EXFOLIATION CORROSION KINETICS OF HIGH STRENGTH ALUMINUM

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

Dr. Rudolph G. Buchheit

Dr. Michael J. Mills
Dr. Stanislav I. Rokhlin

____________________________
Adviser
Graduate Program in
Materials Science and Engineering

ABSTRACT

The objectives of this study were to quantitatively study localized corrosion,

especially exfoliation corrosion (EFC), of high strength aluminum alloys and to
investigate the mechanism of exfoliation corrosion with a focus on the effects of alloy
temper, microstructure, relative humidity (RH) and mechanical stress. A new
technique, Exfoliation of Slices in Humidity (ESH) was developed for the
determination of exfoliation corrosion (EFC) susceptibility and quantification of EFC
kinetics. Two AA7178 plates taken from the wingskin of a retired KC135 airplane
were used as test samples.

Slices of the plate were pretreated by potentiostatic

polarization in chloride solution to develop localized corrosion sites.

Subsequent

exposure to high humidity after pretreatment of properly oriented and unconstrained

samples resulted in the development of EFC at the edges of the slices. The EFC
kinetics was determined by measuring the width of the central unattacked region of the
samples. The ESH results were representative of the different EFC behavior of the two
plates during outdoor exposure. These results show the capability of the ESH test to
discriminate between plates of varying susceptibility and to determine EFC rates
ii

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

Unencased samples given the same

treatment exhibited exfoliation in the high humidity environment.

In situ X-ray

radiography of intergranular corrosion attack provides a wide range of intergranular

corrosion kinetics including the fastest growing sites that can be detected by foil
penetration technique. This method is a good approach for visualizing the EFC process.

iv

To my father, Tianhai Zhao and my mother, Fengqin Zhao

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.

X.Zhao, G.S.Frankel, B.Zoofan, and S.I.Rokhlin, In-Situ X-Ray Radiographic

Study of Intergranular Corrosion in Al Alloys, Corrosion, 59,1012-1018
(2003)

2.

X.Zhao, G.S.Frankel, The Visual Determination of Exfoliation Rate of Al

Alloy Slices in Humidity, The Journal of Corrosion Science and Engineering,
Volume 6 Paper C134, 2003, The paper was presented at the conference
Corrosion Science in the 21st Century held at UMST in July 2003.

3.

X. Zhao, T. Huang, G. S. Frankel, B. Zoofan, and S. I. Rokhlin, " Intergranular

Corrosion Morphology and Growth Kinetics In High Strength Aluminum
Alloys," in Critical Factors in Localized Corrosion IV, S. Virtanen, P. Schmutz,
and G. S. Frankel, eds, PV 2002-04, The Electrochemical Society, 2003.

4.

T.Huang, X.Zhao, Gerald Frankel, B.Zoofan and S.Rokhlin, Growth Kinetics

of Intergranular and Exfoliation Corrosion in AA7178, proceedings of
Triservice Corrosion Conference, Las Vegas, 2003
viii

5.

Tsai-Shang Huang, Xinyan Zhao, and G.S.Frankel, Localized Corrosion

Growth Kinetics in AA7xxx Alloys, 2005 Triservice Corrosion Conference
Proceedings, 2005

6.

G.S.Frankel, Tsai-Shang Huang, and Xinyan Zhao, Localized Corrosion

Growth Kinetics in AA7178, P.Marcus, ed., Elsevier, 2005

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

2.4.2 Alloy Microstructure ..............................................................................34

2.4.2.1 Intermetallic Particles .................................................................34
2.4.2.1.1 Intermetallic Particles in Al-Cu & Al-Cu-Mg Alloys ....35
2.4.2.1.2 Intermetallic particles in Al-Zn-Mg & Al-Zn-Mg-Cu
Alloys...........................................................................38
2.4.2.1.3 Other Intermetallic Particles .........................................39
2.4.2.2 Precipitate Free Zones (PFZ).......................................................41
2.5 INTERGRANULAR STRESS CORROSION CRACKING (IGSCC)...............................42
2.5.1 Mechanisms of SCC ...............................................................................43
2.5.1.1 Anodic Dissolution Mechanism ..................................................44
2.5.1.2 Cathodic Mechanism...................................................................46
2.5.2 Effects of Microstructure and Alloy Chemistry on IGSCC......................47
2.5.3 Relationship between Stress Corrosion Cracking and Exfoliation
Corrosion ...............................................................................................50
2.6 RESEARCH OBJECTIVES ....................................................................................50
3.

QUANTITATIVE STUDY OF EXFOLIATION CORROSION: EXFOLIATION

OF SLICES IN HUMIDITY TECHNIQUE ..........................................................69
3.1 INTRODUCTION ................................................................................................69
3.2 EXPERIMENTAL................................................................................................71
3.2.1 Materials ................................................................................................71
3.2.2 Exfoliation of Slices in Humidity (ESH) Technique................................74
3.3 RESULTS AND DISCUSSION ...............................................................................75
3.3.1 Electrochemical and ESH Characterization.............................................75
3.3.2 Optical Microscopy ................................................................................79
3.3.3 TEM Analysis ........................................................................................80
3.4 CONCLUSIONS..................................................................................................82

4.

EFFECTS OF RH, TEMPER AND STRESS ON EXFOLIATION CORROSION

KINETICS OF AA7178......................................................................................102
4.1 INTRODUCTION ..............................................................................................102
4.2 EXPERIMENTAL..............................................................................................103
4.2.1 Four Point Bending Technique .............................................................107
4.3 RESULTS AND DISCUSSION .............................................................................109
4.3.1 Effect of RH on EFC Kinetics ..............................................................109
4.3.2 Effect of Temper on EFC Kinetics........................................................113
4.3.3 Effect of Stress on EFC Kinetics...........................................................113
xi

4.3.4 Empirical Model for Exfoliation Corrosion...........................................115

4.4 CONCLUSIONS................................................................................................118
5.

IN SITU X-RAY RADIOGRAPHY OF LOCALIZED CORROSION ................141

5.1 INTRODUCTION ..............................................................................................141
5.2 EXPERIMENTAL..............................................................................................142
5.3 RESULTS AND DISCUSSION .............................................................................145
5.4 CONCLUSIONS................................................................................................148

6.

CONCLUSION AND FUTURE WORK.............................................................166

6.1 CONCLUSIONS................................................................................................166
6.2 FUTURE WORK ...............................................................................................168

APPENDIX A: FINITE ELEMENT SIMULATION ...................................................171

A.1 PHENOMENOLOGICAL MODEL OF EXFOLIATION CORROSION ............................171
A.2 MODEL DEVELOPMENT IN ABAQUS............................................................172
A.3 RESULTS AND DISCUSSIONS ..........................................................................174

BIBLIOGRAPHY .......................................................................................................182

xii

LIST OF TABLES

Table

Page

3. 1 Humidity associated with saturated salt solutions. ..................................................86

4. 1 Chemical composition of AA7178. ......................................................................122
4. 2 Data obtained from XPS spectra...........................................................................122
5. 1 Composition of AA2024 and AA7178 plates. ......................................................152
A. 1 Stress strain input data. ........................................................................................177

xiii

LIST OF FIGURES

Figure

Page

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) ............62

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

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.).............................................63

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

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

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

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]..................................................................................................66

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

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

Variation in corrosion potential of FeAl3 and -Al(Fe,Mn)Si.[91] (Reproduced

by permission of The Electrochemical Society, Inc.) ..........................................67

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

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

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. ....................................87

3.2

Schematic drawing of orientation of slices for ESH test relative to elongated

microstructure of AA7178 wingskin sample.......................................................87

3.3

Polarization curves of AA7178 good and bad sample in deareated 1.0 M

NaCl at a scan rate of 0.1 mV/s. .........................................................................88
xv

3.4

Images of AA7178 wingskin samples exposed to 96% humidity following

electrochemical pretreatment. (a) Bad sample. (b) Good sample. .................89

3.5

Average exfoliation corrosion depth for duplicate samples of good (triangles)

and bad (circles) AA7178 wingskin samples exposed to 96% RH...................90

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

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

TEM micrographs of grain-boundary in AA7178 good plate. (a) Near up side

of plate. (b) Near down side of the plate. ............................................................95

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. ...........................96
3.12

TEM micrographs of grain-boundary in AA7178 showing Nano EDS line

profiling across the GB SDZ (a) Near up side of the good plate. (b) Near
down side of the good plate. ...........................................................................98

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. ...............................................99

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

Heat treatment of AA7178 wingskin plate........................................................123

xvi

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

Polarization curve of AA7178 S sample in deareated 1.0 M NaCl at a scan rate

of 0.1 mV/s. .....................................................................................................124

4.4

Four point bending mode. ................................................................................124

4.5

Four point bending jig. .....................................................................................125

4.6

Four point bending test set up...........................................................................125

4.7

SEM image of the surface of AA7178 sample after electrochemical

pretreatment at 5.5 mA/cm2 in 1M NaCl for 7 hours.........................................126

4.8

SEM micrograph of FIB cross section of IGC attack in the same AA7178
sample as in figure 4.7......................................................................................126

4.9

AA7178 wingskin slices after pretreatment and then 11 days in different

constant humidity. The rolling direction is vertical. .........................................127

4.10

AA7178 wingskin slices after pretreatment and then 13 days in 30% constant
humidity. The rolling direction is vertical. .......................................................128

4.11

Width change of central unexfoliated region as a function of time for AA7178

wingskin in various constant RH environments. ...............................................129

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

4.13

X-ray EDS spectrograms of corrosion product in 56%RH. Accelerating

voltage=12kV. .................................................................................................131

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

4.15

Results of cyclic exfoliation test for AA7178 wingskin samples.......................133

4.16

Image of AA7178 sample exposed to 96%humidity following galvanostatic

pretreatment. (a) T6 temper. (b) T7 temper.......................................................134

4.17

Exfoliation corrosion kinetics of AA7178 samples exposed to 96% RH. ..........135

4.18

Image of AA7178 sample under four point bending loading exposed to

96%humidity. ..................................................................................................136

4.19

Exfoliation corrosion kinetics under compression and tension. .........................137

4.20

Exfoliation corrosion kinetics under different strain conditions. .......................138

4.21

Predictive 3D model for EFC kinetics. .............................................................139

4.22

Average kinetics of exfoliation corrosion in 96%RH under tension. .................140

xvii

5.1

Schematic of the microradiographic system......................................................153

5.2

In situ x-ray microradiographic experimental setup ..........................................153

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. ...........................................................................................154

5.4

Metallographic cross section of ST oriented 3 mm wide AA2024-T3 cylinder

exposed to 1.0 M NaCl at -580 mV SCE for 4 hr..............................................154

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

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

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

5.8

Same sample as Figure 7, taken in T direction. Exfoliation is evident..............158

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

5.10

Plot of the depth of various specific sites for L oriented AA7178 wingskin
sample as a function of time. ............................................................................160

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

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

5.13

Metallographic cross-section of AA2024-T3 L sample after 10 days in the

environment of RH 96%, T=22C~24C. ...........................................................163

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

IGC attacks are created in the sample from electrochemical pretreatment.........178

A. 2

EFC process of a sample exposed in high relative humidity. (L section)...........179

A. 3

Two dimensional finite element mesh. .............................................................180

A. 4

2D contour plot showing stress field around intergranular crack tip simulated by
ABAQUS.........................................................................................................181

xix

CHAPTER 1

INTRODUCTION

This dissertation consists of six chapters including this introduction. Chapter 2 is

a literature review on several topics. Exfoliation corrosion is reviewed first including the
effect of alloy microstructure, alloy addition, environmental condition, heat treatment and
techniques for exfoliation corrosion testing. Then intergraunular corrosion and stress
corrosion cracking are discussed and reviewed.

Since there is a close relationship

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.

Wedging stresses that

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

BASIC FORMS OF CORROSION OF ALUMINUM ALLOYS

Summerson and Sprowls have classified the corrosion of aluminum alloys into

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

The Mechanism of Exfoliation Corrosion

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.

2. A preferential anodic path along grain boundaries resulting from the

electrochemical action between precipitates and adjacent solute denuded zones.[21]
3. A Corrosive environment.
The driving force for EFC is the wedging stress that is produced by the insoluble
corrosion product. Pickering clearly stated that the wedging stress could be produced
when the volume of insoluble corrosion product is larger than the metal from which it
formed and the product is formed in a restricted region.[22] The corrosion product in 188 stainless steel was calculated to be 4000 psi to 7000 psi, providing all the energy
required for stress corrosion cracking.[22]
Robinson et al. have measured the corrosion product force in aluminum alloys.[10, 23,
24]

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.

He concluded that the EFC susceptibility of the precipitation

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.

To investigate the correlation between stress

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

wedging action of corrosion product is essential to continued EFC propagation by a stress

corrosion mechanism.[8, 9, 22, 23, 27-29]
The extent of exfoliation is affected by many factors, especially the alloy composition,
grain aspect ratio and heat treatment, which will be discussed in detail in the following.

10

2.3.2

Factors That Influence EFC

2.3.2.1 Effects of Alloying Elements on EFC

Mattsson et al. have stated that an increase in Fe content decreases resistance to
exfoliation corrosion in Al-Zn-Mg alloys.[20] In studying the EFC mechanism of Al5Zn-Mg1, they proposed that the main factor governing the amount of the -Al(Fe,Me)Si
phase is the Fe content. Increasing Fe content thus markedly increases the amount of Al(Fe,Me)Si. This phase is active as a cathode so that increasing Fe will increase the
susceptibility to EFC.

Increasing Mn content will also increase the amount of -

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.

2.3.2.2 Effects of The Grain Shape on Exfoliation Corrosion

Elongated grains are an important requirement for EFC.[21] It was reported that the
severity of exfoliation corrosion depends on the grain aspect ratio of the material.[1, 8, 9,
24] The greater the grain aspect ratio, the more severe the exfoliation attacks.
Robinson developed a mathematical model of the exfoliation process.[9] In his model,
the effect of grain shape on exfoliation susceptibility was studied. Figure 2.3 shows a
12

schematic of intergranular corrosion through the grains of a wrought aluminum alloy. In

this model, the aluminum skin was assumed to deform into a section of a hemispherical
blister, the pressure within the blister is assumed to be uniform. The height of blisters
was calculated from the depth of unattacked metal from which the corrosion product was
formed. Robinson calculated the strain that was generated in an exfoliation blister by
corrosion product through the equation of c=(L-L)/L; the meaning of L and L are
shown in figure 2.4. An iterative process was used to obtain several values for blister
strain that develops with a certain volume of metal converted to corrosion product. When
a blister strain reached the fracture strain of 11%, the blister was said to fracture. Alloys
with larger grain aspect ratios take longer time to reach the 11% fracture strain than
alloys with smaller grain aspect ratios, and a larger diameter of blister develops for the
former. More elongated grains convert more material to reach the surface fracture strain
and result in a more severe surface attack on the surface. The depth of penetration was
shown to be unrelated to the shape of the grains. The model investigated exfoliation
mainly from the mechanical aspect and many assumptions were introduced to simplify
the model. The real situation is more complicated and electrochemical reactions are an
essential element that cannot be neglected throughout the EFC process.
More recently, Robinson and Jackson performed exfoliation tests on Al-Cu-Mg alloys
AA2014-T651 and AA2024-T351.[1, 8] Samples with different sections from each plate
material were exfoliated using an acidified intermittent salt fog test.[35] The corrosion
was assessed by rating the surface appearance with reference to a series of standard
photographs given in ASTM G34[36] and by measuring the depths of EFC attack. The
severity of exfoliation corrosion was shown to be related to the grain aspect ratio of the
13

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.

It was shown that the surface rating of

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

2.3.2.3 Effect of Heat Treatment on Susceptibility to Exfoliation Corrosion

The heat treatment condition of the alloy is thought to be important in exfoliation
corrosion because heat treatment changes the microchemistry.[4, 39, 40]
Winifred carried out a full investigation on the susceptibility of HE15 (Al-CuMg-Mn-Si-Fe) to EFC considering the influence of copper content and aging treatment
on susceptibility to exfoliation corrosion.[40] The laboratory tests clearly showed that
extended aging prevented the exfoliation corrosion. The aging temperature also has an
effect on the extent of EFC. The specimens fully aged at 200C showed more severe
attack than those fully aged at 165C or 185C. Higher copper content in the material is
desirable for beneficial effect of extended aging to resistance to EFC. The extended
aging results in reduction in tensile stress as well as hardness. Metallographic studies on
HE15 were conducted and provided an explanation that as the aging time increased, a
more apparent subgrain structure could be observed. In over aged material, corrosion
takes place entirely at the substructure boundaries and does not lead to EFC.
The grain boundaries of an Al-4%Cu alloy with different aging duration were
observed using TEM by Garner and Tromans.[41] The underaged material showed intact
Al2Cu precipitates after immersion in test solutions, but the grain boundary area had been
attacked.

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.

Sheppard et al. have found that the susceptibility to

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

EFC is generally observed under specific environmental conditions such as high

humidity and salt containing environments.[2, 4, 33, 47] Lifka and Sprowls found that the
corrosivity of an atmosphere is controlled by factors relating to the moisture on a
corroding surface and by atmospheric pollutants such as chlorides.[48] It was also shown
18

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 Testing

It was reported that exfoliation propagated at a nearly linear rate.[50] Robinsons

study on exfoliation extent by measuring depth of penetration during MASTMAASIS
test[35] also showed that rate of penetration of EFC was almost linear.[1] From the
discussion above, we know that, by mechanical processing or heat treatment to eliminate
the preferential corrosion path or change grain structures, we can improve the resistance
to exfoliation corrosion of aluminum alloy. However, their effects on EFC rate have only
been studied qualitatively. Previous test methods were limited to natural environments,
but these are time consuming and accelerated tests are often required.[51] A good
exfoliation test should show enough sensitivity to distinguish between susceptible and
non-susceptible alloys and ideally show results during a short exposure period, which
should correlate well with extended tests such as atmospheric exposure to marine
environments.

Efforts have been devoted to investigate the extent or the rate of

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

2.3.4.1 Immersion Test

The EXCO test according to ASTM G34 is widely used to evaluate the
exfoliation corrosion behavior of 2xxx and 7xxx series aluminum alloys.[36] This
accelerated test method can help people to predict the performance of the products after
many years of service by using an aggressive corrosion solution containing 4.0M NaCl,
0.5M KNO3 and 0.1 M HNO3. The solution has an initial pH of 0.4. After constant
immersion in the EXCO solution for a period of 48 to 96hrs, depending on the alloy
system, the susceptibility to exfoliation is determined by visual assessment of the surface.
The surface is compared to a series of standard photographs of sample surface ratings
from N (No appreciable attack), P(pitting), to various degrees of exfoliation attack EAED, which represent conditions from superficial exfoliation to very severe exfoliation.
During the test, the pH of the solution will increase from 0.4 to about 3. The reason for
the increase in pH was attributed to the hydrolysis of aluminum ions.[51] At low pH,
following reaction occurs:
Al+3H+Al3++3/2H2 [51]
After part of the acid is consumed, the aluminum hydrolyzes and has following
equilibrium reactions:
Al3++H2OAl(OH)2++H+
Al3++2H2OAl(OH)2++2H+
Al3++3H2OAl(OH)3+3H+
Until the pH for the deposition of Al(OH)3 is reached.[51] Consequently, the
attack will continue at a lower rate. EXCO test successfully distinguishes between the
resistant T76 temper and the susceptible T6 temper for AA7075.
20

Exfoliation

performances of AA7075 and AA7178 alloys in atmospheric exposure and laboratoryaccelerated tests were compared.

The EXCO test was successful in predicting the

resistance to exfoliation of AA7075 and AA7178 in seacoast and industrial atmospheric

environments.[7]
However, the EXCO test is generally considered to be unreliable for predicting
the exfoliation corrosion behavior of Al-Li alloys due to the poor correlation with
atmospheric exposure data.[52] One limitation for the EXCO test method is that it is not
identical to the service environment. The EXCO solution is too corrosive for some
materials in the T7 tempers and does not always accurately predict the resistance of
AA2024 and AA7x50 products.[52] The EXCO test could not reveal intermediate
exfoliation corrosion resistance for AA2xxx and AA7xxx series aluminum alloys. Lee
and Lifka proposed a modified EXCO solution with pH adjusted to 3.2 by the addition of
HCl and aluminum ions.[46] The solution contains 2.96g/L AlCl3, 230g/L NaCl, and
60.7g/L KNO3. Results of the modified EXCO solution were in good agreement with the
performance of the alloys in actual service conditions.[46, 52] The modified EXCO test
reproduced the atmospheric data better than standard EXCO test. A good correlation
between the performance in natural environments and in modified EXCO exposure was
reported for Al-Cu-Mg AA2024 and Al-Cu-Li alloy AA2090. However, the Al-Li alloys
seemed to suffer more severe exfoliation than in the standard EXCO solution.
Other immersion tests such as the ASSET test was used by Sprowls et al. to look
at the exfoliation behavior of AA5xxx and AA7xxx aluminum alloys.[1, 38] A solution
containing ammonium chloride and ammonium nitrate with additions of ammonium

21

tartrate and hydrogen peroxide was devised for the ASSET test. This test was found to
give a good correlation with natural exfoliation.

2.3.4.2 Salt Spray Tests

The SWAAT test (Sea Water Acetic Acid Test) was first developed by Romans.[53]
The 5% synthetic seawater was used and was acidified to pH 2.8 by the addition of
glacial acetic acid.

The test solution is introduced onto the surface of aluminum

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.

The standard EXCO test

indicated better exfoliation corrosion behavior of the alloys investigated except for
23

AA8090-T6 and AA7075-T7351 plates. In the modified EXCO test, AA7075-T7351

panels were susceptible to pitting, whereas the other alloys studied generally suffered
more severe EFC than in the standard EXCO test.[52]
Immersion test and salt spray tests are useful for research and development
purposes but should not be used as a method for quality acceptance or quantification.
The discrepancy of the results between accelerated test and environmental exposure tests
as well as the different results among several accelerated tests indicates the limitation of
these tests. The appropriate synthetic environment for use in accelerated exfoliation
corrosion tests is still an open question. Another drawback is the visual assessment of the
corrosion attack, which is subjective.

Furthermore, visual assessment cannot give

sufficient evidence of the depth of penetration or kinetics. To get the full information of
EFC, metallographic sectioning is often required.

2.3.4.3 Electrochemical Impedance Spectroscopy

In an attempt to obtain improved predictive ability for EFC behavior, Keddam et
al. examined the ability of EIS to distinguish exfoliation corrosion from other forms of
corrosion.[54] AA2024-T351, AA2219-T87 and AA6013-T6 were divided into two
groups in terms of their different impedance behavior. In AA2024-T351 and AA6013-T6,
a marked feature corresponding to potential distribution inside pores or lamellae by the
corrosion was observed, indicting the formation of exfoliation or intergranular corrosion.
No such feature was observed in AA2219. They concluded that EIS could distinguish
between formation of pits and deep lamellae due to exfoliation or intergranular
24

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.

They used the

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]

2.3.4.4 Deflection Technique

A non-destructive method based on deflection has been used as a quantitative
measurement of exfoliation corrosion.[4, 8] In this method, the deflection of specimens
25

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.

The assumption of uniform

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.

In conclusion, exfoliation corrosion is a type of intergranular corrosion. The

mechanism of exfoliation corrosion is strongly related with that of IGC and intergranular
tress corrosion cracking. To understand exfoliation corrosion, it is essential to understand
these two types of corrosion.

In the following, intergranular corrosion and stress

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.

Galvanic couple theory.

2.

Precipitate free zone breakdown model.

3.

Anodic dissolution of grain boundary precipitates.

2.4.1.1 Galvanic Couple Theory

Dix et al. first proposed the galvanic couple theory.[60] According to this theory,
IGC susceptibility is attributed to the local galvanic corrosion between the noble grain
matrix and the anodic grain boundary regions. In Al-Mg and Al-Mg-Zn alloys, IGC is
considered to be a result of galvanic corrosion between anodic grain boundary particles
(Al8 Mg2, MgZn2) and the noble grain matrix.[38] In Al-Cu, Al-Cu-Mg and Al-Zn-Mg-Cu

27

alloys, IGC is explained as a result of galvanic corrosion between copper-depleted

(precipitate free) zones and the grain matrix.[38]
The galvanic couple theory was verified and developed by the work of Mears and
Brown.[38] They developed a technique to measure the corrosion potentials of the grain
boundary regions and the grain centers of large grained Al-4%Cu alloy in NaCl-H2O2.
The grain boundary regions were anodic relative to the grain bodies and the potential
difference reached a maximum after aging for about 4 to 8 hrs at 191 C. It was found that
the Al-4%Cu was most susceptible to intergranular attack when the corrosion potential
difference was a maximum. With extended aging, the precipitation within the grain
center began to approach that at the grain boundaries until complete precipitation
occurred and the difference in potential decreased to almost zero. The anodic grain
boundary regions were copper depleted zones. TEM studies by Hunter et al. successfully
related the path of IGC attack to microstructural features of Al-Cu-Mg alloy 2024.[61]
Brown et al. tried to explain IGC in Al-Cu based alloys on galvanic coupling
between the grain boundary constituents and matrix. They stated that the susceptibility of
peak-aged temper to IGC was due to the difference in corrosion potentials of the grain
boundary and the grain interior. Because of Cu depletion, the grain boundary region was
200 mV more active than the grain interior.[62]
Studies on Al-Zn-Mg alloys were also performed.

The potential difference

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.

2.4.1.2 Precipitate Free Zone Breakdown Model.

The galvanic couple mechanism does not explain why halide ions are necessary
for intergranular corrosion.

Galvele and DeMicheli proposed a second theory, the

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).

When the potential is higher than the PFZ

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

dependent on the artificial aging.

The anodic polarization curve in deareated NaCl

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.

They also stated that the

increased susceptibility of AA7075-T7351 to IGC is due to increased differences

between pitting potentials of different regions of the microstructure. The addition of
nitrates to the aqueous solution can increase this difference and hence increases the
susceptibility.[67]
Meng et al. studied the corrosion behavior of AA7xxx series aluminum
alloys.[68] Two breakdown potentials were found in Cu-containing 7xxx alloys in
deareated chloride solution.

They explained that the first breakdown potential

corresponded to transient dissolution associated with attack of the fine hardening

31

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]

2.4.1.3 Anodic Dissolution of Grain Boundary Precipitates

The anodic dissolution of grain boundary precipitates was proposed as an explanation
for intergranular stress corrosion and IGC behavior in Al-Li-Cu and Al-Mg alloys. This
theory suggests that the existence of a preferential anodic path along grain boundaries
results in anodic dissolution of grain boundary precipitates. In the wrought AA5xxx(AlMg) series alloys, it is generally accepted that Mg5Al8 particles can form a continuous
intergranular path, which is active to metal matrix [58] and is corroded preferentially.
Work on AA5083 alloys showed that IGSCC of the alloy was related to the anodic
dissolution of the continuous grain boundary Mg2Al3 when Mg content was greater than
32

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.

Anodic polarization of T1 phase and SHT AA2090 in

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.

The electrochemical roles of the grain boundary

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.

A clear understanding of the electrochemical

behavior of these microstructural factors will help us to understand EFC.

2.4.2.1 Intermetallic Particles

The addition of alloying elements to Al leads to the dispersion of various
intermetallic particles in grain bodies and grain boundaries. The formation of these
particles is due to the low solubility of alloying elements in aluminum. They are two
groups of the intermetallic particles. First group consists of the particles formed from the
melt during solidification and they usually contain impurities. The second group consists
of particles formed during heat treatment in heat treatable alloys. Heat treatable alloys
are strengthened by precipitation hardening. Heat treatment develops large differences in
the behavior of the grain boundary precipitate and precipitates free zones[73], hence
affecting the susceptibility to localized corrosion. Many papers show that the anodic
dissolution of intermetallic particles at the grain boundaries can lead to IGC and SCC.[43,
44] Al alloys that do not form intermetallic particles at grain boundaries, or those in
which the intermetallic particles have potential similar to the grain matrix, have high
resistance to localized corrosion. The electrochemical behavior of intermetallic particles
in aluminum alloy plays an important role on localized corrosion process.[74] The review
below will discuss the major intermatillic particles that exist in Al-Cu and Al-Zn alloys.
34

2.4.2.1.1

Intermetallic Particles in Al-Cu & Al-Cu-Mg Alloys

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

Intermetallic particles in Al-Zn-Mg & Al-Zn-Mg-Cu Alloys

The formation of Zn and Mg bearing intermetallic particles can create differing

electrochemical potentials in AA7xxx aluminum alloys.[32] It has been reported that eta
phase in 7xxx series alloys is active to the Al alloy matrix as it has an OCP of 1100mV
SCE in 1 M NaCl.[20] The electrochemical behavior of Mg (Zn, Cu, Al)2 in AA7150 was
studied by Ramgopal et al. by using thin film composition analogs prepared by flash
evaporation.[83] Since eta phase has considerable solubility for Cu and Al, the generic
composition can be considered to be Mg (Zn, Cu, Al)2.[84] The electrochemical behavior
of eta phase as a function of Cu and Al was studied in deareated neutral and high pH 0.5
M NaCl.

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

OCP playback experiments

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

Other Intermetallic Particles

Intermetallic particles in aluminum alloys commonly contain iron because iron

always exists in the alloys as an impurity. The influence of the Al3 Fe phase and various
39

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.

The first peak at around -1.15V

~ -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

The addition of Si to alloys reduced

significantly the selective dissolution rate of aluminum in Fe containing particles and

caused a negative shift in the corrosion potential. Mn has same effect as Si. Though the
dealloying of Fe-rich compounds catalyzes the cathodic reaction, which is detrimental to
the corrosion behavior of aluminum alloys, Mn or Si addition can suppress this reaction.
The Mg2Si phase is the basis for precipitation hardening in Al-Si-Mg series alloys.
This phase is anodic to aluminum and very reactive in acidic solutions with open circuit
potential of 1590 mV SCE.[65] In addition to Mg2Si, Al3 Fe, -Al-Fe-Mn-Si and MnAl6
are common particles present in Al-Si-Mn-Fe aluminum alloys. These particles are
slightly cathodic to aluminum and tend to increase the frequency of pitting.[2] MnAl6 has
a corrosion potential of 760 mV SCE in a NaCl-H2O2 environment, which is close to
that of the aluminum matrix.
Wrought materials, especially those containing more than 3wt% Mg, have Al8Mg5
and Al2 Mg3 distributed on grain boundaries and are susceptible to corrosion. Al3Mg2 has
a corrosion potential of 1.15V SCE and Al8Mg5 has a potential of 1.05mV SCE in 1M
NaCl-H2O2.[94]
At high Zr contents, primary crystals of the tetragonal phase Al3 Zr are easily
formed. This phase has been found to be cathodic to pure aluminum and supports
cathodic reactions.[77]

2.4.2.2 Precipitate Free Zones (PFZ)

Alloys that are age hardenable tend to produce PFZ parallel to grain boundaries.
The formation of PFZ is generally thought to be related to two mechanisms.[20] One is
due to the vacancy depletion in these areas giving too low diffusion rates for the
41

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

INTERGRANULAR STRESS CORROSION CRACKING (IGSCC)

According to Jones, Stress corrosion cracking is the brittle failure at relatively low

constant tensile stress of an alloy exposed to a corrosive environment.[16] Stress

corrosion cracking can be intergranular or transgranular. In contrast to the transgranular
cracking which is associated with mechanical fractures, IGSCC usually propagates along
electrochemically active ground boundary path. The IGSCC mechanism has been
proposed as one of the possible mechanisms of exfoliation corrosion.[1, 8, 9, 23] Hence,
a fundamental knowledge of the mechanism of IGSCC in high strength aluminum alloy
42

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

Three conditions must be present simultaneously to produce SCC:[16]

1.

Critical environment.

2.

A susceptible alloy.

3.

Some component of tensile stress.

The sequence of events involved in the SCC process is divided into three stages:[12]
1.

Crack initiation and propagation.

2.

Steady-state crack propagation.

3.

Final failure stage.

These three stages are shown in figure 2.12, a schematic diagram of typical crack

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

discontinuities, at pits, or from intergranular corrosion sites.[12] The mechanism of

IGSCC propagation has also been argued for many years, yet there is no one mechanism
universally accepted. General speaking, the mechanisms proposed for SCC propagation
can be classified into two basic categories:
1.

Anodic mechanism, which involve stress assisted anodic dissolution of an active

path.
2.

Cathodic mechanism, or hydrogen embrittlement.

2.5.1.1 Anodic Dissolution Mechanism

The anodic mechanism can be described as the active dissolution and removal of
material from the crack tip. Two main anodic dissolution models have been proposed.
The first is active path intergranular SCC model. In this model, IGSCC occurs by
preferential corrosion of electrochemical active path.

The second is film rupture

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

2.5.1.2 Cathodic Mechanism

Since hydrogen evolution is always observed during localized corrosion of Al
alloys, hydrogen embrittlement is accepted by the majority researchers as the cause of
intergranular stress corrosion cracking in Al-Zn-Mg-Cu system.[107]
Considerable research work indicated that the presence of cathodically charged
internal hydrogen could degrade the ductility of Al-Zn-Mg alloys.[118-121] Some
researchers suggested that the existence of hydrogen causes crack blunting[122],
reduction in cohesive strength[123], and microvoid nucleation[124]. Works have been
done to provide the evidence of hydrogen induced cracking in 7xxx alloys. Hardie et al.
found that embrittlement occurred in tensile tested AA7179-T651 by presoaking the
materials in 70o tap water for enough time and recovered by drying out in lab air or
vacuum.[125] Scamans observed crack arrest markings on the fracture faces of failed
DCB and C-ring samples of AA7071 and AA7078.[126] Chu et al. have studied the
dynamic processes of the nucleation and propagation of hydrogen induced delayed
cracking (HIDC) and SCC of AA7075. They observed the similar fracture features
between SCC and HIDC and concluded that the mechanism of SCC of high strength
aluminum alloys in aqueous solution was hydrogen induced delayed cracking.[127] Tong,
Lin and Hsiao found that the atomic binding energy was reduced by charging hydrogen
into pure aluminum[123] and water vapor accelerated the fatigue crack propagation rate
in pure aluminum due to the rapid diffusion of hydrogen through dislocations into the
interior of the crystal.[128]
The path of hydrogen diffusion into interiors of alloys has been investigated.
Albrecht et al. studied the SCC of AA7075 alloys and concluded that dislocations can
46

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

Effects of Microstructure and Alloy Chemistry on IGSCC

SCC susceptibility strongly depends on alloy microstructure.

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

Relationship between Stress Corrosion Cracking and Exfoliation

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.

Similarly, mechanical stress also affects the

Airplane wingskin plates experience both

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

In order for airplane maintenance to be based on need rather than time, a

corrosion model is required for predicting the corrosion extent under various conditions.
The last objective of this work is to develop a model that will predict how existing
corrosion will grow as a result of the total exposure conditions such as relative humidity
and stress.

52

REFERENCES

M.J. Robinson and N.C. Jackson, Corrosion Science, 41, 1013-1028 (1999).

T.J. Summerson and D.O. Sprowls, Corrosion Behavior of Aluminum Alloys, in

International Conference in Celebration of the Centennial of the Hall-Heroult
Process. 1986. University of Virginia, Charlottesville,Virginia.

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).

M.J. Robinson, Corrosion Science, 22(8), 775-790 (1982).

10

D.J. Kelly and M.J. Robinson, Corrosion, 49(10), 787-795 (1993).

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

G.S. Frankel, Journal of Electrochemical Society, 145(6), 2186-2198 (1998).

53

14

E.H. Hollingsworth and H.Y. Hunsicker, "Corrosion of Aluminum and Aluminum

Alloys", in Metals Handbook, Materials Park, OH: ASM International, 583-609
(1990).

15

S. Maitra and G.C. English, Metallurgical Transactions A, 12A, 535-541 (1981).

16

D.A. Jones, "Principles and Prevention of Corrosion", 2nd ed. Upper Saddle
River, NJ: Prentice-Hall, Inc. (1996).

17

H.P. Godard and W.E. Cooke, Corrosion, 16, 117-123 (1960).

18

L.L. Shreir, F.R.I. C, and F.I. M, "Corrosion", 2nd ed., ed., L.G.Shreir. NewnesButterworth (1976).

19

E. Ghali, "Aluminum and Aluminum Alloys", in Uhlig's Corrosion Handbook, ed.,

R.W. Revie, New York, NY: John Wiley & Sons, INC., 677-716 (2000).

20

E. Mattsson, L.O. Gullman, L. Knutsson, R. Sundberg, and B. Thundal, British

Corrosion Journal, 6, 73-83 (1971).

21

B.W. Lifka and D.O. Sprowls, Corrosion, 22, 7-15 (1966).

22

H.W. Pickering, F.H. Beck, and M.G. Fontana, Corrosion, 18, 230-239 (1962).

23

M.J. Robinson, Corrosion Science, 23(8), 887-899 (1983).

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

D. Adenis and A. Guilhaudis, Memoires Scientifiques de la Revue de Metallurgie,

64(10), 877-888 (1967).

27

K. Ebtehaj, D. Hardie, and R.N. Parkins, British Corrosion Journal, 24(3), 184188 (1989).

28

R. Ambat and E.S. Dwarakadasa, Corrosion Science, 33(5), 681-690 (1992).

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

W. Rosenkranz, Metall, 15(5), 894-899 (1961).

54

31

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.

32

H.A. Holl, Journal of the Institute of Metals, 97(200), 200-205 (1969).

33

M. Posada, L.E. Murr, C.-S. Niou, D. Roberson, D. Little, R. Arrowood, and D.

George, Materials Characterization, 38, 259-272 (1997).

34

B.W. Lifka, "Corrosion of aluminum alloys", in Corrosion Engineering

Handbook, P.A.Schweitzer, Marcel Dekker, New York: 129-133 (1996).

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

B.W. Lifka and D.O. Sprowls, "Significance of Intergranular Corrosion in High

Strength Aluminum Alloy Products", in Localized Corrosion : Cause of Metal
Failure. ASTM STP 516, p. 120 (1972).

39

J. Onoro and C. Ranninger, British Corrosion Journal, 28(2), 137-141 (1993).

40

W.A. Bell and H.S. Campbell, Journal of the Institute of Metals, 89, 464-471
(1960-1961).

41

A. Garner and D. Tromans, Corrosion, 35(2), 55-60 (1979).

42

T. Sheppard and N.C. Parson, Materials Science and Technology, 3, 345-352

(1987).

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

E. Tankins, J. Kozal, and E.W. Lee, Journal of Metals, 9, 40-43 (1995).

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

R. Braun, British Corrosion Journal, 30(3), 203-208 (1995).

53

H.B. Romans, Materials Research & Standards,, 31-34 (1969).

54

M. Keddam, C. Kuntz, H. Takenouti, D. Schuster, and D. Zuili, Electrochimica

Acta, 42(1), 87-97 (1997).

55

A. Conde and J. De Damborenea, Corrosion Science, 42, 1363-1377 (2000).

56

A. Conde and J. De Damborenea, Electrochimica Acta, 43(8), 894-860 (1998).

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

M.O. Speidel and M.V. Hyatt, "Stress-Corrosion Cracking of High-Strength

Aluminum Alloys", in Advances in Corrosion Science and Technology, eds., M.G.
Fontana and R.W. Staehle, New York-London: Plenum Press, p.118 (1972).

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

E.H. Dix Jr., Transactions, American Institute of Mining and Metallurgical

Engineers, 137, 11-40 (1940).

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

R.G. Buchheit, Journal of The Electrochenical Society, 142(11), 3994-3996

(1995).

66

K. Sugimoto, K. Hoshino, M. Kageyama, and Y. Sawada, Corrosion Science, 15,

709-720 (1975).

67

S. Maitra and G.C. English, Metallurgical Transactions A, 13A, 161-166 (1982).

68

Q. Meng and G.S. Frankel, Surface and Interface Analysis, 36, 30-42

69

V. Guillaumin and G. Mankowski, Corrosion Science, 41, 421-438 (1999).

70

V. Guillaumin and G. Mankowski, Corrosion Science, 42, 105-125 (2000).

71

T. Ramgopal, "Intergranular Corrosion of AA7XXX Alloys", MSE, The Ohio State

University: Columbus. (2001).

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

T. Ramgopal, P. Schmutz, and G.S. Frankel, Journal of The Electrochenical

Society, 148(9), B348-B356 (2001).

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

Katsuhiko Urushino and Katsuhisa Sugimoto, Corrosion Science, 19, 225-236

(1979).

79

B. Mazurkiewicz and A. Piotrowski, Corrosion Science, 23(7), 697-707 (1983).

57

80

R.G. Buchheit, M.A. Martinez, and L.P. Montes, Journal of The Electrochemical
Society, 147(1), 119-124 (2000).

81

P. Schmutz and G.S. Frankel, Journal of The Electrochemical Society, 145(7),

2285-2295 (1998).

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

M. Zamin, Corrosion, 37(11), 627-632 (1981).

86

E.I. Storchai and A.V. Turkovskaya, Protection of Metals, 6, 626 (1970).

87

K. Sugimoto and Y. Sawada, Journal of the Japan Institute of Metals, 37, 148
(1973).

88

R. Grauer and E. Wiedmer, Werkstoffe Korros., 31, 550 (1980).

89

A.P. Bond, G.F. Bolling, H.A. Domian, and H. Biloni, Journal of The
Electrochemical Society, 113, 773 (1966).

90

M.C. Reboul and J. Bouvaist, Werkstoffe Korros., 30, 700 (1979).

91

K. Nisancioglu, Journal of The Electrochemical Society, 137(1), 69-77 (1990).

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

G.T. Burstein and G.W. Ashley, Corrosion, 39(6), 241-247 (1983).

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

J. Embury and R. Nicholson, Acta Metallurgica, 13, 403-417 (1965).

96

N. Ryum, B. Haegland, and T. Lindtveit, Zeitschrift fur Metallkunde, 58, 28-31

(1967).

97

E.H. Dix Jr., in Physical Metallurgy of Aluminum Alloys, Cleveland: American

Society for Metals, 214 (1949).

98

S.J. Ketcham and F.H. Haynie, Corrosion, 19(7), 242t (1963).

99

S.J. Ketcham, Corrosion Science, 7(6), 305 (1967).

58

100

C. Kumai, J. Kusiniski, G. Thomas, and T.M. Devine, Corrosion, 45, 294 (1989).

101

D.A. Vermilyea, Journal of The Electrochemical Society, 119, 405-407 (1972).

102

H.L. Logan, Journal of Research of the National Bureau of Standards, 48, 99

(1952).

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

J.C. Scully, Corrosion Science, 15, 207 (1975).

106

R.W. Staehle, in the Theory of Stress Corrosion Cracking in Alloys, North

Atlantic Treaty Organization, p.223 (1971).

107

T.D. Burleigh, Corrosion, 47(2), 89-98 (1991).

108

M. Izu, M. Shiokawa, and E. Sato, Journal of the Japan Institute of Metals, 34(11),
643-649 (1984).

109

W. Gruhl, Zeitschrift fuer Metallkunde, 75(11), 819-826 (1984).

110

J.G. Rinker, M. Marek, and T.H. Sanders, Materials Science and Engineering,
64(2), 203-221 (1984).

111

E.I. Meletis, Materials Science and Engineering, 93(1-2), 235-245 (1987).

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

A.H. Le and R.T. Foley, Corrosion, 40(4), 195-197 (1984).

114

Y. Choi, H.-C. Kim, and S.-I. Pyun, Journal of Materials Science, 19(5), 15171521 (1984).

115

M.B. Hinz, Scripta Metallurgica, 19(12), 1445-1450 (1984).

116

A. Joshi, C.R. Shastry, and M. Levy, Metallurgical Transaction A: Physical

Metallurgy and Materials Science, 12A(6), 1081-1088 (1981).

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

R.J. Gest and A.R. Troiano, Corrosion, 30, 274-279 (1974).

120

T.F. Klimowicz and R.M. Latanision, Metallurgical Transaction A, 9A, 597-599

(1978).

121

J. Albrecht, A.W. Thompson, and I.M. Bernstein, Metallurgical Transaction A,

10A, 1759-1766 (1979).

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

C.P. You, M. Dollar, A.W. Thompson, and I.M. Bernstein, Metallurgical

Transaction A, 22A, 2445-2450 (1991).

125

D. Hardie, N.J.H. Holroyd, and R.N. Parkins, Materials Science, 12(11), 603-610
(1979).

126

G.M. Scamans, Metallurgical Transactions A: Physical Metallurgy and Materials

Science, 11A(5), 846-850 (1980).

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

J. Albrecht, I.M. Bernstein, and A.W. Thompson, Metallurgical Transactions A,

13A, 811-820 (1982).

130

M. Talianker and B. Cina, Metallurgical Transactions A: Physical Metallurgical

and Materials Science, 20A(10), 2087-2092 (1989).

131

A.W. Thompson, M.A. Mueller, and I.M. Bernstein, Metallurgical Transactions A,

24A, 2569-2575 (1993).

132

M.O. Speidel, "Interaction of Dislocations with Precipitates in High Strength

Aluminum Alloys and Susceptibility to Stress Cracking", in Fundamental Aspects
of Stress Corrosion Cracking, ed., R.W.Staehle, 561-579 (1967).

133

K.G. Kent, Journal of the Institute of Metals, 97, 127-128 (1969).

134

A.J. De Ardo Jr. and R.D. Townsend, Metallurgical Transactions, 1, 2573-2581

(1970).

135

M. Puiggali, A. Zielinski, J.M. Olive, E. Renauld, and D. Desjardins, Corrosion

Science, 40(4/5), 805-819 (1998).
60

136

L. Christodoulou and H.M. Flower, Acta Metallurgica, 28, 481-487 (1980).

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

J.R. Pickens and T.J. Langan, Metallurgical Transactions A: Physical

Metallurgical and Materials Science, 18A(10), 135-149 (1987).

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. 10 Variation in corrosion potential of FeAl3 and -Al(Fe,Mn)Si.[91]

(Reproduced by permission of The Electrochemical Society, Inc.)
67

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

QUANTITATIVE STUDY OF EXFOLIATION CORROSION: EXFOLIATION

OF SLICES IN HUMIDITY TECHNIQUE

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.

In these methods, salt solutions are alternately

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.

The grain boundary region of a typical Al alloy can

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]

Evans proposed the possibility of an

electrochemical cell occurring between Mn-rich and Mn-depleted regions.

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

Test samples were machined from a piece of a wingskin of a retired KC135

airplane.

The wingskin section contained two AA7178-T6 plates attached to an

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

thickness cross section was divided into 15 zones.

In each zone, 30 grains were

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

Exfoliation of Slices in Humidity (ESH) Technique

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

upon subsequent exposure to high humidity.

The EFC started at the outer edges,

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

determine the depth of metal consumed by EFC: d i = d 0 d i , where i=1,2,.day.

Since the exfoliation occurred at the two edge surfaces, one half of the width change
d i
represents the average EFC depth.
2

3.3

RESULTS AND DISCUSSION

3.3.1

Electrochemical and ESH Characterization

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.

It is interesting that zones of varying EFC

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

Analytical transmission electron microscopy (TEM) was used to investigate the

composition of the grain boundary constituents in the good plate. Figure 3.10 shows
transmission electron micrographs of samples taken from near the up and downsides of
the good plate. Both micrographs exhibit particles in the matrix as well as on the grain
boundaries and a precipitate free zone (PFZ) on either side of the grain boundaries. The
width of the PFZ near the upside of the good plate was around 90~100 nm, which is
somewhat larger than the PFZ width near the downside, less than 50 nm. There was no
clear difference in the size, distribution, and composition of the large precipitates in the
matrix near the two surfaces. EDS line profiling was performed across grain boundaries
at locations that include precipitates and locations away from precipitates to analyze the
PFZ. Four different precipitates and five different PFZs along grain boundaries near
upside and downside of the good plate were analyzed. Figure 3.11 and 3.12 show
representative precipitates and PFZs that exist near upside and downside of the plate.
The lines in the higher resolution images in these figures show the paths of the line
profiles in one measurement. The concentration profiles in the form of concentration
ratios for various elements relative to Al are shown in figure 3.13 for the line-scans
across the grain boundary precipitates in figure 3.11, and in figure 3.14 for the line scans
across the grain boundary precipitate free zones in figure 3.12. The concentration ratios
were determined using the appropriate k factors and the x-ray intensity ratios. Two kinds
80

of precipitates are found along grain boundaries near both upside and downside of the
plate.

One is Al-Zn-Mg-Cu-Cr precipitate (concentration profiles across these

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.

This element has high diffusion rates and low

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.

A high susceptibility to EFC was associated with a high Zn content in the

precipitate free zone at grain boundaries.

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).

B.W. Lifka and D.O. Sprowls, Corrosion, 22, 7-15 (1966).

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).

R. Braun, British Corrosion Journal, 30(3), 203-208 (1995).

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

M.J. Robinson, Corrosion Science, 22(8), 775-790 (1982).

15

T. Ramgopal, P.I. Gouma, and G.S. Frankel, Corrosion, 58(8), 687-697 (2002).

16

S. Maitra and G.C. English, Metallurgical Transactions A, 12A, 535-541 (1981).

17

T. Ramgopal, P. Schmutz, and G.S. Frankel, Journal of The Electrochemical

Society, 148(9), B348-B356 (2001).

18

E. Mattsson, L.O. Gullman, L. Knutsson, R. Sundberg, and B. Thundal, British

Corrosion Journal, 6, 73-83 (1971).

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

Q. Meng and G.S. Frankel, Journal of The Electrochemical Society, 151(5),

B271-B283 (2004).

22

X. Zhao, Chapter 4 of this dissertation.

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

F. Sato and R.C. Newman, Corrosion, 54(12), 955-963 (1998).

26

F. Sato and R.C. Newman, Corrosion, 55(1), 3-9 (1999).

27

P.L. Bonora, G.P. Ponzano, and V. Lorenzelli, British Corrosion Journal, 9(2),
108-111 (1974).

28

T. Ramgopal and G.S. Frankel, Corrosion, 57(8), 702-711 (2001).

85

TABLES AND FIGURES

Salt

Expected %RH (from CRC)

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

Humidity associated with saturated salt solutions.[20]

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

Potential (mV SCE)

-700
-800
-900
-1000
"bad" plate

-1100
-1200 -11
10

10

-9

10

-7

10

-5

0.001

0.1

Current Density (A/cm )

Figure 3.3
Polarization curves of AA7178 good and bad sample in deareated 1.0
M NaCl at a scan rate of 0.1 mV/s.

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

Average EFC Depth (mm)

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

Average EFC Width (mm)

"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

Grain Size (um)

600

T
L

500
400

L/S

Grain Aspect Ratio T/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

Position Through Thickness (4.1mm)

(a)
700
S

Grain Size (um)

600

T
L

500
400
300

16.2
8.7

Grain Aspect Ratio

20.0 18.5 19.6 21.6
7.6
L/S
6.6 6.7
8.0 17.4 16.9
T/S
16.9 6.0
22.1
7.6
14.5 14.6
20.7
5.2 5.4
8.0
12.2
5.1 15.6
5.9

17.0
17.2
7.4 6.8

200
100
0
T1/15

T8/15

upside

T15/15
downside

Position Through Thickness (4.6mm)

(b)
Figure 3.9
Grain size distribution and grain aspect ratios through the thickness of (a)
good plate and (b) bad plate.
94

(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

Concentration Ratio Profile (C i/CAl)

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

Concentration Ratio Profile (C i/CAl)

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

Concentration Ratio Profile (C i/CAl)

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

EFFECTS OF RH, TEMPER AND STRESS ON EXFOLIATION CORROSION

KINETICS OF AA7178

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

is important to understand the EFC of aircraft materials including the effects of

microstructure, environment and stress on the corrosion progress.
Exfoliation corrosion is affected by environmental conditions such as humidity to
a marked degree.[6, 7] In this work, EFC tests were performed in different humidities to
simulate changing RH found under real conditions.

Predictive models for EFC of

airplane wingskins also require a quantitative understanding of the effects of stress

because aircraft wings are subjected to both compressive and tensile stress, whether on
the ground or in flight. Exfoliation tests were carried out in this study using a four-point
bending apparatus to study the effect of stress on EFC.
Susceptibility of AA7178 materials to exfoliation corrosion is influenced by heat
treatment. The resistance to exfoliation corrosion is usually improved in alloys with
extended aging time.[8-12] For instance, the EXCO test on 7xxx alloys show that the T7
temper is more resistant to EFC than the T6 temper.[13-15] However, for some alloys,
the EXCO test is too corrosive which markedly reduces the ability to distinguish between
all the various commercial tempers.[14] Exfoliation tests were carried out on T7 and T6
samples to investigate the effect of temper on exfoliation corrosion.

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 longitudinal orientation of the microstructure (along the rolling direction).

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

sections of the T6 and T7 temper were metallographically cross-sectioned. The samples

were polished to 1 m and etched in Kellers reagent.
Electrochemical polarization measurements were performed on samples ground to
1200 grit in ethanol. 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.
The Exfoliation of Slices in Humidity (ESH) test was used to measure EFC
kinetics. Slices of Al alloy plates prepared as described above were electrochemicallypretreated and then exposed to a high relative humidity environment where exfoliation
occurred. In earlier work, a potentiostatic pretreatment in chloride solution was used to
104

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

pretreatment current density was selected based on the potentiodynamic polarization

curve measured on AA7178-T6 in 1 M NaCl, Figure 4.3. The polarization curve of
AA7178-T6 sample exhibited two breakdown potentials, and the second breakdown
potential was found to be about -722 to -730 mV SCE. A nominal current density
associated with a potential just above this breakdown potential was desired to get
sustained intergranular corrosion.[18] A current density of 5.5 mA/cm2 was selected, as it
results in a potential between 0.69 and 0.72 V SCE. The electrochemical pretreatment
for the ESH test was thus 7 hours in 1 M NaCl at 5.5 mA/cm2.
Following the pretreatment, the sample was rinsed with DI water, dried with a
stream of air and placed in a humidity chamber, consisting of a sealed beaker containing
a saturated salt solution at room temperature (22-25C). Several salts were used to create
a range of constant humidity: sodium sulfate (K2SO4), ammonium chloride (NH4Cl),
potassium iodide (KI), and magnesium nitrate (Mg(NO3)2) for creating constant relative
humidity of 96%, 76%, 66%, 56%, respectively.
105

During RH exposure, digital

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

1,2,30 positions. The mean value is calculated by: d i =

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

represents the average amount of material consumed on the two sides.

X-ray photoelectron spectrometry measurements were also performed on the
corroded transverse surface of a sample exposed to 56%RH after electrochemical
pretreatment using an AXIS Ultra spectrometer controlled by a VISON data system. A
monochromatic Al Ka X-ray line with energy of 1300 eV and 130 W was used as the
incident radiation source. The binding energies of the measurements of interest were
calibrated with respect to the C 1s line at 284.5 eV.

106

4.2.1

Four Point Bending Technique

The deflection of specimens under four point loading conditions is a function of

the dimensions of the sample and stressing jig.[8] Figure 4.4 is a schematic figure of a
symmetrically loaded specimen supporting four concentrated loads P. The geometry of
the specimen is represented in the figure. From elastic theory, the deflection under
four point loading between two loading points A and B is given by [19]:

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

RESULTS AND DISCUSSION

4.3.1

Effect of RH on EFC Kinetics

Samples cut from the as-received AA7178 wingskin plate were ESH tested in a range
of RH.

Figure 4.7 shows the corrosion morphology after the electrochemical

pretreatment at 5.5mA/cm2 in 1 M NaCl for 7 hours. The corrosion morphology appears

to be a combination of IGC and selective grain attack.[22] To check the depth of attack,
109

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.

The results were rather reproducible for samples pretreated

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.

The critical humidity for partially hydrolyzed

111

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.

To study these factors, samples were exposed to cyclic RH conditions.

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.

The extent of EFC was

This indicates that the EFC

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

Effect of Temper on EFC Kinetics

Metallographic cross sections show that the microstructures of various sections of

AA7178, i.e. the grain size and shape, are similar for the T6 and T7 temper and there was
no recrystallization during heat treatment. Figure 4.16 shows digital photographs of T6
and T7 tempers on the 6th, 11th and 17th day of exposure in 96%RH. Less corrosion
product is evident on the T7 sample surface than on the T6 sample. The width change of
the unattacked region of these samples was determined, and the EFC rate for T7 was
found to be lower than for T6, figure 4.17. The kinetics of the T7 temper in 96%RH was
a little faster than for T6 temper in 76%RH. These ESH test results are consistent with
reported EXCO test results, which also indicated that the T6 temper is more susceptible
to EFC than T7.[13-15]

4.3.3

Effect of Stress on EFC Kinetics

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.

The exfoliation rate at the compressive side increased with

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).

The acceleration of exfoliation on the

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

Empirical Model for Exfoliation Corrosion

Accelerated corrosion testing helps corrosion engineers to obtain information about

corrosion susceptibility of materials within a compressed time. To estimate the corrosion
kinetics at normal stress levels based on the accelerated life testing data, extrapolation is
required. (Note that the term stress here is used generically to represent all parameters
that might accelerate degradation, including mechanical, environmental and thermal
factors.) In accelerated life testing, the nominal life-time is often related to stress levels
by an acceleration equation that predicts the service lifetime of the material or component
based on the lifetime under stress and the level of the stress.[27] In the following, the
effects of time, RH, and mechanical stress will be considered in turn.
It was shown that EFC would propagate above a critical humidity of 56%RH. The
high humidity assures that a wetted film forms and remains on the aluminum surfaces
throughout the test period. As shown in figure 4.11, ignoring the data for the first few
days, the EFC kinetics were found to be approximately constant with time, so the depth
increased linearly with time. The first days seem to be a transition period similar to that
shown in the cyclic test experiments in figure 4.15. Therefore, the ESH data following
the transition period were fitted to d = Kt, resulting in K values of 0.2, 1.2, and 6.2
m/day for 66, 76 and 96% RH, respectively.
An empirical model including RH effects can be developed using the Eyring model of
acceleration. The Eyring model has been utilized to model acceleration process in many
fields[27-32] as this model has a solid theoretical foundation in chemistry and quantum
mechanics.[28] When thermal stress (temperature) is the main acceleration variable in the
115

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)

where A, B, C and are constants, H is the enthalpy change, and k is Boltzmanns

constant. Additional stresses can be added into this equation with more terms in the
exponential.
In this work, temperature was assumed to be constant during humidity exposure.
So exp(

H
) can be included into parameter A. The Eyring model can be simplified to
kT

the following form, including the time dependence:

d = A t exp(B RH)

(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

when = 0 (Unstressed condition)

C3 exp( E )

when > 0 (Tension)

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)

When RH = 0.96, B = 6.44, so:

d = t exp(E ' )

(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

followed a linear relationship with time.

2.

The effect of temper was quantitatively measured, and the EFC rate of AA7178-

T6 was found to be higher than that of AA7178-T7.

3.

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.

An empirical model was developed to describe the combined effects of RH and

stress. The model is more accurate at longer time and higher humidity.

118

REFERENCES

M. Posada, L.E. Murr, C.-S. Niou, D. Roberson, D. Little, R. Arrowood, and D.

George, Materials Characterization, 38, 259-272 (1997).

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).

T.J. Summerson and D.O. Sprowls, Corrosion Behavior of Aluminum Alloys, in

International Conference in Celebration of the Centennial of the Hall-Heroult
Process, 1986, University of Virginia, Charlottesville,Virginia.

E.A.G. Liddiard, J.A. Whittaker, and H.K. Farmery, Journal of the Institute of
Metals, 89, 377-384 (1960-1961).

B.W. Lifka and D.O. Sprowls, Corrosion, 22, 7-15 (1966).

10

W.A. Bell and H.S. Campbell, Journal of the Institute of Metals, 89, 464-471
(1960-1961).

11

H.B. Romans, Materials Research & Standards,, 31-34 (1969).

12

D.J. Kelly and M.J. Robinson, Corrosion, 49(10), 787-795 (1993).

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

M.O. Speidel and M.V. Hyatt, "Stress-Corrosion Cracking of High-Strength

Aluminum Alloys", in Advances in Corrosion Science and Technology, eds., M.G.
Fontana and R.W. Staehle, New York-London: Plenum Press, p.118 (1972).

17

X. Zhao, Chapter 3 of this dissertation.

18

Q. Meng and G.S. Frankel, Journal of The Electrochemical Society, 151(5),

B271-B283 (2004).

19

J.M. Gere and S.P. Timoshenko, "Mechanics of Materials", Fourth Edition,

Boston, MA: PWS Publishing Company (1997).

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

T. Huang and G.S. Frankel, "Sharp Intergranular Corrosion Fissures in AA7178",

Submitted to Corrosion Science, 11/05.

23

N. LeBozec, D. Persson, A. Nazarov, and D. Thierry, Journal of Electrochemical

Society, 149(9), B403-B408 (2002).

24

H.W. Pickering, F.H. Beck, and M.G. Fontana, Corrosion, 18, 230-239 (1962).

25

M.J. Robinson, Corrosion Science, 22(8), 775-790 (1982).

26

M.J. Robinson, Corrosion Science, 23(8), 887-899 (1983).

27

P. Qiu, The Indian Journal of Statistics, 62, 223-235 (2000).

28

S. Glasstone, K.J. Laidler, and H. Eyring, "Electrochemical Processes", in The

Theory of Rate Processes, Frick Chemical Laboratory, Princeton University:
McGRAW-HILL BOOK COMPANY, Inc., 552-599 (1941).

29

G.C. Montanari and A. Motori, Journal of Physics D: Applied Physics, 28, 14331438 (1995).
120

30

G.C. Montanari and A. Motori, Journal of Applied Physics D: Applied Physics,

24, 1172-1181 (1991).

31

G. Mazzanti, G.C. Montanari, and A. Motori, Journal of Applied Physics D:

Applied Physics, 27, 2601-2611 (1994).

32

L. Bitman, Y.T. Choi, S.B. Choi, and N.M. Wereley, Smart Materials and
Microstructures, 14, 237-246 (2005).

33

P. Tobias, "Assesing Product Reliability", in NIST/SEMATECH e-Handbook of

Statistical
Methods,
http://www.itl.nist.gov/div898/handbook/,
Statistical
Engineering Division,

34

B. Reich and E.B. Hakim, Solid State Technology, September, 65-66 (1972).

121

TABLES AND FIGURES

Alloy

Si

Fe

AA7178 0.0392 0.175

Table 4. 1

Element

Cu

Mn

Mg

2.04

0.0125 2.71

Cr

Zn

Ti

Al

0.187

6.46

0.0212 Bal.

Chemical composition of AA7178.

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 exits in two different

Cl-

198.1

1.292

46.4

Cl 2p3/2

chemical states

Cl-

199.9

1.010

20.6

Table 4. 2

Data obtained from XPS spectra.

122

Figure 4.1

Heat treatment of AA7178 wingskin plate.

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

Potential (mV SCE)

-800
-1000
-1200
-1400
-1600 -11
10

10-9

10-7

10-5

10-3

10-1

Current Density (A/cm )

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

Four point bending mode.

124

Figure 4.5

Four point bending jig.

Figure 4.6

Four point bending test set up.

125

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

Half of Width Change

of Unattacked Region (mm)

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

X-Ray Energy (KeV)

Figure 4.13 X-ray EDS spectrograms of corrosion product in 56%RH. Accelerating
voltage=12kV.

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

Binding Energy (eV)

(a)

202

204

206

6500

O 1s

Zn 2p

600

Intensity (Counts)

Intensity (Counts)

200

(b)

800
700

198

Binding Energy (eV)

500
400
300
200

6000

5500

5000

100
0
520

525

530

535

540

4500
1010

545

1020

Binding Energy (eV)

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

Bending Energy (eV)

Intensity (Counts)

(c)
4

1030

Binding Energy (eV)

10

930

940

950

960

970

980

45

50

55

60

Binding Energy (eV)

Binding Energy (eV)

(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

Half Width Change

of Central Unattacked Region (mm)

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

Results of cyclic exfoliation test for AA7178 wingskin samples.

133

(a)

(b)
Figure 4.16 Image of AA7178 sample exposed to 96%humidity following
galvanostatic pretreatment. (a) T6 temper. (b) T7 temper.
134

Half of Width Change

of Unattacked Region (mm)

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

Exfoliation corrosion kinetics of AA7178 samples exposed to 96% RH.

135

Figure 4.18 Image of AA7178 sample under four point bending loading exposed to
96%humidity.

136

0.3
-0.006
-0.005

EFC Depth (mm)

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

Exfoliation corrosion kinetics under compression and tension.

137

Sum of EFC Extent of Both Sides (mm)

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

Exfoliation corrosion kinetics under different strain conditions.

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

EFC Depth on Tension Side (mm)

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

Average kinetics of exfoliation corrosion in 96%RH under tension.

140

CHAPTER 5

IN SITU X-RAY RADIOGRAPHY OF LOCALIZED CORROSION

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.

IGC and EFC are significant sources of life limiting degradation in

A great deal of effort has been devoted over the years to the study of

intergranular and exfoliation corrosion of aluminum alloys.

technique recently has been used to quantify IGC growth rate.[1]

The foil penetration

When performed on

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

technique. It is clearly of interest to obtain information regarding the full population of

corrosion sites, not just the fastest growing sites. Exfoliation is a form of intergranular
corrosion that can occur on the surface of wrought aluminum alloys with elongated grain
structure.[4] However, few reports of quantitative measurements of the kinetics of
exfoliation in Al alloys exit, and little is known about how EFC develops under a wide
range of environmental conditions.[5-7]
Predictive models of the effects of corrosion on the structural integrity of aircraft
require information and prediction of the growth kinetics of the various forms of
localized corrosion. X-ray radiography technology is one approach for providing this
information. The essence of radiography is modulation of the x-ray intensity due to
discontinuities or thickness change caused by corrosion.[8] Owing to the relatively low
density of water, x-ray images can be obtained on metal samples in situ, i.e. immersed in
solution. Radiography can of course be performed in air so, unlike electrochemical
techniques, it can be used on samples exposed to atmospheric conditions, which are more
related to the environment of aircraft components. In

this

work,

in

situ

x-ray

radiography was used to address localized corrosion kinetics and morphology of

AA2024-T3 and AA7178 in different orientations relative to the rolling direction. By
orientating the sample properly, based on the understanding generated in prior work, it is
possible to image defects as they are growing down into a sample.

5.2

EXPERIMENTAL

The schematic of the developed micro-radiographic experimental setup for micro

radiography of corroded samples is shown in figure 5.1. It is located in the Welding
142

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.

The exposed and processed films were back-illuminated and digitized by

illuminating a CCD camera and a Data Translation frame grabber (DT3155).[ 8, 9]

Two different Al alloys, AA2024-T3 of thickness 1.9cm and AA7178 8.9 mm
thick plate removed from the wingskin of a retired KC135 airplane were tested. Their
compositions are shown in Table 5.1. Figure 5.3 shows the anisotropic microstructure of
the two materials and terminology used for the three different sections. The L section is
perpendicular to L direction, which is along the rolling direction. The S section is
perpendicular to S direction, which is through the thickness of the plate. Special cells
and sample configurations were developed in this study to allow the real time in situ
radiography. Pillars of dimension 0.8 mm x 2 mm x 30 mm were cut from AA2024-T3
such that the long dimension of the sample was oriented along the plate rolling direction
(longitudinal, L), the intermediate dimension was along the plate through-thickness
direction (short transverse, S), and the short dimension was along the plate transverse
direction (long transverse, T). This orientation allowed for IGC or EFC growth in the L
direction. Some samples were oriented such that their long dimension was along the T
direction and the short dimension was along the L direction, allowing IGC or EFC
growth in the T direction. Some samples were machined to have a square 2 mm x 2 mm
143

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

RESULTS AND DISCUSSION

Before showing the results of the real time radiography, an example of

metallographic sectioning results will be given.

Figure 5.4 shows an example of a

metallographic section for a 3 mm thick cylindrical AA2024-T3 sample with axial

orientation in the S direction after a 4 h exposure at -580 mV SCE in 1 M NaCl. The
attack was in the L-T plane of the sample, i.e. along the grain boundaries. Samples were
exposed for different times and sectioned. The kinetics determined from the longest site
in each section were slower than those found for L or T oriented samples determined by
the foil penetration technique, which measures the fastest growing site.[1] This proves
that metallographic sectioning does not determine the kinetics of the fastest growing sites.
Figure 5.5 shows an in situ x-ray radiograph of a 2 mm thick AA2024-T3 sample
oriented such that the L direction was vertical. The sample was encased in epoxy with
one face, the L face, exposed to the solution. This sample was exposed for 19 h to the
same conditions as the last sample. The dark lines in the image are corroded grain
boundaries. The depths of several selected growing sites were measured on the films at
various times. Analysis of a series of images taken from this sample at varying exposure
times resulted in the data shown in figure 5.6. The depth of specific IG sites is shown as
145

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

delaminating at the edge of the sample, which is characteristic of EFC.

Figure 5.9 shows an in situ x-ray radiograph of a 2 mm thick AA7178 sample
oriented such that the L direction was vertical. The sample was encased in epoxy with
one face, the L face, exposed to the solution. This sample was exposed for 3 h at 725mV SCE in 1 M NaCl. The growth kinetics of selected corrosion sites along the L
direction were measured from in situ radiographs recorded during the whole 7 hr
potentiostatic treatment, and it was found that the general law d=Atn was obeyed, as
shown in figure 5.10. The value of n was about 0.33 and A was 0.23-0.34 for t hr and d
in mm. This is consistent with results from the foil penetration method.[10]
Figure 5.11a shows an in situ x-ray radiograph of the same AA7178 sample with
long axis oriented in the L direction. This radiograph was taken after 7 h exposure at 725 mV SCE in 1 M NaCl. After 7 hours in 1 M NaCl at 725mV SCE, the sample was
rinsed with DI water and placed in 96%RH. Sharp IGC was evident in radiographs after
147

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

electrochemical treatment (7 h at -710 mV SCE in 1 M NaCl), IGC is evident, as was

found for the free-standing AA2024-T3 sample after short times.

This sample was

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

The intergranular nature of the attack in AA2024-T3 generated different

radiographs depending on the orientation of the sample and the x-rays relative to the
elongated microstructure. IGC appears as broad and diffuse features when looking in the
148

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).

F. Hunkeler and H. Bohni, Corrosion, 37(11), 645-540 (1981).

F. Hunkeler and H. Bohni, Corrosion, 40(10), 534-540 (1984).

"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).

M.J. Robinson, Corrosion Science, 22(8), 775-790 (1982).

D.J. Kelly and M.J. Robinson, Corrosion, 49(10), 787-795 (1993).

B. Zoofan and S.I. Rokhlin, Materials Evaluation, 52(2), 191-194 (1998).

A. Sehgal, G.S. Frankel, B. Zoofan, and S.I. Rokhlin, Journal of The

Electrochemical Society, 147(1), 140-148 (2000).

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

R. Fitzgerald, Physics Today, 53, 23-26 (2000).

14

A. Singirev and I. Snigireva, Review of Scientific Instruments, 66(12), 5486-5492

(1995).

151

TABLES AND FIGURES

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

Composition of AA2024 and AA7178 plates

152

Figure 5.1

Schematic of the microradiographic system.

Figure 5.2

In situ x-ray microradiographic experimental setup

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

Same sample as figure 5.7, taken in T direction. Exfoliation is evident.

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

CONCLUSIONS AND FUTURE WORK

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.

It has also generated a better understanding on how

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.

In the AA7178 wingskin sample studied, a high

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.

A specimen under four point bending condition is under

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

exfoliation rate on a bent sample was independent of strain because the

acceleration effect of compression stress was offset by the opposite effect of
tension stress on the other side.
5. An empirical model for EFC propagation was developed by extrapolating the
experimental data from ESH tests in order to describe the combined effects of RH
and stress. An Eyring relationship was used. The model is more accurate at
longer time and higher humidity.
6. Exfoliation corrosion is a particular kind of intergranular corrosion, which is
observed on the surface of the wrought materials. To study the correlation of
intergranular corrosion and exfoliation corrosion, in-situ X-ray radiography was
utilized. The results on AA2024 and AA7178 showed that in situ radiography of
rectangular samples generated information about the growth kinetics of a large
number of sites with a range of penetration rates, not just the fastest sites that can
be only measured by foil penetration technique. In situ X-ray radiography is a
good approach in studying samples exposed to different controlled conditions
such as immersed condition or humidity exposure conditions.

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

mechanism of exfoliation corrosion. The following procedures may be followed as an

extension to this study.
1. In this study, several relative humidity: 30%, 56%, 66%, 76%, 96% were used to
investigate the threshold humidity for propagation of EFC in AA7178.

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

4. This work confirmed the effect of T6 and T7 temper on exfoliation corrosion.

The quantitative effect of other heat treatment can be evaluated by using ESH
technique.
5. An empirical model of exfoliation corrosion was developed in this study. This
model describes a relationship between the exfoliation corrosion kinetics and
intergranular sharp fissure growth. To realize the model, a numerical simulation
is necessary. Understanding of the development of wedging stress is essential to
perform this simulation.

170

APPENDIX A

FINITE ELEMENT SIMULATION

A.1

PHENOMENOLOGICAL MODEL OF EXFOLIATION CORROSION

A phenomenological model is proposed to explain EFC and IGC process in this

work.

Figure A.1 illustrates a 3D rectangular sample after an electrochemical

pretreatment: selective grain attacks are created during pretreatment.

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

dimension of 2c0=[b-2a(tpre)], where b is the original sample dimension in the T direction.

Subsequent exposure to humidity results in the growth of IGC from these sites. IGC
171

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

MODEL DEVELOPMENT IN ABAQUS

A finite element model was developed and analyzed using the commercial finite
element program ABAQUS.

The meshes consist of 8879 4-node-plane-strain

quadrilateral, reduced integration elements. Meshes of a quarter of the L section of ESH

sample are shown in figure A.3. The model consisted of 5 cracks opened at one side of
the section, each of which has one-grain size in width and 400um in length. In each
crack, a U shape rigid body was inserted to simulate the corrosion product. At room
temperature, the AA7178 has an elastic modulus E=71.7GPa, Poissons raio =0.33, and
172

materials density = 2.8E-9g/cm3. Stressstrain relationship in plastic zone is determined

by the Holloman equation and the calculation is shown as follows:
For

AA7178-T6,

yield

stress

y = 538MPa.

Ultimate

tensile

stress UTS = 607 MPa . Elongation at breaking e f = 0.11

Assume: = 0 + k ( p ) n for 0

(a)

Let 0 = y
To determine k and n, two conditions were used as following:
1.

fp

= f where f is true stress at the necking which is given by

f = UTS (1 + e f )
2.

fp

= f

For AA7178-T6 at the necking: f = 538(1 + 0.11) = 597 MPa

Elastic true strain: ef =

f
E

= 0.008326 .

Total true strain: f = ln(1 + e f ) = 0.104

Plastic strain: fp = f ef = 0.09567
For condition 1, we get:
k ( fp ) n = f 0

(1)

For condition 2, we get:

nk ( fp ) n 1 = f

(2)

Equation (2) divided by equation (1), we get

173

n=

f
f 0

fp

(3)

Substituting equation (2) into equation (1), we get:

k=

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

RESULTS AND DISCUSSIONS

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

( x y ) 2 + ( y z ) 2 + ( z x ) 2 + 6( xy2 + yz2 + zx2 )

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

T. Huang and G.S. Frankel, "Effect of Temper and Potential on Localized

Corrosion Kinetics of AA7075", Submitted to Corrosion, 11/05.

T. Huang and G.S. Frankel, "Sharp Intergranular Corrosion Fissures in AA7178",

Submitted to Corrosion Science, 11/05.

176

TABLES AND FIGURES

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

Stress strain input data.

177

Figure A. 1

IGC attacks are created in the sample from electrochemical pretreatment.

178

(a) Electrochemical pretreatment creates IGC attacks.

(b) IGC continues growing during RH exposure.

(c) IGC grows fastest at the edge of sample.

(d) Layers of materials exfoliate.

Figure A. 2

EFC process of a sample exposed in high relative humidity. (L section)

179

Figure A. 3

Two dimensional finite element mesh.

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

R. Braun, British Corrosion Journal, 30(3), 203-208 (1995).

R. H. Brown, W. L. Fink and M. S. Hunter, Transactions, American Institute of Mining
and Metallurgical Engineers, 143, 115-123 (1941).
R. G. Buchheit, Journal of The Electrochenical Society, 142(11), 3994-3996 (1995).
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).
R. G. Buchheit, M. A. Martinez and L. P. Montes, Journal of The Electrochemical
Society, 147(1), 119-124 (2000).
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).
R. G. Buchheit, J. P. Moran and G. E. Stoner, Corrosion Science, 46(8), 610-617 (1990).
R. G. Buchheit, J. P. Moran and G. E. Stoner, Corrosion Science, 50(2), 120-130 (1994).
R. G. Buchheit, F. D. Wall, G. E. Stoner and J. P. Moran, Corrosion, 51(6), 417-428
(1995).
T. D. Burleigh, Corrosion, 47(2), 89-98 (1991).
G. T. Burstein and G. W. Ashley, Corrosion, 39(6), 241-247 (1983).
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).
V. E. Carter and H. S. Campbell, Journal of the Institute of Metals, 89, 472-475 (1961).
G. S. Chen, M. Gao and R. P. Wei, Corrosion, 52(1), 8-15 (1996).
Y. Choi, H.-C. Kim and S.-I. Pyun, Journal of Materials Science,19(5), 1517-1521 (1984).
L. Christodoulou and H. M. Flower, Acta Metallurgica, 28, 481-487 (1980).
W.-Y. Chu, Y.-B. Wang and C.-M. Hsiao, Corrosion, 38(11), 561-570 (1982).
J. P. Chubb, T. A. Morad, B. S. Hockenhull and J. W. Bristow, International Journal of
Fatigue, 17(1), 49-54 (1995).
A. Conde and J. D. Damborenea, Electrochimica Acta, 43(8), 894-860 (1998).
A. Conde and J. D. Damborenea, Corrosion Science, 42, 1363-1377 (2000).

183

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).
A. J. Cornish and M. K. B. Day, Journal of the Institute of Metals, 97, 44-52 (1969).
J. D. Damborenea and A. Conde, British Corrosion Journal, 30(4), 292-296 (1995).
J. R. Davis, Metals Handbook, Materials Park, OH: ASM International (1998).
J. R. Davis, Corrosion of Aluminum and Aluminum Alloys, Materials Park, OH: ASM
International (1999).
T. J. Davis, D. Gao, T. E. Gureyev, A. W. Stevenson and S. W. Wilkins, Nature, 373(16),
595-598 (1995).
K. Ebtehaj, D. Hardie and R. N. Parkins, British Corrosion Journal, 24(3), 184-188
(1989).
J. Embury and R. Nicholson, Acta Metallurgica, 13, 403-417 (1965).
H. J. Engle, The Theory of Stress Corrosion Cracking in Alloys, North Atlantic Treaty
Organization (1971).
D. G. Evans and P. W. Jeffrey (1974), "Exfoliation Corrosion of AlZnMg Alloys", in U.R.
Evans Conference on Localized Corrosion, NACE-3, Houston, TX, Alcan
International Limited Research Center.
R. Fitzgerald, Physics Today, 53, 23-26 (2000).
H. M. Flower and P. J. Gregson, Materials Science and Technology, 3, 81-90 (1987).
G. S. Frankel, Journal of Electrochemical Society, 145(6), 2186-2198 (1998).
J. R. Galvele and S. M. D. D. Micheli, Corrosion Science, 10, 795-807 (1970).
A. Garner and D. Tromans, Corrosion, 35(2), 55-60 (1979).
J. M. Gere and S. P. Timoshenko, Mechanics of Materials, Boston, MA: PWS Publishing
Company (1997).
R. J. Gest and A. R. Troiano, Corrosion, 30, 274-279 (1974).
E. Ghali, "Aluminum and Aluminum Alloys", in Uhlig's Corrosion Handbook, ed., R. W.
Revie, New York, NY: John Wiley & Sons, INC., 677-716 (2000).

184

S. Glasstone, K. J. Laidler and H. Eyring, "Electrochemical Processes", in the Theory of

Rate Processes, Frick Chemical Laboratory, Princeton University, McGRAWHILL BOOK COMPANY, Inc., 552-599 (1941).
H. P. Godard and W. E. Cooke, Corrosion, 16, 117-123 (1960).
A. I. Golubev and M. N. Ronzhin, Corrosion of Metals and Alloys. London: Oldbourne
Press (1966).
R. Grauer and E. Wiedmer, Werkstoffe Korros, 31, 550 (1980).
W. Gruhl, Zeitschrift fuer Metallkunde, 75(11), 819-826 (1984).
V. Guillaumin and G. Mankowski, Corrosion Science, 41, 421-438 (1999).
V. Guillaumin and G. Mankowski, Corrosion Science, 42, 105-125 (2000).
D. Hardie, N. J. H. Holroyd and R. N. Parkins, Materials Science, 12(11), 603-610 (1979).
D. A. Hardwick, A. W. Thompson and I. M. Bernstein, Corrosion Science, 28(12), 11271137 (1988).
E. W. Hart, Surface and Interfaces II, Syracuse University Press, p.120 (1968).
G. S. Haynes and R. Baboian, "Modified Salt Spray (Fog) Testing", in Laboratory
Corrosion Tests and Standards, ASTM Standard G85-85 A2. STP866,
Philadelphia, PA: ASTM, 578-583 (1985).
M. B. Hinz, Scripta Metallurgica, 19(12), 1445-1450 (1984).
H. A. Holl, Journal of the Institute of Metals, 97(200), 200-205 (1969).
E. H. Hollingsworth and H. Y. Hunsicker, "Corrosion of Aluminum and Aluminum
Alloys", in Metals Handbook, Materials Park, OH: ASM International, 583-609
(1990).
N. J. H. Holroyd and G. M. Scamans, Scripta Metallurgica, 19(7), 915-916 (1985).
J. M. Holt, H. Mindlin and C. Y. Ho, Structural Alloys Handbook, West Lafayette,IN:
CINDAS/Purdue University (1994).
T. Huang and G. S. Frankel, "Effect of Temper and Potential on Localized Corrosion
Kinetics of AA7075". Submitted to Corrosion, 11/05,
T. Huang and G. S. Frankel, "Sharp Intergranular Corrosion Fissures in AA7178".
Submitted to Corrosion Science, 11/05,

185

T. S. Huang and G. S. Frankel (2002). "Localized Corrosion Growth Kinetics in Al

Alloys". 6th Joint FAA/DoD/NASA Aging Aircraft Conference-Sept. 16-19,2002.
F. Hunkeler and H. Bohni, Corrosion, 37(11), 645-540 (1981).
F. Hunkeler and H. Bohni, Corrosion, 40(10), 534-540 (1984).
M. S. Hunter, G. R. Frank and D. L. Robinson, in Second International Congress on
Metallic Corrosion, 1963, National Association of Corrosion Engineer.
M. Izu, M. Shiokawa and E. Sato, Journal of the Japan Institute of Metals, 34(11), 643649 (1984).
D. A. Jones, Principles and Prevention of Corrosion, Upper Saddle River, NJ: PrenticeHall, Inc. (1996).
R. H. Jones, Stress Corrosion Cracking, Materials Park, Ohio: ASM International (1992).
A. Joshi, C. R. Shastry and M. Levy, Metallurgical Transaction A: Physical Metallurgy
and Materials Science, 12A(6), 1081-1088 (1981).
A. J. De Ardo Jr. and R. D. Townsend, Metallurgical Transactions, 1, 2573-2581 (1970).
E. H. Dix Jr., Transactions, American Institute of Mining and Metallurgical Engineers,
137, 11-40 (1940).
E. H. Dix Jr., Physical Metallurgy of Aluminum Alloys, Cleveland, OH: American
Society for Metals, 214 (1949).
E. H. Dix Jr., Transactions of the American Society for Metals, 42, 1057-1127 (1950).
M. Keddam, C. Kuntz, H. Takenouti, D. Schuster and D. Zuili, Electrochimica Acta,
42(1), 87-97 (1997).
D. J. Kelly, "Exfoliation and Stress Corrosion Cracking of the Aluminum Lithium Alloy
8090", Cranfield University, (1991).
D. J. Kelly and M. J. Robinson, Corrosion, 49(10), 787-795 (1993).
K. G. Kent, Journal of the Institute of Metals, 97, 127-128 (1969).
S. J. Ketcham, Corrosion Science, 7(6), 305 (1967).
S. J. Ketcham and F. H. Haynie, Corrosion, 19(7), 242t (1963).
T. F. Klimowicz and R. M. Latanision, Metallurgical Transaction A, 9A, 597-599 (1978).
C. Kumai, J. Kusiniski, G. Thomas and T. M. Devine, Corrosion, 45, 294 (1989).
186

A. H. Le and R. T. Foley, Corrosion, 40(4), 195-197 (1984).

N. LeBozec, D. Persson, A. Nazarov and D. Thierry, Journal of Electrochemical Society,
149(9), B403-B408 (2002).
S. Lee and B. W. Lifka, "Modification of the EXCO Test Method for Exfoliation
Corrosion Susceptibility in 7XXX, 2XXX and Aluminum-Lithium Alloys". New
Methods for Corrosion Testing of Aluminum Alloys, ASTM STP 1134.V. eds.,
S.Agarwala and G.M.Ugiansky. Philadelplia, PA: American Society for Testing
and Materials, 1-19 (1992).
E. A. G. Liddiard, J. A. Whittaker and H. K. Farmery, Journal of the Institute of Metals,
89, 377-384 (1960-1961).
B. W. Lifka, "Corrosion of aluminum alloys", in Corrosion Engineering Handbook, ed.,
P.A.Schweitzer, Marcel Dekker, New York: 129-133 (1996).
B. W. Lifka and D. O. Sprowls, Corrosion, 22, 7-15 (1966).
B. W. Lifka and D. O. Sprowls, "Significance of Intergranular Corrosion in High
Strength Aluminum Alloy Products", in Localized Corrosion: Cause of Metal
Failure, ASTM STP 516, p. 120 (1972).
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", ASTM SPEICIAL TECHNICAL PUBLUCATION 558,
306-333 (1973).
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, Columbus, Ohio (1967)
H. L. Logan, Journal of Research of the National Bureau of Standards, 48, 99 (1952).
L. I. Maissel and R. Glang, Handbook of Thin Film Technology, New York: McGrawHill
(1970).
S. Maitra and G. C. English, Metallurgical Transactions A, 12A, 535-541 (1981).
S. Maitra and G. C. English, Metallurgical Transactions A, 13A, 161-166 (1982).
E. Mattsson, L. O. Gullman, L. Knutsson, R. Sundberg and B. Thundal, British Corrosion
Journal, 6, 73-83 (1971).
B. Mazurkiewicz and A. Piotrowski, Corrosion Science, 23(7), 697-707 (1983).
187

G. Mazzanti, G. C. Montanari and A. Motori, Journal of Applied Physics D: Applied

Physics, 27, 2601-2611 (1994).
D. McNaughtan, M. Worsfold and M. J. Robinson, Corrosion Science, 45, 2377-2389
(2003).
E. I. Meletis, Materials Science and Engineering, 93(1-2), 235-245 (1987).
Q. Meng and G. S. Frankel, Surface and Interface Analysis, 36, 30-42
Q. Meng and G. S. Frankel, Journal of The Electrochemical Society, 151(5), B271-B283
(2004).
G. C. Montanari and A. Motori, Journal of Applied Physics D: Applied Physics, 24,
1172-1181 (1991).
G. C. Montanari and A. Motori, Journal of Physics D: Applied Physics, 28, 1433-1438
(1995).
I. L. Muller and J. R. Galvele, Corrosion Science, 17, 179-193 (1977).
I. L. Muller and J. R. Galvele, Corrosion Science, 17, 995-1007 (1977)
K. Nisancioglu, Journal of The Electrochemical Society, 137(1), 69-77 (1990).
J. Onoro and C. Ranninger, British Corrosion Journal, 28(2), 137-141 (1993).
J. K. Park and A. J. Ardell, Acta Metallurgica et Materialia, 39(4), 591-598 (1991).
J. R. Pickens and T. J. Langan, Metallurgical Transactions A: Physical Metallurgical and
Materials Science, 18A(10), 135-149 (1987).
H. W. Pickering, F. H. Beck and M. G. Fontana, Corrosion, 18, 230-239 (1962).
M. Posada, L. E. Murr, C.-S. Niou, D. Roberson, D. Little, R. Arrowood and D. George,
Materials Characterization, 38, 259-272 (1997).
P. K. Poulose, J. E. Morral and A. J. Mevily, Metallurgical Transactions, 5(6), 1393-1400
(1974).
M. Puiggali, A. Zielinski, J. M. Olive, E. Renauld and D. Desjardins, Corrosion Science,
40(4/5), 805-819 (1998).
P. Qiu, The Indian Journal of Statistics, 62, 223-235 (2000).
K. Rajan, W. Wallace and J. C. Beddoes, Journal of Materials Science, 17, 2817-2824
(1982).
188

T. Ramgopal, Intergranular Corrosion of AA7XXX Alloys, MSE, Columbus, The Ohio

State University, (2001).
T. Ramgopal, P. I. Gouma and G. S. Frankel, Corrosion, 58(8), 687-697 (2002).
T. Ramgopal, P. Schmutz and G. S. Frankel, Journal of The Electrochemical Society,
148(9), B348-B356 (2001).
T. Ramgopal and G.S. Frankel, Corrosion, 57(8), 702-711 (2001).
M. C. Reboul and J. Bouvaist, Werkstoffe Korros, 30, 700 (1979).
B. Reich and E. B. Hakim, Solid State Technology, September, 65-66 (1972).
J. G. Rinker, M. Marek and T. H. Sanders, Materials Science and Engineering, 64(2),
203-221 (1984).
M. J. Robinson, Corrosion Science, 22(8), 775-790 (1982).
M. J. Robinson, Corrosion Science, 23(8), 887-899 (1983).
M. J. Robinson and N. C. Jackson, British Corrosion Journal, 34(1), 45-49 (1999).
M. J. Robinson and N. C. Jackson, Corrosion Science, 41, 1013-1028 (1999).
H. B. Romans, Materials Research & Standards, 31-34 (1969).
W. Rosenkranz, Metall, 15(5), 894-899 (1961).
N. Ryum, B. Haegland and T. Lindtveit, Zeitschrift fur Metallkunde, 58, 28-31 (1967).
F. Sato and R.C. Newman, Corrosion, 54(12), 955-963 (1998).
F. Sato and R.C. Newman, Corrosion, 55(1), 3-9 (1999).
G. M. Scamans, Metallurgical Transactions A: Physical Metallurgy and Materials
Science, 11A(5), 846-850 (1980).
G. M. Scamans, N. J. H. Holroyd and C. D. S. Tuck, Corrosion Science, 27(4), 329-347
(1987).
P. Schmutz and G. S. Frankel, Journal of The Electrochemical Society, 145(7), 22852295 (1998).
J. C. Scully, Corrosion Science, 15, 207 (1975).
J. R. Scully, T. O. Knight, R. G. Buchheit and D. E. Peebles, Corrosion Science, 33, 185195 (1993).
189

A. J. Sedriks, J. A. S. Green and D. L. Novak, "Corrosion Processes and Solution

Chemistry Within Stress Corrosion Cracks in Aluminum Alloys", 1971, in
National Association of Corrosion Engineers, Houston, TX.
A. Sehgal, G. S. Frankel, B. Zoofan and S. I. Rokhlin, Journal of The Electrochemical
Society, 147(1), 140-148 (2000).
T. Sheppard and N. C. Parson, Materials Science and Technology, 3, 345-352 (1987).
L. L. Shreir, F. R. I. C and F. I. M, Corrosion, Newnes-Butterworth (1976).
A. Singirev and I. Snigireva, Review of Scientific Instruments, 66(12), 5486-5492 (1995).
M. O. Speidel, "Interaction of Dislocations with Precipitates in High Strength Aluminum
Alloys and Susceptibility to Stress Cracking", in Fundamental Aspects of Stress
Corrosion Cracking, ed., R.W.Staehle, 561-579 (1967).
M. O. Speidel and M. V. Hyatt, "Stress-Corrosion Cracking of High-Strength Aluminum
Alloys", in Advances in Corrosion Science and Technology, eds., M. G. Fontana
and R. W. Staehle.New York-London: Plenum Press,2, p.118 (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",
Philadelphia, PA: American Society for Testing and Materials, 99-112 (1973).
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).
R. W. Staehle, the Theory of Stress Corrosion Cracking in Alloys, North Atlantic Treaty
Organization, p.223 (1971).
E. I. Storchai and A. V. Turkovskaya, Protection of Metals, 6, 626 (1970).
K. Sugimoto, K. Hoshino, M. Kageyama and Y. Sawada, Corrosion Science, 15, 709-720
(1975).
K. Sugimoto and Y. Sawada, Journal of the Japan Institute of Metals, 37, 148 (1973).
T. J. Summerson and D. O. Sprowls (1986), "Corrosion Behavior of Aluminum Alloys",
in International Conference in Celebration of the Centennial of the Hall-Heroult
Process, University of Virginia, Charlottesville, Virginia.
M. Talianker and B. Cina, Metallurgical Transactions A: Physical Metallurgical and
Materials Science, 20A(10), 2087-2092 (1989).
E. Tankins, J. Kozal and E. W. Lee, Journal of Metals, 9, 40-43 (1995).
190

A. W. Thompson, M. A. Mueller and I. M. Bernstein, Metallurgical Transactions A, 24A,

2569-2575 (1993).
P. Tobias, Assesing Product Reliability.NIST/SEMATECH e-Handbook of Statistical
Methods, http://www.itl.nist.gov/div898/handbook/, Statistical Engineering
Division.
Z.-X. Tong, S. Lin and C.-M. Hsiao, Metallurgical Transactions A, 20A, 921-924 (1989).
Z.-X. Tong, S. Lin and C.-M. Hsiao, Metallurgical Transactions A, 20A, 925-933 (1989).
K. Urushino and K. Sugimoto, Corrosion Science, 19, 225-236 (1979).
D. A. Vermilyea, Journal of The Electrochemical Society, 119, 405-407 (1972).
R.C. Weast, M.J. Astle and W.H. Beyer, CRC Handbook of Chemistry and Physics, Boca
Raton, FL: CRC Press, Inc. (1987).
S. W. Wilkins, T. E. Gureyev, D. Gao, A. Pogany and A. W. Stevenson, Nature, 384(28),
335-338 (1996).
C. P. You, M. Dollar, A. W. Thompson and I. M. Bernstein, Metallurgical Transaction A,
22A, 2445-2450 (1991).
J. Zahavi and J. Yahalom, Journal of The Electrochemical Society, 129(6), 1181-1186
(1982).
M. Zamin, Corrosion, 37(11), 627-632 (1981).
W. Zhang and G. S. Frankel, Electrochemical and Solid-State Letters, 3(6), 268-270
(2000).
X. Zhao, Chapter 3 of this dissertation.
X. Zhao, Chapter 4 of this dissertation.
B. Zoofan and S. I. Rokhlin, Materials Evaluation, 52(2), 191-194 (1998).

191