to grain boundaries opened the lattice and helped the diffusion of hydrogen
into grain boundaries. Talianker and Cina used TEM to study the dislocation density of
the grains and grain boundaries in AA7075, AA7050 and AA7278 in T6 temper and in
retrogression and re-aged treatment. They observed high density of dislocations in the T6
temper, which disappeared after RRA treatment. Their study supports both of the above
theories that grain boundary dislocations transport hydrogen.[130]
Thompson has found out that mode I loading on AA7075 led to the much higher
susceptibility to SCC due to a hydrogen embrittlement mechanism.[131] Other studies
showe that alloys containing large fraction of coherent particles and few active slip
planes have high susceptibility to SCC.[132] Since the copper rich alloys have lower
volume fraction of coherent particles, these alloys have less susceptibility to SCC.
2.5.2
Chemical
composition and metallurgical treatment can change the microstructure and markedly
influence the resistance of an alloy to SCC.
Kent investigated the effect of quench rate on the microstructure and SCC
resistance of Al-Zn-Mg alloys.[133] Slow quenching rate resulted in an increase in the
size and spacing of the precipitates at high angle grain boundaries, an increase in the
47
width of PFZ and consequently a longer time to fail. He claimed that an increase in the
PFZ width and an increase in the size and spacing of grain boundary precipitate lead to
excellent SCC resistance. However, a later investigation conducted by De Ardo indicated
that there was no correlation between PFZ width and the susceptibility, though the size
and spacing of grain boundary precipitate was very important in determining the SCC
susceptibility.[134] Rajan et al. have suggested that coarse grain boundary precipitates
act as hydrogen traps and can create hydrogen bubbles.[122]
Speidel reviewed the literature on SCC of aluminum alloys and concluded that the
over aged temper T76 and T73 provided high strength aluminum alloys with improved
stress corrosion resistance.[132]
Puiggali et al. investigated the susceptibility to stress corrosion cracking of Al-ZnMg-Cu alloy of different temper (under aged T351, peak aged T651 and over aged
T7451) by slow strain rate testing.[135] They found that T7451 alloy was relatively
resistant to corrosion and immune to SCC and T651 alloy was resistant to corrosion to a
similar extent but susceptible to SCC. They explained that during aging, the GP zones
might develop into MgZn2 and precipitates whose volume increases with prolonged
aging. Large MgZn2 precipitates became obstacles for the crack initiation and trapped a
great amount of hydrogen, thereby decreasing its interstitial concentration below a critical
value. As a result, the susceptibility to SCC was decreased.[136] T351 alloy had the
smallest precipitates and showed high susceptibility to pure corrosion, which seemed to
decrease the relative risk of SCC. The pitting potential was the most negative for T351
temper alloy and the most positive for T7451 temper alloy indicating their decreasing
tendency to pitting from which the cracks may initiate.
48
Lin et al. have studied the effect of minor alloying addition of Mn, Cr, Zr and Ti
on SCC of Al-Zn-Mg-Cu alloys. They concluded that Mn, Cr, Zr and Ti could all reduce
the susceptibility to SCC but by different mechanisms.
Hardwick, Thompson and Bernstein catholically charged Al-6Zn-2Mg alloys,
AA7050 and low copper AA7050 alloys with hydrogen to determine the effect of copper
and microstructure on the hydrogen embrittlement of these alloys. The results showed
that AA7050 was susceptible to SCC only for under aged condition and was highly
resistant to SCC for peak aged and over aged conditions. Low copper AA7075, on the
other hand, was embrittled for all the tempers, which indicated a beneficial effect of
copper additions for increasing SCC resistance.[137]
Scamans and Holroyd proposed that the magnesium promotes hydrogen diffusion
into aluminum.[126, 138] They stated, SCC only occurs when cracks can be nucleated
on grain boundary precipitates[139]. The enrichment of magnesium increases corrosion
activity and enhanced hydrogen entry by forming magnesium hydride.[138] However,
Pickens and Langan reported different results.[140] They conducted SCC tests on boltloaded DCB specimen and measured the grain boundary composition. No correlation
between Mg segregation and SCC susceptibility was found in their study.
Microstructural features of aluminum alloys play an important role in SCC, yet
many issues remain unresolved. There is no agreement on the composition of the solute
depleted zone and role of grain boundary precipitates in SCC, both of which are
extremely important in the understanding of SCC.
49
2.5.3
Corrosion
IGSCC and EFC in high strength aluminum alloys have been thought to have close
relationship.[10] They both can initiate from intergranular corrosion and both propagate
by a stress assisted intergraunlar corrosion mechanism.[1, 9, 10, 23] In the case of SCC,
the stress is generally applied through an external load whereas in exfoliation the stress
arises internally from the wedging effect of the voluminous corrosion products. Both
stress corrosion cracking and exfoliation are strongly influenced by their copper content
and heat treatment conditions. It has been shown that some factors that lead to an
increase in the rate of grain boundary corrosion in exfoliation also have a similar effect
on SCC. For example, SCC also propagates more rapidly on more elongated grain
structure. At low stress intensities, SCC velocities are comparable with the rate of
intergranular penetration for exfoliation corrosion.[23]
2.6
RESEARCH OBJECTIVES
While the qualitative effects of grain shape and heat treatment condition on
exfoliation susceptibility are well known, there has been little attempt to measure these
effects quantitatively. The reason is that no satisfactory method to quantify the extent of
exfoliation exists. What is the best accelerated environment for laboratory testing of EFC
is still an open question. The first objective of this work is to develop a laboratory testing
method and to quantitatively investigate the effect of grain shape and heat treatment on
exfoliation corrosion kinetics.
50
It has been noticed that some materials can be free from exfoliation during
exposure in one location for many years, and attacked by exfoliation within several
months in another place. This can be explained by a threshold humidity, below which the
exfoliation cannot occur. Little attempt has been made to investigate the threshold
humidity of exfoliation corrosion and few reports about the effects of relative humidity
on susceptibility to exfoliation corrosion exist. The effect of humidity is studied in this
work and the threshold humidity for EFC propagation is determined.
To understand the nature of exfoliation, it is important to understand the role of
the grain boundary constituents, namely the behavior of the PFZ and the precipitates, in
determining EFC susceptibility. From the literature review above, materials having same
grain structure could have different susceptibility to EFC.[1] The question can probably
be solved by looking at the microchemistry in the grain interior or at the grain boundary.
In the proposed research, TEM and FIB are utilized to characterize the grain boundary
precipitates and PFZ to investigate the correlation between microchemistry and EFC
susceptibility.
The existence of exfoliation corrosion on aircraft structure is detrimental to its
fracture and fatigue behavior.[3, 6]
propagation of exfoliation corrosion.
compressive and tensile stress either during flight or on the ground. It is important to
understand the effect of mechanical stress on exfoliation corrosion. In this work, a
special apparatus is designed to apply surface strain. The effects of compression and the
tension on exfoliation corrosion kinetics are quantitatively studied.
51
52
REFERENCES
M.J. Robinson and N.C. Jackson, Corrosion Science, 41, 1013-1028 (1999).
J.P. Chubb, T.A. Morad, B.S. Hockenhull, and J.W. Bristow, International Journal
of Fatigue, 17(1), 49-54 (1995).
E.A.G. Liddiard, J.A. Whittaker, and H.K. Farmery, Journal of the Institute of
Metals, 89, 377-384 (1960-1961).
J. Zahavi and J. Yahalom, Journal of The Electrochemical Society, 129(6), 11811186 (1982).
P.C. Conor, A.D. James, and R.N. Collier, "Evaluation of Exfoliation Corrosion
Damage to 7075-T6 Aluminum Alloy Aircraft Structural Components", Materials
and Structures Group, Defense Technology Agency (2004).
D.O. Sprowls, J.D. Walsh, and M.B. Shumaker, "Simplified Exfoliation Testing of
Aluminum Alloys", in Localized Corrosion-Cause of Metal Failure, ASTM STP
516, American Society for Testing and Materials, 38-65 (1972).
M.J. Robinson and N.C. Jackson, British Corrosion Journal, 34(1), 45-49 (1999).
10
11
W. Zhang and G.S. Frankel, Electrochemical and Solid-State Letters, 3(6), 268270 (2000).
12
R.H. Jones, Stress Corrosion Cracking. Materials Park, Ohio: ASM International,
1-41 (1992).
13
53
14
15
16
D.A. Jones, "Principles and Prevention of Corrosion", 2nd ed. Upper Saddle
River, NJ: Prentice-Hall, Inc. (1996).
17
18
L.L. Shreir, F.R.I. C, and F.I. M, "Corrosion", 2nd ed., ed., L.G.Shreir. NewnesButterworth (1976).
19
20
21
22
H.W. Pickering, F.H. Beck, and M.G. Fontana, Corrosion, 18, 230-239 (1962).
23
24
D. McNaughtan, M. Worsfold, and M.J. Robinson, Corrosion Science, 45, 23772389 (2003).
25
G. Bassi and J.J. Theler, Zeitschrift fur Metallkunde, 60(3), 179-184 (1969).
26
27
K. Ebtehaj, D. Hardie, and R.N. Parkins, British Corrosion Journal, 24(3), 184188 (1989).
28
29
B.W. Lifka, D.O. Sprowls, and J.G. Kaufman, Exfoliation and Stress-Corrosion
Characteristics of High Strength, Heat Treatable Aluminum Alloy Plate, in Joint
Conference of North Central and Northeast Regions, National Association of
Corrosion Engineers, 1967, Columbus, Ohio.
30
54
31
32
33
34
35
G.S. Haynes and R. Baboian, "Modified Salt Spray (Fog) Testing, Laboratory
Corrosion Tests and Standards", in ASTM Standard G85-85 A2. STP866,
Philadelphia, PA: ASTM, 578-583 (1985).
36
"Standard Test Method for Exfoliation Corrosion Susceptibility in 2xxx and 7xxx
Series Aluminum Alloys (EXCO Test)", in Annual Book of ASTM Standards,
Philadelphia, PA: The American Society for Testing and Materials, 114-119
(1990).
37
D.J. Kelly, "Exfoliation and Stress Corrosion Cracking of the Aluminum Lithium
Alloy 8090", Cranfield University. (1991).
38
39
40
W.A. Bell and H.S. Campbell, Journal of the Institute of Metals, 89, 464-471
(1960-1961).
41
42
43
R.G. Buchheit, J.P. Moran, and G.E. Stoner, Corrosion Science, 50(2), 120-130
(1994).
44
R.G. Buchheit, J.P. Moran, and G.E. Stoner, Corrosion Science, 46(8), 610-617
(1990).
45
H.M. Flower and P.J. Gregson, Materials Science and Technology, 3, 81-90
(1987).
55
46
S. Lee and B.W. Lifka, "Modification of the EXCO Test Method for Exfoliation
Corrosion Susceptibility in 7XXX, 2XXX and Aluminum-Lithium Alloys", in New
Methods for Corrosion Testing of Aluminum Alloys, ASTM STP 1134, eds.,
V.S.Agarwala and G.M.Ugiansky, Philadelplia, PA: American Society for Testing
and Materials, 1-19 (1992).
47
48
D.O. Sprowls, T.J. Summerson, and F.E. Loftin, "Exfoliation Corrosion Testing
of 7075 and 7178 Aluminum Alloys-Interim Report on Atmospheric Exposure
Tests", in American Society for Testing and Materials, Philadelphia, PA: 99-112
(1973).
49
V.E. Carter and H.S. Campbell, Journal of the Institute of Metals, 89, 472-475
(1961).
50
J.R. Davis, Corrosion of Aluminum and Aluminum Alloys. Materials Park, OH:
ASM International, p.68 (1999).
51
J.D. Damborenea and A. Conde, British Corrosion Journal, 30(4), 292-296 (1995).
52
53
54
55
56
57
F.H. Cao, Z. Zhang, J.F. Li, Y.L. Cheng, J.Q. Zhang, and C.N. Can, Materials and
Corrosion, 55(1), 18-23 (2004).
58
59
D.R. Baer, C.F. Windisch, M.H. Engelhard, M.J. Danielson, and R.H. Jones,
Journal of Vacuum Science and Technology A, 18(1), 131-136 (2000).
60
61
M.S. Hunter, G.R. Frank, and D.L. Robinson,, in Second International Congress
on Metallic Corrosion. 1963. National Association of Corrosion Engineer.
56
62
R.H. Brown, W.L. Fink, and M.S. Hunter, Transactions, American Institute of
Mining and Metallurgical Engineers, 143, 115-123 (1941).
63
E.H. Dix Jr., Transactions of the American Society for Metals, 42, 1057-1127
(1950).
64
J.R. Galvele and S.M. De De Micheli, Corrosion Science, 10, 795-807 (1970).
65
66
67
68
Q. Meng and G.S. Frankel, Surface and Interface Analysis, 36, 30-42
69
70
71
72
R.G. Buchheit, F.D. Wall, G.E. Stoner, and J.P. Moran, Corrosion, 51(6), 417-428
(1995).
73
A.J. Cornish and M.K.B. Day, Journal of the Institute of Metals, 97, 44-52 (1969).
74
75
R.G. Buchheit, L.P. Montes, M.A. Martinez, J. Michael, and P.F. Hlava, Journal
of The Electrochenical Society, 146(12), 4424-4428 (1999).
76
R.G. Buchheit, R.P. Grant, P.F. Hlava, B. McKenzie, and G.L. Zender, Journal of
The Electrochemical Society, 144(8), 2621-2628 (1997).
77
J.R. Scully, T.O. Knight, R.G. Buchheit, and D.E. Peebles, Corrosion Science, 33,
185-195 (1993).
78
79
57
80
R.G. Buchheit, M.A. Martinez, and L.P. Montes, Journal of The Electrochemical
Society, 147(1), 119-124 (2000).
81
82
G.S. Chen, M. Gao, and R.P. Wei, Corrosion, 52(1), 8-15 (1996).
83
L.I. Maissel and R. Glang, "Handbook of Thin Film Technology", New York:
McGrawHill (1970).
84
T. Ramgopal, P.I. Gouma, and G.S. Frankel, Corrosion, 58(8), 687-697 (2002).
85
86
87
K. Sugimoto and Y. Sawada, Journal of the Japan Institute of Metals, 37, 148
(1973).
88
89
A.P. Bond, G.F. Bolling, H.A. Domian, and H. Biloni, Journal of The
Electrochemical Society, 113, 773 (1966).
90
91
92
A.I. Golubev and M.N. Ronzhin, "Corrosion of Metals and Alloys", eds.,
N.D.Tomashov and E.N.Mirolyubev. London: Oldbourne Press (1966).
93
94
W.W. Binger, E.H. Hollingsworth, and D.O. Sprowls, "Aluminum", ed., K.R.Van
Horn. Metals Park, OH: American Society for Metals (1967).
95
96
97
98
99
100
C. Kumai, J. Kusiniski, G. Thomas, and T.M. Devine, Corrosion, 45, 294 (1989).
101
102
103
E.W. Hart, "Surface and Interfaces II", Syracuse University Press, p.120 (1968).
104
H.J. Engle, "The Theory of Stress Corrosion Cracking in Alloys", North Atlantic
Treaty Organization (1971).
105
106
107
108
M. Izu, M. Shiokawa, and E. Sato, Journal of the Japan Institute of Metals, 34(11),
643-649 (1984).
109
110
J.G. Rinker, M. Marek, and T.H. Sanders, Materials Science and Engineering,
64(2), 203-221 (1984).
111
112
A.J. Sedriks, J.A.S. Green, and D.L. Novak, Corrosion Processes and Solution
Chemistry Within Stress Corrosion Cracks in Aluminum Alloys, in National
Association of Corrosion Engineers, 1971, Houston, TX.
113
114
Y. Choi, H.-C. Kim, and S.-I. Pyun, Journal of Materials Science, 19(5), 15171521 (1984).
115
116
117
P.K. Poulose, J.E. Morral, and A.J. Mevily, Metallurgical Transactions, 5(6),
1393-1400 (1974).
118
I.L. Muller and J.R. Galvele, Corrosion Science, 17, 179 (1977).
59
119
120
121
122
K. Rajan, W. Wallace, and J.C. Beddoes, Journal of Materials Science, 17, 28172824 (1982).
123
Z.-X. Tong, S. Lin, and C.-M. Hsiao, Metallurgical Transactions A, 20A, 921-924
(1989).
124
125
D. Hardie, N.J.H. Holroyd, and R.N. Parkins, Materials Science, 12(11), 603-610
(1979).
126
127
W.-Y. Chu, Y.-B. Wang, and C.-M. Hsiao, Corrosion, 38(11), 561-570 (1982).
128
Z.-X. Tong, S. Lin, and C.-M. Hsiao, Metallurgical Transactions A, 20A, 925-933
(1989).
129
130
131
132
133
134
135
136
137
D.A. Hardwick, A.W. Thompson, and I.M. Bernstein, Corrosion Science, 28(12),
1127-1137 (1988).
138
N.J.H. Holroyd and G.M. Scamans, Scripta Metallurgica, 19(7), 915-916 (1985).
139
G.M. Scamans, N.J.H. Holroyd, and C.D.S. Tuck, Corrosion Science, 27(4), 329347 (1987).
140
61
FIGURES
Figure 2. 1
Exfoliation resulting from rapid lateral attack of selective boundaries or
strata forming wedges of corrosion product that force layers of metal upward giving rise
to a layered appearance.[36] (Reprinted, with permission, from ASTM G34-90 Standard
Test Method for Exfoliation Corrosion Susceptibility in 2XXX and 7XXX Series
Aluminum Alloys (EXCO Test), copyright ASTM International, 100 Barr Harbor Drive,
West Conshohocken, PA 19428)
Figure 2.2
Schematic graph of the variation of corrosion product wedging force as a
function of time.[23] (Reprinted from Corrosion Science, Vol. 23, M.J. Robinson, The
Role of Wedging Stresses in The Exoliation Corrosion of High Strength Aluminum
Alloys, Pages: 887-899, Copyright (1983), with permission from Elsevier.)
62
Figure 2.3
(a) Model of the intergranular penetration of a wrought alloy showing the
boundary of the attack marked by the dotted line. (b) Improved model in which the
corrosion product tapers uniformly towards the tips.[9] (Reprinted from Corrosion
Science, Vol. 22, M.J. Robinson, Mathematical Modelling of Exfoliation Corrosion in
High Strength Aluminum Alloys, Pages: 775-790, Copyright (1982), with permission
from Elsevier.)
Figure 2.4
(a) Dimensions of a thin walled hemispherical shell used to calculate the
excess pressure in a blister. (b) Section through a blister diameter.[9] (Reprinted from
Corrosion Science, Vol. 22, M.J. Robinson, Mathematical Modelling of Exfoliation
Corrosion in High Strength Aluminum Alloys, Pages: 775-790, Copyright (1982), with
permission from Elsevier.)
63
Figure 2.5
Surface rating for 2024 in the MASTMASSIS test. (P=pitting,
EA=Superficial EFC,EB=Moderate EFC,EC=Severe EFC,ED=Very Severe EFC). [1]
(Reprinted from Corrosion Science, Vol. 41, M.J. Robinson and N.C. Jackson, The
Influence of Grain Structure and Intergranular Corrosion Rate on Exfoliation and Stress
Corrosion Cracking of High Strength Al-Cu-Mg Alloys, Pages: 1013-1028, Copyright
(1999), with permission from Elsevier.)
64
Figure 2.6
(a) Corrosion product forces generated on T/2 specimens of 8090 plates.
(b) Corrosion product forces generated on T/4 specimens of 8090 plate.[10] (Reprinted
from Corrosion Science, Vol. 49, D.J. Kelly and M.J. Robinson, Influence of Heat
Treatment and Grain Shape on Exfoliation Corrosion of Al-Li Alloy 8090, Pages: 787795, Copyright (1993), with permission from Elsevier.)
65
Figure 2.7
The schematic equivalent circuit for the immersion of AA7075 in EXCO
solution. (Rs: solution resistance; Cdl: origin surface double layer capacitive; Rp:
polarization resistance; Clf: low frequency capacitive; Rt: charge transfer resistance).
[57]
Figure 2. 8
Schematic illustration of a mechanism for redistribution of Cu by
dissolution of large Al2CuMg and Al2Cu intermetallic particles in Al alloys. [80]
(Reproduced by permission of The Electrochemical Society, Inc.)
66
Figure 2. 9
Potentiodynamic polarization scans for bulk Al2CuMg in 0.5M NaCl
solution open to air, actively aerated, and actively deaerated by sparging with N2.[75]
(Reproduced by permission of The Electrochemical Society, Inc.)
Figure 2. 11 Anodic part of cyclic polarization curve for Al3 Fe in deareated 0.1N
NaOH solution. The curve corresponds to the first cycle starting with an as polished
surface at 1.5V. In this particular run, the sweep was reversed at 0V. Continuous curve:
forward sweep. Dashed curve: reverse sweep. [91] (Reproduced by permission of The
Electrochemical Society, Inc.)
Figure 2.12 Schematic diagram of typical crack propagation rate as a function of crack
tip stress intensity behavior illustrating the regions of stage 1,2 and 3 crack propagation
as well as identifying the plateau velocity and the threshold stress intensity.[12]
(Reprinted with permission of ASM International. All rights reserved)
68
CHAPTER 3
3.1
INTRODUCTION
Exfoliation corrosion (EFC) is a particular form of intergranular corrosion that is
often observed on the surface of wrought aluminum alloys with an elongated grain
structure.[1] When intergranular corrosion proceeds along intergranular paths parallel to
the surface, the hydrated aluminum oxide corrosion product has a higher volume than the
alloy matrix from which the product formed. Consequently, relatively large wedging
stresses are produced, which lift the surface grains, giving rise to a layered appearance.
EFC is significant source of life-limiting degradation in airframes.[2] It consumes load
bearing cross section and increases the stress in the remaining intact material, resulting in
a loss in mechanical properties of Al alloys.
The susceptibility of Al alloys to EFC can be assessed by exposure testing.
ASTM G-34, known as the EXCO test, involves exposure to an oxidizing acidic chloride
solution and comparison of the resulting surface to standard photographs.[1, 3] Other
69
tests involving exposure to aggressive environments have been devised, including the
ASSET and MASTMAASIS tests.
sprayed onto the surface of the specimens and the resulting surfaces are evaluated to
determine the EFC susceptibility.[4-7] The behavior of Al alloys in these accelerated
environments has been correlated to long-term exposure in less-aggressive natural
environments.[8-10] As a result, these tests are very useful for assessing susceptibility to
EFC attack. However, they do not provide quantitative measurements of susceptibility or
growth kinetics, which are required for predictive modeling of corrosion development.
Liddiard and coworkers have used a deflection technique to quantify exfoliation extent
and determine EFC kinetics.[11] In this technique, the effective remaining load-bearing
section of specimens having undergone EFC was determined from their compliance
under four-point bending. The rate of EFC can be assessed from periodic measurements.
The deflection technique is valid only when the thinning of the specimen during
corrosion is uniform.
The alloy grain shape is thought to be an important factor in exfoliation corrosion.
Robinson et al. found that the severity of exfoliation corrosion is related to the grain
aspect ratio of the material.[12-14] More elongated grain shape results in severe
exfoliation corrosion. Besides elongated grain structure, grain boundary chemistry also
plays an important role in the EFC susceptibility. It is well known that grain boundaries
are often more susceptible to corrosion than the grain interiors because of the
microstructrual heterogeneity.
contain precipitates on the grain boundary and a precipitate free zone (PFZ) next to the
grain boundary, which have very different electrochemical behaviors than the grain
70
interior region.[15] Maitra and English have attributed the IGC susceptibility of AA7075T6 to Mg and Zn solute segregation or enrichment in the grain boundary region.[16]
Ramgogal et al. proposed that IGC in 7075-T6 temper was caused by anodic dissolution
of Mg(ZnCuAl)2.[15, 17] Other investigations have suggested that the cause of
exfoliation is an electrochemical potential developed between primary or secondary
particles ,which contain some or all of the elements Al, Fe, Si, Mn, Mg, and Zn, and the
surrounding AlZnMg solid solution.[18, 19]
The Mn
bearing particles precipitate preferentially in the center of the grains compared to the
region near grain boundaries, resulting in differing electrochemical potential.[5]
Determination of EFC kinetics is critical for the development of predictive failure
models. The aim of this work was to develop a technique and analysis methodology that
can provide quantitative measurement of exfoliation corrosion kinetics. The EFC rates of
different AA7178 wingskin plates were determined and related to the alloy
microstructure and compositions of the grain boundary constituents determined from
analytical TEM techniques.
3.2
EXPERIMENTAL
3.2.1
Materials
underlying support beam by steel rivets. The wingskin section was divided into two parts.
One part of the wingkin section was exposed in the uncoated condition at an atmospheric
71
exposure test site near Daytona Beach, FL by W. Abbott of Battelle. The other part of
the section was used to perform laboratory testing. Figure 3.1 shows the section after 9
months exposure at Daytona Beach. The two plates in the section exhibited vastly
different exfoliation behavior during the exposure near the seacoast. One of the plates
exfoliated badly next to the steel rivets and the other plate only developed cosmetic
surface attack. They are referred to as good and bad plates in reference to their EFC
susceptibility.
Samples were machined from the good and bad plates in the shape of
rectangular slices, as illustrated in figure 3.2. The slices were 3-4 cm long, and oriented
such that the long axis of each slice was in the longitudinal orientation of the
microstructure (along the rolling direction). The slice thickness was around 1 mm, and
the slice thickness was oriented in the plate transverse direction. The width of each slice
was the full plate through-thickness in the short transverse direction: 4.1 and 4.6 mm for
the good and bad plates, respectively. The slice edges, which were the original outer
surfaces of the plate, were lightly polished. All other faces were ground in ethanol to 800
grit finish, cleaned ultrasonically in ethanol, and finally dried by an air stream.
Detailed metallographic analysis was performed on T (long transverse), S (short
transverse), and L (longitudinal) sections of the good and bad plates. Samples were
polished to 1 micron and chemically etched with Kellers reagent (2mL HF, 3mL HCl,
5mL HNO3 and 190 mL H2O) to reveal the microstructure. Optical microscopy methods
were used to examine metallographic sections in three perpendicular orientations. The
grain dimension in three orientations as a function of the position along through thickness
direction was determined with Clemex Vision image analysis software. The through
72
measured and the mean values were calculated to represent the grain size in each zone.
Analytical transmission electron microscopy (TEM) was performed to investigate the
local chemistry at the grain boundaries. A 10 mm long sample was cut from a T section
of the good plate. The width of the sample was the whole thickness of the plate. The
sample surface was polished to 0.5 m by alumina suspended in ethanol and slightly
etched with acid reagent (6mL HF, 40mL H2SO4, 360 mL H2O) for 20 seconds to reveal
the grain boundaries on the surface. Cross sectional TEM samples were prepared by an
FEI Strata Dual Beam 235M scanning electron microscope (SEM) / focused ion beam
(FIB) tool using a 30 keV Ga ion beam and a 5 keV electron beam. The TEM samples
were made such that they contained a grain boundary located 10~30 m away from the
surface. TEM samples were made close to the upside surface (corresponding to the
surface that was exposed to the Daytona Beach environment) and close to the downside
surface (corresponding to the surface that was not exposed to the Daytona Beach
environment).
Prior to sectioning, specific sites near the specimen edge were located
and identified by SEM/EDS in the FIB and then covered by deposition of a 1.5 m-thick
Pt layer to protect the specimen surface during the FIB sectioning. The membrane had an
area of 15 m x 5 m and was thinned in the FIB to a thickness of 100 nm for electron
transparency.
The membrane was plucked out of the bulk sample under an optical
microscope using a sharp Pyrex needle of 1 um in diameter and placed on a 200 mesh Au
TEM grid with a formvar/ carbon support film for TEM analysis. The TEM observation
was made in the region of the membrane above a hole in the mesh grid. TEM
characterization of the FIB sectioned membranes was conducted using an FEI Tecnai
73
TF20 scanning transmission electron microscope (STEM) operating at 200 kV. The
probe size was <2 nm and the step size was 5 nm.
Electrochemical polarization measurements were performed on samples ground to
1200 grit. 1 M NaCl solution was deareated with Ar gas to decrease the corrosion
potential and allow for clear observation of the breakdown potentials. Potentiodynamic
scans were performed at a rate of 0.1 mV/s. A Pt counter electrode and saturated calomel
reference (SCE) were used. All potentials in this paper are referenced vs. SCE.
3.2.2
Samples sliced from the plates as described above were given a potentiostatic
electrochemical pretreatment in 1 M NaCl at a potential of 710 mV SCE for 7 h. This
potential is above the second breakdown potential for this material in this environment,
which is -725 mV SCE as will be shown below. The purpose of the pretreatment was to
initiate localized corrosion attack and develop an aggressive environment in the corrosion
sites. Following the pretreatment, the sample was rinsed with DI water and placed in a
humidity chamber, consisting of a sealed beaker containing a saturated salt solution at
room temperature (22-25C). The samples were held in the air space above the solution
and a graph paper was placed behind them to facilitate determination of EFC extent.
Different saturated salt solutions were used to create a range of constant humidity:
sodium sulfate (Na2SO4), ammonium chloride (NH4Cl), potassium iodide (KI), potassium
carbonate (K2CO32H2O), and calcium chloride (CaCl26H2O). The relative humidities
above saturated solutions of these salts measured by an RH meter were similar to
handbook values[20], as is reported in Table 3.1. Pretreated samples developed EFC
74
corresponding to the original plate surfaces, and moved inward. Images of the samples
were recorded by digital photography through the glass walls of the humidity chamber.
The contrast between the boundary of the outer exfoliated and inner unattacked regions
was sufficient to allow tracking of the EFC kinetics by analysis of the photographs.
The width change of the inner unattacked region of an ESH sample can be related to
the material loss due to exfoliation corrosion. The photographs of the ESH samples were
analyzed to measure the unattacked width as a function of time. One measurement of the
unattacked width was made on each sample at a position 10 mm from the top of the
sample.
This value was subtracted from the original width at this position, d0, to
3.3
Potentiodynamic polarization curves of the good and bad plates were obtained in
deareated 1 M NaCl at a potential sweep rate of 0.1 mVs-1, figure 3.3. Both plates
exhibited two breakdown potentials as has been observed for AA7xxx alloys in the
underaged and peak aged conditions.[21] The values of the breakdown potentials are
similar for the two plates, so the difference in exfoliation susceptibility cannot be
attributed to a difference in breakdown potentials. There was, however, a difference in
75
the magnitude of both the transient peak above the first breakdown potential and the
current flowing at higher potentials; both were smaller for the good plate. This reflects
some differences in the microstructure.
Samples of the good and bad plates from the AA7178 wingskin section were
ESH-tested by electrochemical pretreatment and then exposure to 96% RH. Figure 3.4
shows images of the samples over a period of time from 0 (as-pretreated) to 52 days of
high humidity exposure. The samples are oriented such that the rolling direction is
vertical and the plate thickness is horizontal, with the original plate surfaces on the left
and right edges of the samples. After several days, EFC was evident on both edges of the
slice from the bad plate and it continued to progress more or less evenly on both edges
over the exposure period of almost 2 months, figure 3.4a. The behavior of the slice
removed from the good plate behaved quite differently. One side of the sample (the
left side of the images in figure 3.4b) exhibited severe EFC, while the other side of the
sample was practically unattacked. The side of the slice that was not attacked in the ESH
test corresponds to the side of the good plate that was exposed upwards to the
environment at Daytona Beach. So the high resistance of the good plate to EFC was
reproduced in the lab in that the surface exposed at Daytona Beach reacted slowly in the
ESH test. This resistance was only representative of one surface; the other side of the
good sample corresponding to the side of the plate that faced downward and was not
exposed to the elements at Daytona Beach, showed rapid attack in the ESH test. It is
expected that the good plate would have exfoliated rapidly at Daytona Beach if the
plate had been oriented such that the down side had been facing upward. Overall, the
bad sample was attacked much faster than the good sample.
76
As seen in figure 3.4, corrosion product exuded out of the surfaces (corresponding
to transverse face of the microstructure) of both good and bad samples upon exposure to
96% RH. This generation of corrosion product is evidence of continued corrosion at
localized sites from the aggressive environment deposited in the microstructure during
the electrochemical pretreatment. However, the EFC proceeded inward from the edges.
The kinetics of EFC can be determined by measuring the change in width of the
central unexfoliated region. Figure 3.5 shows the average exfoliation depth, which is
equal to half of the change in width of the inner unexfoliated region. The good sample
exfoliated primarily on one side, so the real EFC depth for the attacked side of the good
sample is approximately twice the value given in the figure 3.5. Considering this fact, the
depth of EFC on the susceptible side of the good sample was approximately the same
as the depth on both sides of the bad sample, which is evident in figure 3.4. Figure 3.5
shows that the rate of EFC was not constant with time. For 2 samples, it started out
rapidly, with a change of about 80 m in the first day and then slowed to about 4 m/day.
The other two samples did not experience the initial rapid EFC and started off with a
slow rate. It is interesting that the two good and two bad samples did not behave the
same at the beginning. One of each type started out quickly and one started slowly.
After about 2 weeks, the EFC rate for the bad samples and one of the good samples
increased to a higher value, about 20 m/day. The rate for each of these samples then
decreased again to a lower value of about 2 m/day. The other good sample (the one
shown in figure 3.4b) did not exhibit an increase again until after 35 days. These results
indicate that there are zones of the plate through the thickness with varying susceptibility
and EFC kinetics. The apparent resistance of the good plate during outdoor exposure
77
at Daytona Beach was caused by the seemingly chance location of a resistant zone at the
upward-facing surface of that plate.
susceptibility exist through the thickness of a plate. This has never been reported, but
other techniques, such as the EXCO tests, do not have the sensitivity to find these zones.
This varying susceptibility could be caused by variations in microstructure or residual
stress in the plate.
ESH testing was performed on good and bad samples under a range of constant
humidity, figure 3.6. The two plates exhibited different exfoliation behavior in different
humidities. In the lowest humidity (30%, CaCl2), the surfaces of both samples were still
shiny after 60 days exposure, and no EFC was evident. This humidity is apparently
below the critical humidity of the environment deposited in the localized corrosion sites
created by pretreatment. This environment dried completely, preventing further attack.
In about 50% RH, some product exuded out of the surface and EFC was seen on the
downside of the "good" sample after 3 days. However, the EFC ceased after that time,
perhaps because it took some time for the environment in the pretreatment sites to
equilibrate with the humidity in the chamber and dry out. Above around 50% RH,
samples exfoliated more severely with increasing humidity. The average EFC rate was
3.0, 4.0, and 14.0 m/day for the bad samples in 66, 76, and 96% RH, respectively.
For good samples, the values were 0.6, 0.8, and 5.0 m/day, respectively, for the same
humidities. The effect of humidity and other factors on EFC rate will be covered in detail
in a subsequent communication.[22] However, it is clear from these data that the rate of
EFC increases with increasing humidity above a critical humidity.
78
3.3.2
Optical Microscopy
The microstructures of the good and bad plates are given in figure 3.7, along with
the convention utilized for the various sections. Figure 3.8 is a montage of highermagnification micrographs of the T sections of the two plates covering the full cross
section from one side of the plate to the other. Each section is given in two parts, with
some overlap of the top left and bottom right. The grains in the bad plate are larger and
more elongated in the L direction than in the good plate. In particular, the near-surface
regions of the good plate exhibit much smaller grains with a smaller aspect ratio. The
grain sizes of the good and bad plates as a function of the position through thickness
direction were obtained by the analysis of L, T, and S cross-sections and are shown in
figure 3.9. The grain aspect ratios in longitudinal and transverse orientations are reported
in figure 3.9 with the L/S ratio shown above the T/S ratio. The bad plate has a higher
grain aspect ratio than the good plate. In particular, the size of grains in the L direction
is about 3 times greater in the bad plate. It is known that a microstructure with grains
that are less-elongated should be more resistant to exfoliation (but should exhibit faster
IGC in the S direction). The larger grain size and higher grain aspect ratio of the bad
plate contribute to its higher EFC susceptibility. There is a small difference in the grain
aspect ratio near the upside and downside of the good sample. However, it is unlikely
that the slightly smaller grain aspect ratio at the upside is the sole explanation for the
much better EFC resistance there. Furthermore, the grain size and grain aspect ratio near
the downside of the good sample are much smaller than those of the bad sample, yet
the EFC rate was similar. Therefore, factors other than grain size and shape must play a
79
determining role in the EFC susceptibility. Possibilities include residual stress in the
plate and local variations in grain boundary microchemistry.
3.3.3
TEM Analysis
of precipitates are found along grain boundaries near both upside and downside of the
plate.
precipitates are shown in figure 3.13a near upside of the plate and in figure 3.13c near the
downside of the plate) and the other is Cu-Al precipitate (concentration profiles across
these precipitates are shown in figure 3.13b near upside of the plate and figure 3.13d near
the downside of the plate). There are no obvious differences in the precipitates near the
up- and downsides of the plate. Furthermore, it is unlikely that a difference in the
composition of grain boundary precipitates can explain the difference in EFC behavior
because the precipitates are isolated along the boundaries in this material, i.e. they do not
form a continuous pathway on the grain boundaries.
The concentration profiles in the form of concentration ratios for the line-scans in
figure 3.12 across PFZ regions are shown in figure 3.14. Zn is seen to be depleted in the
PFZ near the grain boundary at the upside surface relative to the nearby lattice, figure
3.14a. In contrast, there is no apparent Zn depletion in the PFZ near the grain boundary
at the downside surface, figure 3.14b. As noted above, only a limited number of different
grain boundaries were analyzed (five near the upside and five near the downside). It is
always possible to question the validity of using a limited number of very local TEM
measurements to explain bulk properties. Nonetheless, the profiles shown in figure 3.14
are representative of the regions studied, and a difference in Zn concentration in the PFZ
regions near the upside and downside was clearly evident.
It is known that Zn, the main alloying element in 7xxx alloys, diffuses readily at
normal homogenizing temperatures.
solubility in Al, resulting in the formation of Mg-Zn containing phases along grain
81
boundaries and the depletion of Mg and Zn in the area adjacent to the grain
boundaries.[5] Park and Ardell have used analytical transmission electron microscopy to
show that the grain boundary region in AA7150-T6 exhibited considerable Zn depletion
relative to the matrix.[23] Many studies have shown that the alloying of aluminum with
Zn reduces the pitting potential and repassivation potentials.[24-27] Ramgopal et al.
investigated the dissolution kinetics of binary Al alloys containing Cu, Zn and Mg in
chloride solution and found out Zn addition decreased the repassivation potential by
increasing the dissolution kinetics at a given potential. [28]
The TEM results in this study suggest that the difference in Zn concentration in
the PFZs near the upside and downside surfaces of the good plate is at least partly
responsible for the difference in EFC behavior. This different distribution of Zn might be
caused by details of the plate processing. The Cr and Cu enrichment in GB precipitates
results in depletion of these two elements in grain boundary area. The PFZ is anodic to
the rest of the grain and is preferentially attacked. In PFZs near the upside surface of this
plate, the absence of solute Zn increased the corrosion resistance of the PFZ and reduced
the susceptibility to IGC and thus EFC. On the other hand, the presence of Zn in the PFZ
near the downside surface sharpened the electrochemical heterogeneity between the grain
boundaries and the adjacent regions and hence enhanced EFC. The grain boundary
chemistry seems to be an important factor controlling the susceptibility to EFC.
3.4
CONCLUSIONS
1.
The Exfoliation of Slices in Humidity (ESH) test was used to assess the
susceptibility of AA7178 samples to exfoliation corrosion (EFC) and to determine
82
the rate of EFC. In this test, samples sliced in a particular orientation are given an
electrochemical pretreatment and then exposed in a constant humidity
environment.
2.
The ESH test was able to reproduce exfoliation behavior of plates observed
during outdoor exposure.
3.
EFC was not observed below 50% RH and the EFC kinetics increased with
increasing humidity above this critical humidity.
4.
EFC susceptibility was found to depend on grain size and shape as well as grain
boundary composition.
5.
83
REFERENCES
"Standard Test Method for Exfoliation Corrosion Susceptibility in 2xxx and 7xxx
Series Aluminum Alloys (EXCO Test)", in Annual Book of ASTM Standards,
Philadelphia, PA: The American Society for Testing and Materials, 114-119
(1990).
J.P. Chubb, T.A. Morad, B.S. Hockenhull, and J.W. Bristow, International Journal
of Fatigue, 17(1), 49-54 (1995).
S. Lee and B.W. Lifka, "Modification of the EXCO Test Method for Exfoliation
Corrosion Susceptibility in 7XXX, 2XXX and Aluminum-Lithium Alloys", in New
Methods for Corrosion Testing of Aluminum Alloys, ASTM STP 1134,
V.S.Agarwala and G.M.Ugiansky, Philadelplia: American Society for Testing and
Materials, 1-19 (1992).
D.G. Evans and P.W. Jeffrey, Exfoliation Corrosion of AlZnMg Alloys, in U.R.
Evans Conference on Localized Corrosion, NACE-3, 1974, Houston, TX: Alcan
International Limited Research Center.
G.S. Haynes and R. Baboian, "Modified Salt Spray (Fog) Testing, Laboratory
Corrosion Tests and Standards", in ASTM Standard G85-85 A2. STP866,
Philadelphia, PA: ASTM, 578-583 (1985).
D.O. Sprowls, J.D. Walsh, and M.B. Shumaker, "Simplified Exfoliation Testing of
Aluminum Alloys", in Localized Corrosion-Cause of Metal Failure, ASTM STP
516, American Society for Testing and Materials, 38-65 (1972).
D.O. Sprowls, T.J. Summerson, and F.E. Loftin, "Exfoliation Corrosion Testing
of 7075 and 7178 Aluminum Alloys-Interim Report on Atmospheric Exposure
Tests", in American Society for Testing and Materials, Philadelphia, PA: 99-112
(1973).
10
B.W. Lifka and D.O. Sprowls, "Relationship of Accelerated Test Methods for
Exfoliation Resistance in 7xxx Series Aluminum Alloys with Exposure to a
Seacoast Atmosphere", in ASTM SPEICIAL TECHNICAL PUBLUCATION 558,
306-333 (1973).
84
11
E.A.G. Liddiard, J.A. Whittaker, and H.K. Farmery, Journal of the Institute of
Metals, 89, 377-384 (1960-1961).
12
M.J. Robinson and N.C. Jackson, Corrosion Science, 41, 1013-1028 (1999).
13
M.J. Robinson and N.C. Jackson, British Corrosion Journal, 34(1), 45-49 (1999).
14
15
T. Ramgopal, P.I. Gouma, and G.S. Frankel, Corrosion, 58(8), 687-697 (2002).
16
17
18
19
G. Bassi and J.J. Theler, Zeitschrift fur Metallkunde, 60(3), 179-184 (1969).
20
R.C. Weast, M.J. Astle and W.H. Beyer, CRC Handbook of Chemistry and
Physics, Boca Raton, FL: CRC Press, Inc. (1987).
21
22
23
J.K. Park and A.J. Ardell, Acta Metallurgica et Materialia, 39(4), 591-598 (1991).
24
I.L. Muller and J.R. Galvele, Corrosion Science, 17, 995-1007 (1977).
25
26
27
P.L. Bonora, G.P. Ponzano, and V. Lorenzelli, British Corrosion Journal, 9(2),
108-111 (1974).
28
85
Salt
Measured %RH at RT
Na2SO4
93 at 20oC
96
NH4Cl
79.3 at 20oC
76.5
KI
56.2 at 100oC
65.1
K2CO32H2O
43 at 24.5oC
49~50
CaCl26H2O
32.3 at 20oC
30.1
Table 3. 1
86
bad
good
Figure 3.1
Photo of AA7178 wingskin sample after 9 months of exposure at Daytona
Beach. Sample was uncoated and had steel rivets attaching plates to understructure.
Image provided by W. Abbott, Battelle.
Figure 3.2
Schematic drawing of orientation of slices for ESH test relative to
elongated microstructure of AA7178 wingskin sample.
87
-600
"good" plate
-700
-800
-900
-1000
"bad" plate
-1100
-1200 -11
10
10
-9
10
-7
10
-5
0.001
0.1
88
4.07mm
0day
7days
15days
27days
52days
(a)
3.86mm
0day
2days
6days
29days
52days
(b)
Figure 3.4
Images of AA7178 wingskin samples exposed to 96% humidity following
electrochemical pretreatment. (a) Bad sample. (b) Good sample.
89
0.5
0.4
"bad" samples
0.3
"good"
samples
0.2
0.1
0
0
10
20
30
40
50
60
Time (days)
Figure 3.5
Average exfoliation corrosion depth for duplicate samples of good
(triangles) and bad (circles) AA7178 wingskin samples exposed to 96% RH.
90
0.5
"good" "bad"
96% RH
0.4
76% RH
65% RH
0.3
0.2
0.1
0
0
10
15
20
25
30
35
Time (days)
Figure 3.6
Average exfoliation depths of good (open symbols) and bad (closed
symbols AA7178 wingskin samples exposed to different humidities: circles 96%,
squares 76%, triangles 65%.
91
Figure 3.7
Metallographic sections of good plate and bad plate. Also given is the
terminology used for the different sections. (a) bad plate. (b) good plate.
92
0.2 mm
a.
UP
SIDE
b.
Figure 3.8
Metallographic sections of AA7178 wingskin plate. (a) Bad plate. (b)
Good plate. The sections are through-thickness montages, starting at the right of the
top image in each pair and then wrapping around to end at the left side of the bottom
image in each pair.
93
700
S
600
T
L
500
400
L/S
300
8.6
7.5 7.2
8.7
6.4 6.1 7.2 6.6 7.7
7.8
8.2
6.5
7.1 7.2 6.5
7.0 7.8
7.3 6.6
6.3 5.7 5.7
6.6
4.7 5.4
5.0 6.1
5.1
5.8
4.8
200
100
0
T1/15
T8/15
upside
T15/15
downside
600
T
L
500
400
300
16.2
8.7
17.0
17.2
7.4 6.8
200
100
0
T1/15
T8/15
upside
T15/15
downside
(a)
(b)
Figure 3.10 TEM micrographs of grain-boundary in AA7178 good plate. (a) Near up
side of the good plate. (b) Near down side of the good plate.
95
(a)
(b)
96
(c)
(d)
Figure 3. 11 TEM micrographs of grain-boundary in AA7178 showing Nano EDS line
profiling across two different types of the GB precipitates (a,b) near up side of the good
plate and (c,d) near down side of the good plate.
97
(a)
(b)
Figure 3.12 TEM micrographs of grain-boundary in AA7178 showing Nano EDS line
profiling across the GB PFZ (a) Near up side of the good plate. (b) Near down side of
the good plate.
98
0.25
CZn/Al
CMg/Al
CCu/Al
CCr/Al
CFe/Al
CTi/Al
CSi/Al
CMn/Al
0.2
0.15
Zn
0.1
Cr
0.05
Mg
Cu
0
50
100
150
200
250
Position (nm)
Concentration Ratio Profile (C i/CAl)
(a)
0.7
CZn/Al
CMg/Al
CCu/Al
CFe/Al
CCr/Al
CSi/Al
CTi/Al
CMn/Al
Cu
0.6
0.5
0.4
0.3
0.2
Mg
0.1
Zn
0
50
100
150
Position (nm)
(b)
99
200
250
0.25
Zn/Al
Mg/Al
Cu/Al
Cr/Al
Fe/Al
Ti/Al
Si/Al
Mn/Al
Zn
0.2
0.15
0.1
Cr
0.05
0
Cu
Mg
0
50
100
150
200
250
Position (nm)
Concentration Ratio Profile (C i/CAl)
(c)
0.7
0.6
CZn/Al
CMg/Al
CCu/Al
CFe/Al
CCr/Al
Csi/Al
CTi/Al
CMn/Al
Cu
0.5
0.4
0.3
0.2
Zn
0.1
Mg
0
50
100
150
200
250
Position (nm)
(d)
Figure 3.13 Nano-EDS line profile of grain boundary precipitates. (a, b) linescans
across grain boundary precipitates in figure 3.11a and b near upside of the good plate.
(c ,d) ) linescans across grain boundary precipitates in figure 3.11c and d near downside
of the good plate. The data are reported as ratio of the x-ray intensities for each
element relative to that of Al.
100
0.12
CZn/Al
CMg/Al
CCu/Al
CCr/Al
CFe/Al
CTi/Al
CSi/Al
CMn/Al
PFZ
0.1
0.08
Zn
0.06
0.04
Mg
0.02
0
Cu
50
100
150
200
250
Position (nm)
Concentration Ratio Profile (C i/CAl)
(a)
0.3
0.25
CZn/Al
CMg/Al
CCu/Al
CFe/Al
CCr/Al
CSi/Al
CTi/Al
CMn/Al
PFZ
0.2
0.15
Zn
0.1
Mg
0.05
Cu
0
50
100
150
200
250
Position (nm)
(b)
Figure 3.14 Nano-EDS line profile of solute depleted zone around grain boundary. The
data are reported as ratio of the x-ray intensities for each element relative to that of Al. (a)
Near up side of the good plate (shown in figure 3.12a). (b) Near down side of the
good plate (shown in figure 3.12b).
101
CHAPTER 4
4.1
INTRODUCTION
Exfoliation corrosion (EFC) is a particular kind of intergranular corrosion often
found in the wrought Al alloys (AA2024 and AA7178) used in wingskins and fuselages
of airplanes.[1] It proceeds along elongated grain paths parallel to the surface. The
internal stresses caused by voluminous corrosion products force metal grains away from
the bulk giving rise to a layered appearance. EFC usually does not directly lead to failure
of Al components on aircraft, but it can decrease the load-bearing cross section and lead
to stress corrosion cracking or corrosion fatigue.[2, 3]
A great deal of work has been devoted to address the importance of elongated
grain structure on EFC propagation.[4, 5] However, a lack of quantitative understanding
of the environment and applied stress conditions hinders the development of predictive
models and protection schemes to prevent EFC. To assure aircraft structural integrity, it
102
4.2
EXPERIMENTAL
An AA7178 plate of thickness 0.89 cm cut from the wing of a decommissioned
airplane was studied. The nominal composition of AA7178 (in wt %) was determined by
Wavelength Dispersive X-ray Fluorescence (at the OSU Department of Geological
Sciences) and is shown in table 4.1. Samples were machined from the plates in the shape
of rectangular slices. The slices were oriented such that the long axis of each slice was in
103
The
sample thickness, oriented in the plate transverse direction, was around 2-3 mm. The
width of each sample was the full plate through-thickness in the short transverse direction.
The sample edges, which were the original outer surfaces of the plate, were lightly
polished to remove the surface coating. All other faces were ground to 800 grit in
ethanol, cleaned ultrasonically in ethanol, and finally dried by an air stream.
Samples were primarily tested in the as-received condition. The original temper
of the AA7178 plate was probably T6, but the natural aging over the decades of service
might have altered the microstructure and properties. To test the effect of temper on EFC
susceptibility, the as-received AA 7178 plates were re-heat-treated to the T6 and T7
tempers.[16] Details of the heat treatments are summarized in Figure 4.1.
T, S, L
initiate corrosion attack in the samples.[17] The applied potential was a relatively high
value above the second breakdown potential. However, it was subsequently found that
the results of the ESH were sometimes quite scattered using a potentiostatic pretreatment
because this method can create different amounts of pre-corrosion if the samples have
slightly different electrochemical properties. Figure 4.2 shows the current measured
during a potentiostatic experiment at -710 mV SCE in 1 M NaCl on two nominally
identical AA7178 samples with the same exposed area. The current measured, the charge
passed, and thus the damage created, were different for the two samples. In this work, a
galvanostatic pretreatment was used to obtain more reproducible results.
The
photographs were taken of the T face of the samples. The EFC started at the outer edges
of the samples, corresponding to the original plate faces, and moved inward.
The
contrast at the boundary of the outer exfoliated and inner unattacked regions was
sufficient to allow tracking of the EFC kinetics by analysis of digital photographs of the
sample taken through the glass walls of the humidity chamber. The EFC kinetics can be
determined by measuring the width change of central unattacked region of the sample. In
this work, the width of the unattacked region was measured at 30 evenly distributed
positions along the long axis of the sample. From these measurements, the average width
change as a function of time were determined from d0j dij = dij, i = 1,2,3days, j =
30
d
j =1
ij
30
, i = 1,2,3.......days .
Since the exfoliation occurred at two surfaces, one half of the width change
d i
2
106
4.2.1
Pa
(3Lx 3 x 2 a 2 )
6 EI
(1)
where is the deflection in the y direction, x is the distance from the support O
( a x L a ), L is the separation of the stressing points on the tensile side and a is the
separation distance of the opposing loading points (Figure 4.4). The maximum normal
stress in the specimen is given by
max
h
My( )
2 = Pah
=
I
2I
(2)
where M y is the bending moment in the y direction; I is the moment of inertia, and h is
the width of the specimen.[19] Since the bending moment between the two loading points
A and B is constant, the maximum normal stress is also constant in this region.
Considering the deflection at point A or B, where x=a or x=L-a, equation (1) becomes:
Pa
(3La 4a 2 )
6 EI
(3)
Solving equation (2) for Pa/I and substituting into equation (3), the relationship between
deflection and maximum stress is obtained:
max
3hE
(3La 4a 2 )
(4)
107
Assuming elastic behavior and substituting max = E max into equation (4), the initial
outer fiber strain of the specimen, max, can be determined from the specimen geometry
and deflection in y direction, :
max =
3h
3La 4a 2
(5)
A four-point bend test frame was constructed from stainless steel as shown in
figure 4.5. Ceramic rods were used to apply stress to the sample, and a dial gauge was
attached to the screw used to adjust the separation of the load frame plates.
One
revolution of the screw resulted in a 1.27 mm displacement of the plates and the accuracy
of the displacement based on the dial gauge markings, was 1 or 3.53 m. The samples
were cut from the AA7178 plate so that the length (about 60 mm) was along the L or
rolling direction, the width was the full thickness of the plate in the S or through
thickness direction (about 7.5 mm), and the thickness was 2 mm in the transverse
direction. The surfaces of the original plate were coated, one side with gray paint and the
other side with a chromate conversion coating. These coatings were removed from the
samples by light polishing prior to the test, and the identities of the two sides of the
original plate were tracked through the tests. The other surfaces were ground in alcohol
to 1200 grit. The samples were given a galvanostatic pretreatment as described above,
rinsed with DI water, and placed in the frame so that the bending axis was along the short
transverse axis of the plate. The bending stress was applied by displacing the rod support
plates by an amount of using the dial gauge as a monitor. The maximum strain, max,
was determined from equation (5) where h was approximately 7.5 mm, L was 63.5 mm,
and a = (L L)/2 where L, the separation of stressing points on the compressive side,
108
was 31.75 mm, so a = 15.875 mm. Previously reported values for the modulus of
elasticity, E, and yield strength, Y, in compression and tension for AA7178-T6 were
used: 71 GPa, 530 MPa, and 540 MPa, respectively.[20, 21] Using Hookes Law, the
yield strain in compression and tension was calculated as 0.0075 and 0.0076, respectively.
The applied maximum strains were selected to be 0.004, 0.005, and 0.006 to stay in
the elastic region. The loaded frame was placed in a 96% RH chamber so that the sample
length was vertical, similar to the unstressed ESH tests, as shown in figure 4.6. Digital
photographs of the T face were taken every other day through the wall of the glass
chamber.
Because one side of the stressed sample was in tension and the other side was in
compression, they exfoliated at different rates. It was therefore necessary to distinguish
the EFC rate on each side, rather than use the average rate calculated from the change in
width of the inner unattacked region. The sample was surrounded by mm scale during
the RH exposure. The inward movement of the boundary between exfoliated region on
each side and central unattacked region could be tracked by the horizontal scale giving
the exfoliation rate on each side.
4.3
Samples cut from the as-received AA7178 wingskin plate were ESH tested in a range
of RH.
the same sample was cross-sectioned using a focused ion beam (FIB) tool. As shown in
figure 4.8, intergranular attack can be seen in the underlying region in the FIB-etched
section.
Figure 4.9 shows photographs of samples in 56, 66, 76.6, and 96 %RH on the
11th day of exposure. There is a clear effect of RH on EFC behavior. Below 56% RH,
samples did not exfoliate at all and the surface remained shiny and unattacked. Figure
4.10 shows a sample exposed to 30%RH at room temperature for 13 days; no change at
all was observed. The sample exposed at 56 %RH did not exhibit sustained EFC. Only a
small amount of corrosion product was observed on the sample surface. The sample
exhibited some initiation of EFC at the top edge during the first several days of exposure,
but the exfoliation stopped growing and did not progress for the rest of the exposure time.
The critical RH for EFC appears to be close to 56%. As the RH increased above this
critical humidity, the extent of EFC increased and more corrosion product was found to
decorate the transverse surface. The width change of central unexfoliated region was
determined as a function of time in various constant humidity environments, and the
average exfoliation depth, equal to half of the width change, is shown in Figure 4.11.
Replicate tests were performed for each humidity, and the average values from each test
are shown as symbols.
galvanostatically. The lines in Figure 4.11 represent average values from the replicate
runs. Two sets of error bars are shown in Figure 4.11, representing the standard error for
each measurement and the standard error in the whole population. Clearly, the EFC rate
increased with RH.
110
Prior work on other AA7178 plates indicated that the rate of EFC, given by the
slope of the EFC depth curve, did not always vary smoothly.[17] In those plates, the rate
increased and decreased with time as the EFC propagated through regions of the plates
with varying susceptibility. The plate used in the current study seems to be more uniform,
and the rate of EFC was rather constant as the attack proceeded into the plate.
ESH tests in 96%, 76%, 66% and 56%RH were continued until 100 days of
exposure. The samples were removed from the humidity chambers and the sample
surfaces (T face) were observed by SEM. For the sample exposed to 56%RH, most
corrosion sites formed during the pretreatment were free from corrosion product.
However, the sample exposed to 66%RH clearly shows corrosion product exuding out of
the initiation sites, Figure 4.12a, and the amount of corrosion product on the sample
surface increased with increasing RH. The sample exposed to 76%RH exhibited a high
density of corrosion product and the surface was rather rough in appearance, Figure 4.12b.
The sample exposed to 96 % RH exhibited fewer but larger corrosion product aggregates
that sat on top of what appears to be a uniform layer of corrosion product as shown in
Figure 4.12c.
EDS (shown in figure 4.13) and XPS analysis of an AA7178 sample exposed to
56%RH (shown in figure 4.14) shows that the corrosion product consisted of Al, O and
Cl. Note that the sample was rinsed with DI water after pretreatment, so the Cl was not a
remnant of the NaCl pretreatment solution. The EDS and XPS results indicate that the
corrosion product formed during pretreatment, or dried to form hydrated aluminum
chloride or aluminum oxychloride. The detailed results obtained from analysis of the
XPS spectra are listed in table 4.2.
aluminum chloride Al(OH)3-xClx has been reported to be in the range between 40 95%RH, depending on the degree of hydrolysis.[23] This is consistent with our
observation that exfoliation corrosion only propagated at humidity above 56%RH. At
lower humidities, the localized corrosion environment generated by the pretreatment
completely dried, preventing further attack.
In real exposure environments, the RH is not constant with time and might vary
above and below the critical humidity. It is of interest to know if EFC will continue to
propagate when a sample is moved from a high to a low RH environment and, if after
drying out at low RH, it can rehydrate and resume EFC at the same rate as exhibited
previously.
Figure 4.15 shows the results of a cyclic RH ESH test where the RH was switched
between 96% and 56% by quickly moving the sample from one chamber to another. EFC
initiated as usual during the first exposure to the 96%RH environment. After 10 days, the
sample was moved to 56%RH. The exfoliation continued for one or two days and then
almost stopped growing for the rest of exposure period of 20 days. When the sample was
moved back to high humidity again, exfoliation resumed very quickly and developed
with the same kinetics as in a continuous 96% RH exposure test. In fact, during the test
shown in figure 4.15, the sample was exposed to 96% RH for a total of 34.5 days and the
total amount of EFC, about 0.25 mm, is equal to the amount of EFC developed during an
exposure to a constant RH of 96% for that same period.
controlled by the total exposure time at high RH.
environment quickly equilibrates with the exterior environment, which is very different
from the behavior of sharp intergranular corrosion fissures that form in this alloy under
112
similar conditions.[22] The speed of the equilibration of the local and exterior
environments might be partly the result of the configuration of the samples. The throughplate slices exposed the entire transverse surface to the environment, whereas a real
component is primarily exposed at the S surface. Nonetheless, it is interesting that the
corrosion product in the EFC cracks responsible for creating the wedging stress that
propagates attack can dehydrate and rehydrate quickly to create similar conditions
resulting in similar rates of EFC.
4.3.2
4.3.3
A sample mounted in the four point bending jig as described above was exposed in
96%RH for a month. Figure 4.18 shows a stressed sample with applied maximum strain
113
of 0.006. Within days, exfoliation was apparent on both sides of the T face of the sample.
The images in Figure 4.18 are oriented such that the tensile side is on the left and the
compressive side is on the right. It is clear that the extent of EFC on the compressive
side was greater than on the tensile side, and this is supported by a detailed analysis of the
images, Figure 4.19.
increasing strain, while the exfoliation rate at the tensile side was not affected much by
tensile stress. The total amount of exfoliation on both sides, as measured by the total
width change of the unexfoliated region, is shown in Figure 4.20. The total amount of
EFC was similar for all applied strains, and was similar to the unstressed case.
Unstressed ESH tests on samples from the same plate showed that the material exfoliated
similarly on both sides, indicating that the two sides of the sample have the same EFC
susceptibility. Therefore, the different exfoliation rates on the two sides result from the
effects of the applied strain (and stress).
compression side was more or less balanced by the decreased exfoliation on the tension
side so that the total exfoliation rate on a bent sample was independent of strain. It is
known that the wedging stresses introduced by insoluble corrosion product at the grain
boundary can provide the energy required for lifting the surface gains and creating a
layered attack.[24-26] Due to the Poisson effect, the compressive side of the 4-point bend
specimen experienced lateral expansion normal to the direction of compression, which
helped push up the surface grains, consequently accelerating the exfoliation corrosion
rate.[19] Conversely, the tensile side of the specimen experienced lateral contraction
normal to the direction of the tension, which counteracted the wedging effect of the
corrosion product, resulting in lower exfoliation kinetics.
114
4.3.4
process, an Eyring relationship is often considered. The Eyring model also can be used
for stress variables other than temperature such as relatively humidity and mechanical
stresses. Considering temperature, T, and one additional non-thermal stress, S1, the time
to failure, tf, takes the general form[33]:
t f = AT exp{
H
C
+ ( B + )S1 }
kT
T
(1)
H
) can be included into parameter A. The Eyring model can be simplified to
kT
(2)
This exponential dependence on RH has also been applied to predict the failure rate of
integrated circuits.[34] Fitting the ESH results to this equation gives the following values
and standard deviations of the constants A and B for d in mm, RH in fractions, and t in
days:
A = 1.48E-05 mm/day ( = 2.96 E-06),
B = 6.44 ( = 0.21)
A fitted surface representing Eqn 2 is shown in Figure 4.21 along with the ESH data.
Mechanical stress effects may be included in the relationship by assuming that A
in Eqn 2 is stress dependent. The ESH results show that compressive and tensile stresses
have different effects on exfoliation corrosion kinetics. Compression accelerated the
116
EFC kinetics, whereas tension decreased EFC kinetics and was independent of the
magnitude of tension. A conditional function can be constructed to describe the different
effects of stress. It is expected that corrosion rate under different stress conditions should
also follow a linear relationship with respect to time for long exposure period. The
conditional function f ( ) is constructed as follows:
C1 exp( D ) when < 0 (Compression)
A = f ( ) =
C2
C3 exp( E )
where Ci, D, and E are constants. Combining this function into equation 2:
d = f ( ) t exp(B RH)
(3)
When = 0 , the sample is in the unstressed condition. For the unstressed condition, as
shown above, C2 = A = 1.48E-05 mm/day and B = 6.44. When < 0 , the sample is
under compression and Equation 3 becomes:
d = C1 t exp( B RH + D )
(4)
In our work, ESH tests under stress were carried out only at 96%RH, so RH=0.96. B has
same value as when = 0 , which is 6.44. Fitting the ESH data to this equation with in
fractional units results in C1 = 0.88E-05 mm/day ( = 0.29E-05) and D = 0.149E+03 ( =
0.6E+02). Since the 4-point bend ESH tests were only performed in one RH, the fitted
parameters might not be accurate at other RH values. However, we can still get general
information of corrosion kinetics from this model. When > 0 , the sample was under
tensile stress, the corrosion kinetics were found to be independent of stress levels. As
117
shown in figure 4.22, the average value under tension follows: d = 0.037+0.0019t.
Equation (3) becomes:
d = C 3 t exp( B RH E )
(5)
(6)
where E =( ln C3 + 6.18 - E). Fitting the average values under tensile stress into the
model, we get E = 3.9 (=0.014). Again for accurate values of the constants, tests
should be performed at different RH values.
4.4
CONCLUSIONS
Exfoliation corrosion (EFC) of AA7178 was studied using the exfoliation of slices in
humidity test, in which a through-thickness slice of a plate is pretreated electrochemically
and then exposed in a high humidity. The followings were found:
1.
EFC rate increased with humidity above the critical humidity of around 56% and
The effect of temper was quantitatively measured, and the EFC rate of AA7178-
The effects of stress were determined using a 4-point bend apparatus. The EFC
rate increased with increasing compressive stress. Tension decreased the rate of EFC,
independent of the stress level over the range studied.
4.
stress. The model is more accurate at longer time and higher humidity.
118
REFERENCES
J.P. Chubb, T.A. Morad, B.S. Hockenhull, and J.W. Bristow, International Journal
of Fatigue, 17(1), 49-54 (1995).
K. Ebtehaj, D. Hardie, and R.N. Parkins, British Corrosion Journal, 24(3), 184188 (1989).
M.J. Robinson and N.C. Jackson, Corrosion Science, 41, 1013-1028 (1999).
M.J. Robinson and N.C. Jackson, British Corrosion Journal, 34(1), 45-49 (1999).
D.O. Sprowls, T.J. Summerson, and F.E. Loftin, "Exfoliation Corrosion Testing
of 7075 and 7178 Aluminum Alloys-Interim Report on Atmospheric Exposure
Tests", in American Society for Testing and Materials, Philadelphia, PA: 99-112
(1973).
E.A.G. Liddiard, J.A. Whittaker, and H.K. Farmery, Journal of the Institute of
Metals, 89, 377-384 (1960-1961).
10
W.A. Bell and H.S. Campbell, Journal of the Institute of Metals, 89, 464-471
(1960-1961).
11
12
13
B.W. Lifka and D.O. Sprowls, "Relationship of Accelerated Test Methods for
Exfoliation Resistance in 7xxx Series Aluminum Alloys with Exposure to a
Seacoast Atmosphere", in ASTM SPEICIAL TECHNICAL PUBLUCATION 558,
306-333 (1973).
119
14
S. Lee and B.W. Lifka, "Modification of the EXCO Test Method for Exfoliation
Corrosion Susceptibility in 7XXX, 2XXX and Aluminum-Lithium Alloys", in New
Methods for Corrosion Testing of Aluminum Alloys, ASTM STP 1134, eds.,
V.S.Agarwala and G.M.Ugiansky, Philadelplia: American Society for Testing and
Materials, 1-19 (1992).
15
D.O. Sprowls, J.D. Walsh, and M.B. Shumaker, "Simplified Exfoliation Testing of
Aluminum Alloys", in Localized Corrosion-Cause of Metal Failure, ASTM STP
516, American Society for Testing and Materials, 38-65 (1972).
16
17
18
19
20
J.M. Holt, H. Mindlin, and C.Y. Ho, "Structural Alloys Handbook", West
Lafayette,IN:CINDAS/Purdue University,c1994.
21
J.R. Davis, Metals Handbook. Materials Park, OH: ASM International, (1998).
22
23
24
H.W. Pickering, F.H. Beck, and M.G. Fontana, Corrosion, 18, 230-239 (1962).
25
26
27
28
29
G.C. Montanari and A. Motori, Journal of Physics D: Applied Physics, 28, 14331438 (1995).
120
30
31
32
L. Bitman, Y.T. Choi, S.B. Choi, and N.M. Wereley, Smart Materials and
Microstructures, 14, 237-246 (2005).
33
34
B. Reich and E.B. Hakim, Solid State Technology, September, 65-66 (1972).
121
Alloy
Si
Fe
Table 4. 1
Element
Cu
Mn
Mg
2.04
0.0125 2.71
Cr
Zn
Ti
Al
0.187
6.46
0.0212 Bal.
Compound
Oxidation
BE (eV)
State
FWHM
(eV)
Al3+
73.8
1.058
23.2
compound, Al3+
76.8
4.893
40.4
Al 2p
Al2O3/Al(OH)3
Al 2p
Unknown
perhaps Al (OH)3-xClx
Al 2p
Al(OH)3
Al3+
74.5
1.278
36.4
O 1s
Al(OH)3
Al3-
531.9
1.889
44.0
O 1s
Al2O3/Al(OH)3
O2-/OH-
531.3
2.031
56.0
Cl 2p3/2
Cl-
198.1
1.292
46.4
Cl 2p3/2
chemical states
Cl-
199.9
1.010
20.6
Table 4. 2
122
Figure 4.1
0.012
Current (A)
0.01
0.008
0.006
0.004
0.002
0
Time (h)
Figure 4.2
Current response of two AA7178 samples during potentiostatic
pretreatment at -710 mV SCE in 1 M NaCl. Both samples had the same exposed area.
123
-600
-800
-1000
-1200
-1400
-1600 -11
10
10-9
10-7
10-5
10-3
10-1
Figure 4.3
Polarization curve of AA7178 S sample in deareated 1.0 M NaCl at a scan
rate of 0.1 mV/s.
Figure 4.4
Figure 4.5
Figure 4.6
Figure 4.7
SEM image of the surface of AA7178 sample after electrochemical
pretreatment at 5.5 mA/cm2 in 1M NaCl for 7 hours.
Figure 4.8
SEM micrograph of FIB cross section of IGC attack in the same AA7178
sample as in figure 4.7.
126
Figure 4.9
AA7178 wingskin slices after pretreatment and then 11 days in different
constant humidity. The rolling direction is vertical.
127
Figure 4.10 AA7178 wingskin slices after pretreatment and then 13 days in 30%
constant humidity. The rolling direction is vertical.
128
96%RH
Standard error in
single measurement
76%RH
66%RH
0.8
0.6
0.4
Standard error in
whole population
0.2
0
0
20
40
60
80
100
Time (days)
Figure 4.11 Width change of central unexfoliated region as a function of time for
AA7178 wingskin in various constant RH environments.
129
(a)
(b)
(c)
Figure 4.12 (a) SEM micrographs of corrosion product of AA7178 during 66%RH
exposure following electrochemical pretreatment. (b) SEM micrographs of corrosion
product of AA7178 during 76%RH exposure following electrochemical pretreatment. (c)
SEM micrographs of corrosion product of AA7178 during 96%RH exposure following
electrochemical pretreatment.
130
3000
2500
Al
Counts
2000
1500
O
1000
Cl
500
0
131
900
1150
700
Intensity (Counts)
Intensity (Counts)
800
1200
Al 2p
600
500
400
300
200
Cl p
3/2
1100
1050
1000
950
900
850
100
65
70
75
80
85
800
192
90
194
196
(a)
202
204
206
6500
O 1s
Zn 2p
600
Intensity (Counts)
Intensity (Counts)
200
(b)
800
700
198
500
400
300
200
6000
5500
5000
100
0
520
525
530
535
540
4500
1010
545
1020
1.1 10
1 10
1050
1060
(d)
Mg
Cu p 3/2
9000
60
Cu p
1/2
8000
7000
50
40
30
20
6000
5000
920
1040
70
Intensity (Counts)
(c)
4
1030
10
930
940
950
960
970
980
45
50
55
60
(e)
(f)
Figure 4.14 XPS spectra measured from samples of AA7178 exposed to 56%RH after
electrochemical pretreatment in 1 M NaCl for 7 hours. (a) Al 2p, (b) Cl 2p, (c) O 1s, (d)
Zn 2p, (e)Cu 2p, (f) Mg 2p.
132
0.25
56%RH
0.2
56%RH
96%RH
0.15
0.1
0.05
56%RH
96%RH
0
96%RH
10
20
30
40
50
60
70
80
Time (days)
Figure 4.15
133
(a)
(b)
Figure 4.16 Image of AA7178 sample exposed to 96%humidity following
galvanostatic pretreatment. (a) T6 temper. (b) T7 temper.
134
0.14
T6, 96% RH
0.12
T7, 96% RH
0.1
0.08
0.06
T6, 76% RH
0.04
0.02
0
0
10
15
20
25
Time (days)
Figure 4.17
135
Figure 4.18 Image of AA7178 sample under four point bending loading exposed to
96%humidity.
136
0.3
-0.006
-0.005
0.25
-0.004
0.2
compression
0.15
unstressed
0.1
tension
0.05
0.006
0.005
0.004
0
0
10
15
20
25
30
35
Time (days)
Figure 4.19
137
0.4
0.35
0.3
unstressed
0.25
0.2
0.15
0.1
0.006
0.005
0.004
0.05
0
0
10
15
20
25
30
35
40
Time (days)
Figure 4.20
138
Figure 4.21 Predictive 3D model for EFC kinetics. Solid square: ESH data in 96%RH;
Solid circle: ESH data in 76%RH; Solid triangle: ESH data in 65%RH.
139
0.12
0.1
0.08
0.06
d=0.037+0.0019t
0.04
0.02
0.006
0.005
0.004
0
-0.02
-5
10
15
20
25
30
Time (days)
Figure 4.22
140
CHAPTER 5
5.1
INTRODUCTION
Intergranular corrosion (IGC) and exfoliation corrosion (EFC) are often found in
the high strength Al alloys (e.g. AA2024 and AA7178) used in wingskins and fuselages
of airplanes.
airframes.
A great deal of effort has been devoted over the years to the study of
foils with a range of thickness, foil penetration experiments determine the growth kinetics
of the fastest growing localized corrosion sites.[2, 3] However, there are many other sites
in such samples that do not penetrate, and those sites have a set of different growth
kinetics. It is possible that many sites simply grow slower than the fastest sites. Others
might repassivate and stopped growing before they penetrate the foil. Still other sites
might have started growing after the initiation of the experiment. The latter two types of
sites could have had faster growth kinetics than what is determined by the foil penetration
141
this
work,
in
situ
x-ray
5.2
EXPERIMENTAL
Engineering Building at OSU. The microfocal radiographic system used in this study had
a 225 kV, 5 m (0.0002 in.) x-ray microfocal source. Samples were held with a specially
designed chuck that allowed for precise positioning between the x-ray source and x-ray
film holder as shown in figure 5.2. The chuck allows for accurate repositioning so that
intermittent x-ray radiography can be performed on a sample over an extended period of
time. Eastman Kodak AA film was used to reduce the exposure time as it is fast and high
contrast.
cross-section rather than a rectangular section. The long direction of the sample was
positioned vertically and sealed through the bottom of a plastic cell with epoxy.
Some AA2024-T3 samples were first encased in epoxy, except for the top face,
which was polished clean. IGC was then formed by polarization in 1 M NaCl at a
potential of 580 mV SCE. The attack was constrained to IGC in the encased samples.
Other samples were freely exposed to the solution with no epoxy encasement. Freelyexposed samples did not have the physical constraint of the epoxy, and the attack was in
the form of exfoliation rather than IGC. These samples had square cross sections with
their long axis orientated either in the L or LT direction.
IGC and EFC attack (on epoxy-encased and freely-exposed samples, respectively)
were studied by in situ x-ray radiography on samples exposed to a high humidity
environment after initiation in chloride solution. These samples were AA7178 pillars cut
from 8.9 mm thick plate removed from the wingskin of a retired KC135 airplane.
Samples were machined using a band saw or electric discharge machining (EDM) into
long pillars with their long axis orientated in either the L or T direction. Some samples
had square cross-sections of 2 mm x 2 mm and were exposed freely in the solution to
promote EFC. Other samples had cross-sections of 0.8 mm x 2 mm and were encased in
epoxy leaving only L face polished and exposed to promote IGC rather than EFC. The
sample surface was first ground in alcohol to 800 grit. The cell was the same as was used
for the AA2024-T3 samples. A 7 hour potentiostatic pretreatment in 1 M NaCl at -725
and -710 mV SCE was performed on the encased and free standing samples, respectively,
to initiate intergranular corrosion. From foil penetration experiments, it is expected that
this electrochemical treatment will generate attack about 0.46 mm in the AA7178
144
samples.[10] Following the pretreatment, the samples were rinsed with DI water and
placed in a sealed desiccator containing saturated Na2SO4 solution at room temperature.
This saturated solution created an environment with constant RH of 96%. Intermittent xray radiography was performed on these samples to obtain information on how IGC and
EFC propagated in high humidity.
5.3
a function of time in that figure. The rates comprise a band that falls just beneath the line
representing an extrapolation of the data from foil penetration experiments.[1] These
results show that in situ x-ray radiography provides information on the full range of
growth kinetics, including those for the fastest growing sites.
It should be noted that the circular features in this image are hydrogen bubbles.
Localized corrosion of Al alloys always results in the formation of hydrogen. Aggressive
dissolution of the Al alloys for a long period resulted in the formation of a foam in the
solution. The low density of the hydrogen bubbles causes them to appear dark, and the
bright line along their perimeter is from phase contrast, which highlights the boundary
between the bubble and solution.[11-14]
Figure 5.7 shows radiographs of a similarly oriented sample with no epoxy
encasement exposed under the same conditions. This sample had a square 2x2 mm crosssection. Figure 5.7a shows a radiograph prior to the electrochemical treatment. After 4
hr, the attack is apparent in the radiographs, figures 5.7b and 5.7c. Since the sample was
free standing with no epoxy encasement, attack proceeded from both the L face on the
top and the free T faces on the sides. The period of exposure was too short to generate
EFC, so the attack was IGC in nature. However, the IGC propagated inward from three
of the five exposed surfaces of the pillar sample: the top L surface and the two side T
surfaces. The orthogonal views in figures 5.7b and 5.7c show the morphology of attack.
In figure 5.7b, the sample is oriented so that the x-rays are parallel to the long transverse
direction. IGC can be seen as sharp dark lines. The lines are darker at the top of the
sample because the IGC penetrated from the top along the whole thickness. The x-rays
integrate the attack throughout the whole thickness of the samples. The fact that the IGC
146
propagated from three faces is clear in figure 5.7c, which is taken at 90 to figure 5.7b, i.e.
with the x-rays parallel to the through-thickness or S orientation of the microstructure.
The attack is seen from the top L surface and the two side T surfaces. The front of attack
coming from all faces of the sample was evident, and is highlighted in figure 5.7c by the
dotted lines.
With time, the attack on this free-standing sample developed into EFC. Figure
5.8 shows the radiograph after 22 hours of polarization at -580 mV SCE. The orientation
of this image is the same as in figure 5.7b, i.e. with the x-rays oriented along the T
direction of the microstructure. The many dark lines indicate that the IGC increased in
severity.
The higher magnification image in figure 5.8b shows that grains are
several days of exposure in the high humidity and it continued to grow at high rate.
Figure 5.11b shows a radiograph of the same sample after 48 days in the high humidity
environment. Tight fissures grow several mm into the sample. Figure 5.12 shows SEM
image of sharp intergranular fissure on AA7178. The fissure is around 1-2 um wide and
filled of corrosion product.
Such sharp fissures can also be found in metallographic cross-sections for
AA2024-T3 samples after a similar treatment, as shown in figure 5.13. This sample was
encased in epoxy, pretreated in 1 M NaCl at 580 mV SCE for 7 h, and placed in
96 %RH for 10 days.
Figure 5.14 and figure 5.15 shows x-ray radiographs for a free-standing AA7178
samples with long direction oriented in L and T direction respectively.
After the
exposed to the 96% RH environment, and rapid EFC is evident. These radiographs
clearly visualize the exfoliation corrosion and record the progress of the corrosion. As is
shown in the figures, the sample with long axis along the L direction exfoliated much
more severe than the sample with long axis along the T direction along. This is due to the
higher length to width ratio in L oriented sample than T oriented sample.
5.4
CONCLUSIONS
S direction, and as focused sharp features in the L or T direction. Novel cells that utilized
samples with specific size, shape, and orientation were used to perform in situ, real time
radiographic measurements. In situ radiography of such samples generated information
about the growth kinetics of a large number of sites with a range of penetration rates, not
just the fastest sites. In situ X-ray radiography is a good approach in studying immersed
samples exposed to different controlled conditions. It has great advantages in visualizing
the corrosion progress and hence makes it possible to quantify the exfoliation corrosion.
149
REFERENCES
W. Zhang and G.S. Frankel, Electrochemical and Solid-State Letters, 3(6), 268270 (2000).
"Standard Test Method for Exfoliation Corrosion Susceptibility in 2xxx and 7xxx
Series Aluminum Alloys (EXCO Test)", in Annual Book of ASTM Standards,
Philadelphia, PA: The American Society for Testing and Materials, 114-119
(1990).
M.J. Robinson and N.C. Jackson, Corrosion Science, 41, 1013-1028 (1999).
10
T.S. Huang and G.S. Frankel, Localized Corrosion Growth Kinetics in Al Alloys,
in 6th Joint FAA/DoD/NASA Aging Aircraft Conference-Sept. 16-19, 2002.
11
T.J. Davis, D. Gao, T.E. Gureyev, A.W. Stevenson, and S.W. Wilkins, Nature,
373(16), 595-598 (1995).
150
12
S.W. Wilkins, T.E. Gureyev, D. Gao, A. Pogany, and A.W. Stevenson, Nature,
384(28), 335-338 (1996).
13
14
151
Alloy
Fe
Cu
Mn
Mg
Cr
Zn
Ti
Al
AA2024 0.11
0.25
4.5
0.57
1.45
0.01
0.09
0.02
Bal.
AA7178 0.04
0.28
1.95
0.056
2.76
0.19
6.91
0.053
Bal.
Table 5. 1
Si
152
Figure 5.1
Figure 5.2
153
Figure 5.3
Metallographic sections of AA2024-T3 and AA7178-T6 wingskin, and the
terminology used for the three orientations, L: longitudinal, T: long transverse, S: short
transverse.
Figure 5.4
Metallographic cross section of S oriented 3 mm wide AA2024-T3
cylinder exposed to 1.0 M NaCl at -580 mV SCE for 4 hr.
154
250 m
Figure 5.5
In situ x-ray radiograph of L oriented AA2024-T3 sample exposed to 1.0
M NaCl at -580 mV SCE for 19 hr.
155
2
From foil penetration:
d = 0.3244 t
Depth (mm)
1.5
1/2
1
1/2
d = 0.2138 t
slowest growing site
0.5
12
16
20
24
28
32
Time (h)
Figure 5.6
Plot of the depth of various specific sites for L oriented AA2024-T3
sample as a function of time. Also shown is the curve for L oriented foil penetration
samples [1].
156
t=0
2 mm
a
t=4h
90
Figure 5.7
Radiographs of 2x2 mm sample of AA2024-T3 exposed to 1.0 M NaCl at
580 mV SCE. Sample not encased in epoxy. (a) Image was taken before exposure; (b,c)
images were taken after 4 h. Long axis (vertical orientation of sample) is L direction.
90 image was taken in T direction, 0 image was taken in S direction.
157
Figure 5.8
158
Figure 5.9
In situ x-ray radiograph of L oriented AA7178 sample exposed to 1.0 M
NaCl at -725 mV SCE for 3 hrs.
159
0.8
0.7
d=0.34t
Depth (mm)
0.6
1/3
0.5
0.4
d=0.23t
1/3
0.3
0.2
0.1
0
Time (hrs)
Figure 5.10 Plot of the depth of various specific sites for L oriented AA7178 wingskin
sample as a function of time.
160
Figure 5.11 Samples were put in 96%RH after electrochemical treatment. Sharp IGC
continued along L direction at high rate. (a) X-ray radiograph taken at 0 days. (b) X-ray
radiograph taken at 48days.
161
Figure 5.12 SEM image of intergranular sharp fissure in AA7178-T6 after 7 hours in 1
M NaCl at a potential of 710mV SCE.
162
Figure 5.13
Metallographic cross-section of AA2024-T3 L sample after 10 days in the
environment of RH 96%, T=22C~24C.
163
Figure 5.14
Samples were put in 96%RH after above electrochemical treatment.
Sharp IGC continued along L direction at high rate. (a) 0 day in 96%RH (after 7 hours in
1 M NaCl, -710mV SCE). (b) 6 days in 96%RH. (c) 30 days in 96%RH.
164
Figure 5.15 Samples were put in 96%RH after above electrochemical treatment. Sharp
IGC continued along T direction at high rate. (a) 0 day in 96%RH (after 7 hours in 1 M
NaCl, -710mV SCE). (b) 6 days in 96%RH. (c) 48 days in 96%RH.
165
CHAPTER 6
6.1
CONCLUSIONS
In this work, the exfoliation corroaion (EFC) of high strength Al alloys was
studied quantitatively. The effects of alloy temper, microstructure, local chemistry at the
grain boundary, relative humidity and applied mechanical stress on EFC behavior was
characterized. A new technique: exfoliation of slices in humidity (ESH) was developed
and X-ray radiography was utilized to quantify the corrosion behavior under various
conditions. A four point bending jig was used to study the effect of the mechanical stress
on corrosion behavior. This study provides good methods to quantitatively investigate
exfoliation corrosion kinetics.
environmental conditions and the alloy microstructure affect the exfoliation corrosion
kinetics. The followings are the main findings:
1. The initiation and propagation of exfoliation corrosion strongly depends upon
environmental condition, especially relative humidity. EFC of sliced samples of
166
AA7178 was not observed below a critical relative humidity of about 56% and the
EFC kinetics increased with increasing humidity above this critical humidity.
2. EFC susceptibility was found to depend on grain size and shape as well as grain
boundary composition. The more elongated grains or higher grain aspect ratio
and bigger grain size result in higher susceptibility to EFC, while small grain
aspect ratio will more likely develop pitting attack. Samples with similar grain
aspect ratio but different grain boundary composition will show different
susceptibility to EFC.
susceptibility to EFC was associated with a high Zn content in the precipitate free
zone at grain boundaries.
3. The effect of temper on exfoliation corrosion has been qualitatively studied. In
this study, a quantitative evaluation of the effect of T6 and T7 temper on AA7178
was performed by utilizing ESH technique. The EFC rate of AA7178-T6 was
found to be higher than that of AA7178-T7, which is consistent with EXCO test
results.
4. To simulate the stress condition of airplane wingskin, a 4-point bend apparatus
was designed.
compression on one side and under tensile stress on the other side. The effects of
surface strain on exfoliation corrosion kinetics were determined by placing 4point bend apparatus in constant humidity chamber. The EFC rate was observed
to increase with increasing compressive stress. Tension decreased the rate of
EFC, but level of tension stress over the range that was studied has little effect.
The summation of exfoliation corrosion of both sides showed that the total
167
It has great
advantages in visualizing the corrosion progress and hence can be used to observe
the transition between intergranular corrosion and exfoliation corrosion.
6.2
FUTURE WORK
This work has successfully introduced a new technique and methodology to study
the exfoliation corrosion kinetics leading to a better understanding of quantitative
168
The
critical relative humidity was determined roughly based on the results of the
exposure in these humidity values at room temperature. ESH tests on more RH
values at varied temperature are necessary in order to get detailed information of
threshold humidity and the effect of relative humidity on EFC.
2. Because of the rapid response of the EFC to RH, the empirical model developed
in this study can be used to predict the exfoliation extent for a sample exposed to
varying RH conditions. However, the responsiveness of this system to changes in
the environment might result from the specific sample geometry and
configuration used in the experiment. The full section of the slices was exposed
to the environment, whereas a real structure is typically only exposed on the top
surface. In future work, different sample geometry should be tested in ESH test to
clarify how the geometry will affect the EFC kinetics.
3. In order to obtain a complete predictive model for exfoliation corrosion as a
function of stress, more stress level should be used for determining the stress
effect on exfoliation corrosion. The exposure time in humidity under stressed
condition should be extended in order to get a reliable time vs. kinetics
relationship under stressed condition.
169
170
APPENDIX A
A.1
Subsequent
exposure to humidity results in the growth of IGC from these sites. Figure A.2 shows an
L cross-section of a specimen exposed in relative humidity illustrating the EFC process.
The phenomenological model is developed based on the following assumptions:
1.
Selective grain attack along T direction was created during pretreatment (figure
A.2a). The depth of attack can be obtained from foil penetration experiments: a(m) =
0.228 t(h)1/3.[1]
2.
After pretreatment for time tpre, the unattacked region along T direction has the
grows faster at the edge of sample (figure A.2b). We assume that the kinetics of end
grain attack is linear: de = ket, where de is the extent of end grain corrosion attack in the T
direction and ke is a constant. It is further assumed that all grains other than at the end of
the sample exhibit sharp IGC fissure growth at the much slower rate given by the IGC
sharp fissure kinetics of constrained samples: dT (m) = 5.14 t(h)0.58 (along T
direction).[2] When IGC at the two edges growing inwards in the T direction meet each
other (figure A.2c), the outermost layer disconnects from the bulk material as exfoliated
material shown in A.2d. The second outermost grain layer then becomes the end grain
and starts to follow end grain kinetics. The process is repeated for each underlying layer.
We assume each exfoliated layer has the thickness of one-grain dimension along the S
direction. For the first layer, the distance that end grain sharp fissure travels along T
direction is c0, which is the unattacked thickness right after the pretreatment.
The
unattacked thickness for each underlying layer decreases with time due to sharp fissure
growth and can be calculated by: dei=c0-5.14t0.58, i=1,2,3n.
A.2
A finite element model was developed and analyzed using the commercial finite
element program ABAQUS.
AA7178-T6,
yield
stress
y = 538MPa.
Ultimate
tensile
(a)
Let 0 = y
To determine k and n, two conditions were used as following:
1.
fp
f = UTS (1 + e f )
2.
fp
= f
f
E
= 0.008326 .
(1)
(2)
n=
f
f 0
fp
(3)
f 0
( fp ) n
(4)
Solving the above equations, we obtain n = 0.88 and k = 523 Mpa. The stress strain
relationship can be determined by equation (a). The input stress-strain data is listed in
table A.1.
During finite element simulation, the boundary condition is semi-infinite
condition. The left side and top side of the model was fixed in x and y directions. A
force of 0.9N calculated from measured wedging stress was applied on the U-shape,
forcing it to move towards the crack tip. This mimics the wedging action of the corrosion
product, forcing the crack to open.
A.3
The Von-Mises stress was used to show the stress field. The von-Mises stress is a
scalar measure of the stress state (the normal and shear stresses) at any point within a
body. It is a stress quantity that is proportional to the strain energy density associated
with a change in shape (with a zero volume change) at a material point:
vm =
1
2
In figure A.4, a 2D contour plot from finite element simulation shows that high stresses
develop around the crack tip at the edge during EFC. The stress around the crack tip is
decreasing towards the center of the sample. The Von Mises stress at the crack tip of the
174
edge can be as high as 500MPa. While the Von Mises stress for the inner crack tip is
only between 40-90MPa. The finite element simulation of the stress field during EFC is
consistent with the experimental results which showed that unconstrained sample
developed exfoliation corrosion.
175
REFERENCES
176
Stress(Mpa)
538
547.0887
554.7266
561.8984
568.7833
575.4624
581.982
588.3719
594.6527
600.8398
606.9448
612.9769
618.9434
624.8504
630.7031
636.5058
642.2623
647.9757
653.6489
659.2844
664.8844
Table A. 1
Strain
0
0.01
0.02
0.03
0.04
0.05
0.06
0.07
0.08
0.09
0.1
0.11
0.12
0.13
0.14
0.15
0.16
0.17
0.18
0.19
0.2
177
Figure A. 1
178
Figure A. 2
Figure A. 3
180
Figure A. 4 2D contour plot showing stress field around intergranular crack tip
simulated by ABAQUS.
181
BIBLIOGRAPHY
"Standard Test Method for Exfoliation Corrosion Susceptibility in 2xxx and 7xxx Series
Aluminum Alloys (EXCO Test)", in Annual Book of ASTM Standards,
Philadelphia, PA: The American Society for Testing and Materials, 114-119
(1990).
D. Adenis and A. Guilhaudis, Memoires Scientifiques de la Revue de Metallurgie, 64(10),
877-888 (1967).
J. Albrecht, I. M. Bernstein and A. W. Thompson, Metallurgical Transactions A, 13A,
811-820 (1982).
J. Albrecht, A. W. Thompson and I. M. Bernstein, Metallurgical Transaction A, 10A,
1759-1766 (1979).
R. Ambat and E. S. Dwarakadasa, Corrosion Science, 33(5), 681-690 (1992).
D. R. Baer, C. F. Windisch, M. H. Engelhard, M. J. Danielson and R. H. Jones, Journal of
Vacuum Science and Technology A, 18(1), 131-136 (2000).
G. Bassi and J. J. Theler, Zeitschrift fur Metallkunde, 60(3), 179-184 (1969).
W. A. Bell and H. S. Campbell, Journal of the Institute of Metals, 89, 464-471 (19601961).
W. W. Binger, E. H. Hollingsworth and D. O. Sprowls, Aluminum, Metals Park, OH:
American Society for Metals (1967).
L. Bitman, Y. T. Choi, S. B. Choi and N. M. Wereley, Smart Materials and
Microstructures, 14, 237-246 (2005).
A. P. Bond, G. F. Bolling, H. A. Domian and H. Biloni, Journal of The Electrochemical
Society, 113, 773 (1966).
P. L. Bonora, G. P. Ponzano and V. Lorenzelli, British Corrosion Journal, 9(2), 108-111
(1974).
182
183
184
185
191