Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

ThesisBeatriz Almeida Ferreira
Aerospace Engineering
Supervisors: Prof. Luís Filipe Garlão Reis Eng. Bruno Aires de Albuquerque e Concha de Almeida
Examination Committee
Chairperson: Prof. Filipe Szolnoky Ramos Pinto Cunha Supervisor: Prof. Luís Filipe Garlão Reis
Member of the Committee: Prof. Maria de Fátima Reis Vaz
November 2017
iii
iv
Acknowledgments
There are many people that helped me through this Master Thesis. To them, I am sincerely grateful.
First of all, I would like to thank CEiiA for giving me this opportunity and financing my thesis. It was
an amazing experience that made me grow up as a person and as a professional.
Moreover, I would like to express my gratitude to all CEiiA’s Team and CEiiA’s interns that were doing
their master thesis at the same time as me.
At CEiiA’s Aerospace & Ocean team I would like to highlight Bruno Albuquerque, that was always
available and pacient with me, Tiago Rebelo and Paulo Figueiredo for accompanying me throughout the
whole process.
At the Production & Test department a special thanks to Carla Silva, Sara Sampaio, Joao Torrejon
and Manuel Oliveira that helped me with the experimental part of my work and, without them, none of
this would be possible.
For the numerical part of this document, the support of Ines Martins, also from CEiiA, was essential.
I can not forget my supervisor at IST, Prof. Lus Reis for all the good advises that he gave me in
addition to all the time he invested in me, including also an appreciation to the University of Lisbon.
Finally, a big thanks to my family for letting me get this far and supporting me no matter what. To
them I am forever grateful.
v
vi
Resumo
As ligas de alumnio sao muito utilizadas na industria aeronautica especialmente na estrutura principal
dos avioes. Estas ligas estao sujeitas a corrosao, e consequente degradacao do material, que se nao
for tratada pode tornar uma aeronave inapta para voar.
Consequentemente, o comportamento corrosivo da liga de alumnio 7075-T651, com e sem anodizacao,
numa camara de nevoeiro salino em ciclo molhado/seco foi estudado em termos de perda de propriedades
mecanicas e vida a fadiga. Entre o perodo seco e o molhado tres metodos de lavagem foram considerados:
acido ntrico, agua doce e nao lavar.
Os efeitos da corrosao nas propriedades mecanicas revelaram um decrescimo geral, com o aumento
da corrosao, para a tensao de rotura e de cedencia. O modulo de Young nao teve variacao significativa
e as medidas de ductilidade decresceram exponencialmente, no geral.
A vida a fadiga foi reduzida consideravelmente pela anodizacao decrescendo exponencialmente
com a corrosao em resposta a iniciacao de fendas prematuras provenientes das covas formadas que
atuaram como pontos de concentracao de tensao.
Da analise da superfcie detetou-se que o principal mecanismo de corrosao foi o pitting. As covas
comecaram a aparecer em lugares aleatorios na superfcie exposta e o seu desenvolvimento em
numero e dimensao, levou a que muitas se agrupassem, preferencialmente na direcao do grao, com
covas vizinhas criando, gradualmente, covas maiores e mais profundas.
Foi desenvolvida uma ferramenta usando os resultados experimentais para caracterizar o material
corrodo e gerar as suas propriedades na linguagem do programa Nastran. Estas propriedades podem
ser usadas para qualquer simulacao em qualquer peca feita de alumnio 7075-T651.
Palavras-chave: Liga de Alumnio 7075-T651, Teste em Camara de Nevoeiro Salino, Corrosao,
Pitting, Propriedades de Tracao, Vida a fadiga
vii
viii
Abstract
Aluminum alloys are widely used in the aeronautical industry especially in the main airframe. These
alloys are susceptible to corrosion, accompanied by embrittlement, that, if left untreated, can make an
aircraft unairworthy.
With this in mind, the corrosion behavior of bare and anodized 7075-T651 aluminum alloy under
a wet/dry cycle salt spray test in terms of loss of mechanical properties and fatigue life was studied.
Between the wet and the dry period, three different washing methods were considered: nitric acid,
freshwater, and no wash.
The effects of corrosion on the tensile properties revealed a general decay with increasing exposure
time for ultimate tensile strength and yield stress. The Young’s modulus did not have a significant
variation and the ductility measures decreased exponentially generally.
The fatigue life was reduced drastically by the anodization and tended to decrease exponentially with
corrosion as a result of premature crack initiation caused by pitting where pits are considered stress
concentrations.
From the surface analysis, the primary corrosion mechanism detected was pitting. The pits started
to appear in random places all over the exposed surface and they coalesced, preferably in the grain
direction, with neighboring pits creating bigger and deeper pits.
A practical application was developed using the experimental results to characterize the corroded
material and generate the necessary material properties in the FEM solver standard language. These
properties can be used for any simulation in any structure made out of the 7075-T651 aluminum alloy,
within the test domain.
Properties, Fatigue Life
2.2 Historical Background of Aluminum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.2.1 Aluminum Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8
2.3 Aluminum Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9
2.4 Types of Aluminum Corrosion Mechanisms . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.4.1 General Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.4.2 Galvanic Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 11
2.4.3 Localized Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 12
2.5 Corrosion Mitigation Strategies in Aluminum and Aluminum Alloys . . . . . . . . . . . . . 14
2.5.1 Anodizing Aluminum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.7 Typical Tests for Corrosion Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15
2.7.1 Testing Modes . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 16
2.8 Tensile Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
2.9 Fatigue Tests . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 17
3 Experimental Procedure 19
3.1 Specimen Description . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19
3.1.3 Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.1.4 Quantity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21
3.1.5 Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22
3.2.3 Dry Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26
4.1 Corrosion Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1.1 Surface Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.1.5 Valley Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.2.2 Yield Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49
xii
5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59
5.2 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.3 Meshing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60
5.5 Analysis Type and Material Properties . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.5.1 Linear Static Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 63
5.5.2 Nonlinear Static Approach . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 64
5.6 Practical Application of Experimental Results . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.6.1 Program’s Inputs and Outputs . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 69
5.6.2 Program Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
5.6.3 Code Development . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
A.1 Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81
B Equipment, Software and Support Material List 93
B.1 Corrosion Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93
B.2 Tensile Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
B.3 Fatigue Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94
xiii
xiv
3.1 Chemical Composition of 7075-T651 Aluminum Alloy (wt.%) . . . . . . . . . . . . . . . . . 19
3.2 Test Specimen Dimensions (mm) . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.3 Quantity and Variations of the Corroded Samples for Each Exposure Time . . . . . . . . . 21
3.4 Quantity and Variations of the Undamaged Samples . . . . . . . . . . . . . . . . . . . . . 22
3.5 Quantity and Variations of the Samples used to built S-N curves . . . . . . . . . . . . . . 22
3.6 Technical Features of the Dry Corrosion Test Cabinet . . . . . . . . . . . . . . . . . . . . 24
3.7 Technical Features of the Aralab Climatic Chamber Fitoclima 500 EP20 (Source: [40]) . . 26
4.1 Average Tensile Property Values of the Uncorroded Specimens . . . . . . . . . . . . . . 48
4.2 Average Fatigue Life of the Uncorroded Specimens . . . . . . . . . . . . . . . . . . . . . 54
5.1 Stress and Strain Value Comparison for Linear Static Analysis . . . . . . . . . . . . . . . 63
5.2 Stress and Strain Value Comparison of a Linear Point in Nonlinear Static Analysis . . . . 66
5.3 Stress and Strain Value Comparison of a Nonlinear Point in Nonlinear Static Analysis . . 67
A.1 Anodized Specimen Label’s Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . 81
A.2 Not Anodized Specimen Label’s Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . 81
A.3 S-N Curve Specimen Label’s Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . 81
A.4 Specimen’s Label . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82
A.5 S-N Curve Specimen’s Label . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
A.6 Tensile Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
A.7 Ultimate Tensile Strength Averages and Standard Deviations . . . . . . . . . . . . . . . . 87
A.8 Yield Stress Averages and Standard Deviations . . . . . . . . . . . . . . . . . . . . . . . . 87
A.9 Young’s Modulus Averages and Standard Deviations . . . . . . . . . . . . . . . . . . . . . 88
A.10 Elongation at Fracture Averages and Standard Deviations . . . . . . . . . . . . . . . . . . 88
A.11 Reduction Area Averages and Standard Deviations . . . . . . . . . . . . . . . . . . . . . . 89
A.12 Fatigue Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90
A.13 Fatigue Life Averages and Standard Deviations . . . . . . . . . . . . . . . . . . . . . . . . 91
A.14 Fatigue Test Results for S-N Curve Specimens . . . . . . . . . . . . . . . . . . . . . . . . 91
xv
xvi
2.1 Summary of the Background Chapter . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7
2.2 Pourbaix diagram (potential vs pH) for aluminum showing the conditions of corrosion,
immunity and passivation of aluminum at 25oC. . . . . . . . . . . . . . . . . . . . . . . . . 11
2.3 Mechanism of Pitting Corrosion of Aluminum . . . . . . . . . . . . . . . . . . . . . . . . . 12
3.1 Rectangular Tension Test Specimen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20
3.2 Half Cuted Aluminum Sheet on top of the Vaccuum Table . . . . . . . . . . . . . . . . . . 21
3.3 Masked and Labeled Specimens (Not Anodized - Top; Anodized - Bottom) . . . . . . . . . 23
3.4 ACS Dry Corrosion Test Cabinet 1200 . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.5 Empty Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 24
3.6 Filled Support . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 25
3.7 Scheme of the Test Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28
3.8 MTS Landmark Servohydraulic Testing Machine with a Load Cell of 100kN, Model 370.25 29
3.9 MTS Landmark Servohydraulic Testing Machine’s Grips and Load Cell . . . . . . . . . . . 30
3.10 Specimen during a tensile test with a biaxial extensometer . . . . . . . . . . . . . . . . . 30
3.11 Engineering Stress-Strain relationship under uniaxial tensile loading . . . . . . . . . . . . 31
3.12 Stress-Strain Diagram for Determination of Yield Strength by the Offset Method . . . . . . 31
3.13 MTS Landmark Servohydraulic Testing Machine with a Load Cell of 50kN, Model 370.10 33
4.1 Uncorroded and Anodized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.2 Uncorroded and Not Anodized . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35
4.3 A - 0 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.4 AB - 1 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.5 AB - 3 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.6 AB - 9 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.7 AB - 15 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.8 AB - 20 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
4.9 A - 0 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.10 AC - 1 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
4.11 AC - 3 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
xvii
4.39 Effect of corrosion on Maximum Valley Depth . . . . . . . . . . . . . . . . . . . . . . . . . 45
4.40 Effect of corrosion on Average Valley Depth . . . . . . . . . . . . . . . . . . . . . . . . . . 46
4.41 Effect of corrosion on Valley Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47
4.42 Engineering Stress-Strain Curve Evolution for Anodized Washed with Nitric Acid (AB)
Specimens . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48
4.43 Effects of corrosion on Ultimate Tensile Strength of 7075-T651 Aluminum Alloy . . . . . . 49
4.44 Effects of corrosion on Yield Stress of 7075-T651 Aluminum Alloy . . . . . . . . . . . . . . 50
4.45 Effects of corrosion on Young’s Modulus of of 7075-T651 Aluminum Alloy . . . . . . . . . 51
4.46 Effects of corrosion on Elongation at Fracture of 7075-T651 Aluminum Alloy . . . . . . . . 52
4.47 Effects of corrosion on Reduction Area of 7075-T651 Aluminum Alloy . . . . . . . . . . . 53
4.48 Effects of corrosion on Fatigue Life of 7075-T651 Aluminum Alloy . . . . . . . . . . . . . . 55
4.49 Effects of corrosion on Fatigue Life of 7075-T651 Aluminum Alloy . . . . . . . . . . . . . . 56
xviii
4.50 Specimen’s Cross Section After Fatigue Test . . . . . . . . . . . . . . . . . . . . . . . . . 56
4.51 Effect of Corrosion on the S-N Curve of 7075-T651 Aluminum Alloy . . . . . . . . . . . . . 57
5.1 CQUAD4 Element Geometry and Coordinate Systems . . . . . . . . . . . . . . . . . . . . 61
5.2 Mesh Convergence Study . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 61
5.3 Meshed Specimen . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62
5.4 Load and Boundary Conditions Applied on the Specimen . . . . . . . . . . . . . . . . . . 62
5.5 Solution Scheme of a Nonlinear Problem in Nastran . . . . . . . . . . . . . . . . . . . . . 65
5.6 Stress-Strain Curve Definition . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
5.7 Effects of Corrosion on the Slope of the Plastic Line . . . . . . . . . . . . . . . . . . . . . 68
5.8 Effects of Corrosion on the y-intercept of the Plastic Line . . . . . . . . . . . . . . . . . . . 68
5.9 MATLAB R© Input Window - Example for 15 Corrosion Cycles and AB Case Specimen . . . 69
5.10 MATLAB R© Graph Window - Example for 15 Corrosion Cycles and AB Case Specimen . . 70
5.11 Program Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 71
C Elasticity Matrix.
P External Loads.
xxii
Glossary
acid A chemical substance that yields hydrogen ions (H+) when dissolved in water, represented by low
values of pH. 12
anaerobic Free of air or uncombined oxygen. 13
atm Atmosphere (unit); Unit of pressure defined as 101325 Pa, sometimes used as a reference or
standard pressure. 10
basic A chemical substance that yields hydroxyl ions (OH−) when dissolved in water, represented by
high value of pH. 12
electrode An electronic conductor in contact with an ionic conductor. 10, 14
electrolyte A chemical compound or mixture of compounds that when molten or in solution will conduct
an electric current. 9, 11, 12, 14
hydrogen embrittlement A process resulting in a decrease in the toughness or ductility of a metal due
to the presence of atomic hydrogen. 2, 56
hydrolysis Decomposition or alteration of a chemical substance by water. 12, 13
M Molar (unit); Unit for molar concentration which is defined as the molar by liter. 10
quenching Rapid cooling of metals from a suitable elevated temperature. This generally is accomplished
by immersion in water, oil, polymer solution, or salt, although forced air is sometimes used.. 19
rolling direction The parallel direction with the structure lines of the surface. 13
specific gravity The ratio of the density of a substance to the density of a water at 4oC. 25
xxiii
xxiv
Acronyms
AUV Autonomous Underwater Vehicle. 3, 5, 74
CASS Copper-accelerated Acetic Acid Salt Spray. 16
CPU Central Processing Unit. 65
DCTC Dry Corrosion Test Cabinet. 23, 24, 89
FEM Finite Element Method. 59, 60, 62, 64, 70, 72, 75
HCF High Cycle Fatigue. 18, 22, 57
LCF Low Cycle Fatigue. 18
LVDT Linear Variable Differential Transformer. 29, 33, 90
NSS Neutral Salt Spray. 16
RH Relative Humidity. 26
UTS Ultimate Tensile Strength. 48, 68, 70, 72
xxv
xxvi
1.1 Motivation
In March 2016 the National Association of Corrosion Engineers (NACE) released the ”International
Measures of Prevention, Application and Economics of Corrosion Technology (IMPACT)” study, in which
it estimates the global cost of corrosion to be US$2.5 trillion annually [1].
Every transportation, infrastructure, and utilities, which supply oil, gas, water, electricity, among
others, are susceptible to corrosion. All of this important structures to human life and comfort can
fail at any moment because of this phenomenon, sometimes this failure is catastrophic.
The importance of corrosion management, monitoring, and prevention has become more and more
important and the question is how much money it will cost to reduce the cost of corrosion? [1].
It is necessary to study and evaluate new ways of managing corrosion and prevent it. It is also
crucial to study the corrosion behavior of the materials so their service life, in different conditions, can
be predicted. The knowledge of the component’s material performance under corrosion in any structure
can prevent a major failure.
From the earliest constructions to the development of new technology, corrosion has been an endless
matter, because corrosion costs money, corrosion costs jobs, but most of all corrosion costs lives.
1.2 Topic Overview
In 2004, Ghiocel and Tuegel [2] broadly characterized corrosion in aluminum alloys into general corrosion,
pitting, exfoliation and intergranular, being pitting the most common one.
In 1996, Chen et al., [3], identified two types of constituent particles in the investigation of the
role of microconstituents in pitting corrosion in Aluminum Alloy 2024-T3. Each type played a different
role in inducing pitting corrosion. Particles containing Al, Cu, Fe, and Mn acted as cathodes and
promoted matrix dissolution at their periphery. Particles containing Al, Cu, and Mg showed anodic
behavior and dissolved with a preferential dissolution of Mg and Al. It was also evident that individual
particle-nucleated pits coalesced, laterally and in depth, to form larger pits.
1
Frankel, in 1998 [4], studied the factors that affect pitting corrosion and explained its phenomena. The
author presents a concise outline of the phenomenology and stages of pitting showing that constituent
particles play a significant role in the corrosion process of aluminum alloys.
Later, in 1998, Harlow and Wei [5] presented a probabilistic model for the growth of corrosion pits
in aluminum alloys in aqueous environments. The authors’ objective was to estimate the cumulative
distribution function for the size of corrosion pits at a given time for use in multi-site damage and crack
growth analyses.
The relation between pitting and fatigue did not take long to realize, and in 1996 Chen, Wan et al., [6],
observed that fatigue cracks typically nucleated from one or two of the larger pits, and the size of the pit
at which the fatigue crack nucleates is a function of the stress level and load frequency. The observations
indicated that the nucleation of corrosion fatigue cracks essentially results from a competition between
the processes of pitting and crack growth. Pitting predominates in the early stage of the corrosion fatigue
process and is replaced by corrosion fatigue crack growth.
In 1998 Du et al., [7], stated the extent of synergistic activity between corrosion and fatigue effects.
In their study, the authors used the surface roughness as a revealing parameter in corrosion-fatigue
interaction.
In 2000, Brian Obert, [8], tested on his master thesis the effects of corrosion on the static strength
and fatigue life of the 7075-T6 aluminum alloy. After corrosion, specimens were tested in tension and
fatigue. The effect of corrosion on the tensile strength resulted in a large initial drop in strength, then a
linear reduction in strength as mass loss increased. The tensile strength decreased significantly at low
mass loss levels. The reduction of fatigue life due to corrosion tended to follow an exponential reduction
as mass loss increased. Even small amounts of corrosion in specimens reduced the fatigue life of the
aluminum alloy drastically.
Later, Sankaran et al., 2001 [9], studied the effects of pitting corrosion on the fatigue behavior of
bare 7075-T6 aluminum alloy. Pitting corrosion decreased the fatigue lives by a factor between 6 and
8. The measured fatigue lives generally agreed with the predictions using the average rather than the
maximum pit size as the initial crack size. This result could be explained by the pit size distributions
offering a significantly larger population of pits near the average size.
Pits were identified as crack origins in all corroded specimens also by Jones and Hoeppner, 2004
[10]. They realized that it was not always the largest pit that formed/nucleated fatigue cracks. The
combined effects of pit depth, pit surface area, and proximity to other pits were found to reduce fatigue
life considerably.
To analyze environmental factors that affect fatigue life Chlistovsky et al., 2007 [11] subjected specimens
to fatigue testing while they were fully immersed in an aerated and recirculated 3.5 wt% NaCl simulated
seawater solution. A damage analysis showed that the presence of the corrosive environment accelerated
the damage accumulation rate to a greater extent than that observed in air, particularly at low stress
ranges. This resulted in a reduction in the fatigue strength of the material. The reduced fatigue life
was due primarily to corrosion pit formation and a combination of anodic dissolution at the crack tip and
hydrogen embrittlement. Cavanaugh et al., 2010 [12], included in his study variables like temperature,
2
pH, [Cl– ], exposure time, and orientation. Effects of these variables in affecting maximum pit depth and
maximum pit diameter for high-strength aluminum alloy 7075-T651 were studied and reported and it was
found that exposure time and pH, followed by temperature, were the most significant variables impacting
pit depth.
In 2011 Liu et al.,[13], showed that the corrosion rate of 2A12 aluminum alloy was very high in the
first 24 h, and then decreased with test duration.
Arriscorreta and Hoeppner, in 2012 [14], investigated the effects of prior corrosion and stress in
corrosion fatigue of aluminum alloy 7075-T6 and concluded that the effect of stress appears to be more
detrimental than corrosion time on the fatigue life.
The effect of anodic oxidation on fatigue performance of 7075-T6 alloy for pre-corroded and non-corroded
specimens has been investigated by Cirik and Genel, 2008 [15], and the results indicate that the
anodization has a tendency to decrease the fatigue performance. Fatigue strength was reduced with
increasing coating thickness; approximately 40% reduction for a 23 µm thick coating was obtained.
However, it was observed that oxidation mitigated pitting corrosion.
In 2013, Hemmouche et al. [16], studied the effects of some heat treatments and anodizing processes
on fatigue life of aluminum alloy 2017. The result of fatigue tests showed a decrease in fatigue life of
anodized specimens as compared to untreated ones. The decrease in the fatigue life could mainly be
attributed to the brittle nature of oxide layer and to the heterogeneous microstructure of the film.
Later, in 2015 [17], Dejun and Jinchun did a salt spray corrosion test on 7475 aluminum alloy bare
and anodized. The results showed that the corrosion in the original sample surface is severe after salt
spray corrosion, while the anodic oxide film was only slightly corroded, owing it to Al2O3 which prevents
Cl− to contact the base metal effectively.
1.3 Framework
This MSc Thesis was made in partnership with CEiiA. CEiiA is a ”Portuguese center of engineering
and product development that designs, implements and operates innovative products and systems,
alongside with its partners, for high-tech industries, such as, the aeronautics, automotive, smart mobility,
oceans and space”.
CEiiA’s vision is to “Establish Portugal as a reference within the mobility industries, particularly in
the development of technologies, products and systems, conceived, industrialized and operated from
Portugal.”.
Inside CEiiA, the area in which the research was carried out was Ocean & Space Engineering.
The project in which this thesis is inserted is called MEDUSA DEEP SEA, the Autonomous Underwater
Vehicle (AUV) presented in Figure 1.1.
3
Figure 1.1: Medusa Deep Sea [Courtesy of CEiiA]
The goal of this project is to reinforce the national capacity for mobile autonomous and extended
range deep-sea exploration and monitoring, affording scientists and commercial operators means to
open and explore the deep sea frontier and contribute for the Good Environmental Status in oceanic
and coastal areas. This is achieved by developing a system of multiple autonomous vehicles for ocean
exploration and monitoring, capable of operating at water depths of up to 3000 meters.
The main applications are water column profiling, seabed mapping, oil and gas survey, geographical
survey, oceanographic survey and search.
The main characteristics are presented in table 1.1.
Table 1.1: Medusa Deep Sea Main Characteristics [Courtesy of CEiiA]
Type Double hull AUV Size 2.8m x 1.65m x 0.75m
Weight 350kg Endurance 7h
Nominal Speed 1.25 m/s
The material of the main structure of the autonomous underwater vehicle is the anodized 7075-T651
aluminum alloy. This is a very used material for aeronautical applications and CEiiA is trying to use it for
ocean engineering as well.
Aluminum alloys have been the main airframe materials since they started replacing wood in the late
1920’s [18]. The 7075 aluminum alloy is normally applied in fuselage stringers and frames, upper wing
stringers, floor beams, and seat rails and is subjected to corrosion. The corrosive action, which causes
embrittlement, begins on the metal surface and moisture in the air is often sufficient to start corrosion.
When an airplane structure, constructed of several metals, is exposed to corrosive environments such as
exhaust gases, moisture, waste water, and spillage, necessary factors for corrosive action are present
and some areas of the airplane are exposed to more corrosive contaminants than others [19]. Left
untreated, corrosion can make an aircraft unairworthy in just a few years.
4
This master thesis was created, to generate a general corrosion model of this material using the AUV
case study as a reference.
As for scientific value, the tests presented in this dissertation were never done before (at least
published, based on the author’s research) and it can encourage other scientists to invest more in the
conditioning variation test cycles instead of continuous exposure. In reality, there are many variations of
many parameters and the test has to be as accurate as possible.
1.4 Objectives
The objective of this thesis is to study the corrosion behavior of bare and anodized 7075-T651 aluminum
alloy under a wet/dry cycle salt spray test, with different washing methods. This quantification is based
in loss of mechanical properties and fatigue life.
The aim is to provide specific knowledge on the corrosion resistance of this particular aluminum alloy
in aggressive environments, as it is the salt spray exposure.
To have knowledge of the corrosion behavior of any material is important so that adequate materials
are chosen for different situations. Knowing their operational time, maintenance and restoration processes
can be scheduled and applied correctly which will help to prevent failures.
1.5 Thesis Outline
• Chapter 1 - Introduction - presentation of the motivation, main objectives, and introduction to the
approached subject;
• Chapter 2 - Background - introductory concepts and background theory about corrosion, aluminum
and aluminum corrosion. Some corrosion control strategies and typical tests for corrosion evaluation
are also briefly explained;
• Chapter 3 - Experimental Procedure - description of the used methods, materials, equipment and
techniques. Conditions for every test performed are presented;
• Chapter 4 - Experimental Results and Discussion - presentation of test results and analysis of
results. This chapter includes an observation of the corroded surface; a surface roughness profile
study and comparison; tensile test results for ultimate tensile strength, yield stress, Young’s modulus,
elongation at fracture and reduction area; fatigue test results with the decrease in fatigue life as
well as the evolution of the S-N curve;
• Chapter 5 - Finite Element Analysis - includes all the details of a tensile test simulation, on a
test specimen, with the finite element method by a linear static and nonlinear static approach.
Development of a practical application using experimental results with MATLAB R© and MSC Nastran;
5
• Chapter 6 - Conclusions - summary of the important points of the analysis made on the results
chapter, achievements, suggested improvements and follow-up work that can be made on this
topic.
6
Background
In this section, it will be presented some background information on aluminum and its alloys, as well
as the types of corrosion suffered by this material and typical mitigation strategies. In the end, there is
a brief explanation of some typical tests for the evaluation of corrosion effects in seawater, as well as
some concepts about fatigue and tensile tests useful to demonstrate the consequences of corrosion on
the properties of aluminum (Figure 2.1).
Figure 2.1: Summary of the Background Chapter
2.1 Definition of Corrosion
Corrosion is the deterioration of metals by an electrochemical process that degrades materials’ properties
because of an inevitable reaction with the environment. The term corrosion is often associated with
metallic materials, but all material types are susceptible to this phenomenon, even ceramics can undergo
degradation by selective dissolution. Every material has a strong driving force to achieve a lower energy
oxide state, hence the ability to corrode [20].
7
2.2 Historical Background of Aluminum
Due to the difficulty of extracting aluminum from its ore, the metal only became an economic competitor
in engineering applications towards the end of the 19th century [21] being the third most abundant
(7.5–8.1%) metal in the crust of the earth, almost twice as plentiful as iron. It is also second only to iron
as the most important metal used in industry and commerce [22].
With an annual world consumption of 25 million tons, aluminum is the leader in the metallurgy of
non-ferrous metals [23] and because of his properties, aluminum and its alloys have several engineering
applications in many different fields (e.g., transport, building, electrical engineering, and mechanical
construction).
Some of this advantageous properties are, [23]:
• lightness (its density is approximately 2700 kg/m3, which is almost three times less than that of
steel);
• thermal conductivity (with roughly 60% of the thermal conductivity of copper);
• electrical conductivity (around two-thirds of the electrical conductivity of copper);
• suitability for surface treatments;
• diversity of aluminum alloys;
• recyclable;
• non-toxic.
There are two ways to obtain aluminum: primary and secondary production.
The most widely used technology for the primary aluminum production involves two steps: extraction
and purification of aluminum oxide (alumina) from ores (primarily bauxite although alternate materials
can be used like kaolinite or nepheline) [21].
The secondary aluminum production consists of aluminum recovered from scrap. The energy required
to remelt secondary aluminum is only 5% of that required to produce new (primary) aluminum [21] which
is why the secondary industry is growing. However, secondary alloys always contain more undesirable
impurity elements that often lead to inferior properties when compared to the primary alloys.
2.2.1 Aluminum Alloys
Aluminum alloys are divided into two major categories: wrought composition and cast composition.
Cast aluminum is aluminum that has been melted, poured into a mold, and allowed to cool. Wrought
aluminum is aluminum that has been heated and then worked with mechanical tools [22].
The aluminum tested on this MSc Thesis is a wrought aluminum which is why it will be a focus point.
There is a total of nine series or families of wrought aluminum alloys that offer a big variety of
compositions, properties, and uses. The constant progress of the aluminum metallurgy industry leads to
the creation of better and better alloys for all types of applications. Different series can represent a huge
8
difference in the mechanical behavior of the alloys because different alloying addictions have a different
influence on strength and response to heat treatment [23].
The system for designating aluminum and aluminum alloys is standardized by the American National
Standards Institute (ANSI) Standard H35.1. For wrought aluminum alloys this nomenclature consists
of four numeric digits where the first digit indicates the principal alloying group or constituent(s). The
following itemization presents the major alloying element of each series as well as some main applications,
[22]:
• 1xxx, pure aluminum (≥ 99.00%), used primarily in the electrical and chemical industries;
• 2xxx, copper alloys, widely used in aircraft where there high strengths (yield strengths as high as
455MPa) are valued;
• 4xxx, silicon alloys, used in welding rods and brazing sheet;
• 5xxx, magnesium alloys, used in boat hulls, gangplanks and other products exposed to marine
environments ;
• 6xxx, silicon and magnesium alloys, commonly used for architectural extrusions;
• 7xxx, zinc alloys, used in aircraft structural components and other high-strength applications;
• 8xxx, other elements not covered by the other series mainly tin, lithium and/or iron ;
• 9xxx, unused series.
2.3 Aluminum Corrosion
The corrosion of metals is caused by an electrochemical process also know as oxidation-reduction
reaction in which the released energy is converted to electricity [24]. In this reaction, electrons are
transferred from one substance to another oxidizing the one that lost the electrons and reducing the
one that received the electrons. The site where oxidation occurs is called the anode and the one where
reduction occurs is called the cathode.
As an electrochemical process corrosion needs the presence of a metal, oxygen, and an electrolyte
(usually water) [24]. The metal being corroded, the anode, oxidizes releasing electrons. An example of
an oxidation half-equation is the aluminum atom becoming a positively charged ion (Equation 2.1 [24]).
Al(s)→ Al3+(aq) + 3e− (2.1)
Note: Aluminum has three oxide states (+1,+2 and +3), being the +3 the most common one.
A reduction half-equation is, for example, the reaction of oxygen with water to form hydroxide, which
consumes electrons (Equation 2.2 [24]).
9
O2(g) +H2O(l) + 4e− → 4OH−(aq) (2.2)
One of the reactions present on the aluminum corrosion process consists of the reduction of oxygen
and oxidation of aluminum that results in the aluminum oxide (Al2O3) formation (Equation 2.3 [24]).
4Al(s) + 3O2(g)→ 2Al2O3(s) (2.3)
Each half-reaction of the redox equation has a standard-state value of reduction potential (E0), the
value is the voltage associated with a reduction reaction at an electrode when all solutes are 1 M and all
the gases are at 1 atm. If this value is positive the reaction will occur spontaneously and if it is negative
it will occur spontaneously on the opposite way.
The half-reaction with the higher value of reduction potential will behave as the cathode and it is
considered the least active metal, the other substance with the highest oxidation potential (equal to the
negative value of his reduction potential) will behave as the anode. The overall reaction potential is
calculated by the sum of the two halves [24].
In terms of comparison with the formation of rust on iron, that is probably the most familiar example
of corrosion, aluminum has a more negative standard reduction potential meaning a greater tendency
to oxidize than iron. The reason why we do not see many aluminum corrosion products is that when
it oxidizes it creates a layer of aluminum oxide (Al2O3) that protects the metal surface from further
corrosion. The rust of the iron, however, is too porous to protect the iron and flakes off exposing
constantly new metal to corrosion [24].
Exposed to oxygen or to an oxidizing environment, aluminum and its alloys have excellent resistance
to corrosion, due to the protective film, and many of its applications depend on this very property. The
oxide film is, however, generally dissolved in most strong acids and bases which make the corrosion
process very fast.
2.3.1 Passivation of Aluminum
In the pH range of about 4 to 8.5, aluminum is passive (protected by its oxide film) (Figure 2.2). This
range varies somewhat with temperature, with the specific form of the layer and with the presence of
substances that can form soluble complexes or insoluble salts with aluminum. Out of the passivation
limits, aluminum generally corrodes in aqueous solution due to the dissolution of its oxides in acids or
bases, however, sometimes, even though it is outside of the passivation zone, the film is insoluble or the
oxidizing nature of the solution maintains the oxide layer [21].
At low nonoxidizing potential the metal itself, Al, is stable, this is called the immunity zone. On the
passivity zone, the aluminum oxide, Al2O3, is stable which origins the protective layer. At low pH and
high pH the aluminum ion, Al3+, and the aluminate ion, AlO− 2 , respectively, are stable, which means the
metal is subjected to corrosion on these areas [14].
10
Figure 2.2: Pourbaix diagram (potential vs pH) for aluminum showing the conditions of corrosion, immunity and passivation of aluminum at 25oC. (Source:[21])
2.4 Types of Aluminum Corrosion Mechanisms
Due to his high affinity to oxygen, when aluminum is in the presence of an oxidizing agent it acquires
rapidly the thin (about 0,5µ in air,[25]), compact and tightly adhering film mentioned in the previous
section. This layer also has a self-healing property, but, it can become unstable when exposed to
several special cases, like extreme pH levels, and its automatic renewal may not be fast enough to
prevent corrosion. The breakdown of the barrier translates on localized corrosion like pitting corrosion,
crevice corrosion, intergranular corrosion or stress corrosion, depending on the circumstances involved.
2.4.1 General Corrosion
General corrosion is the most common form of corrosion and is responsible for the greatest material loss.
In aluminum, this corrosion is rare except in highly acid or alkaline environments (Figure 2.2) where the
solubility of the oxide film is high. The dissolution rate can vary depending on the nature of the acid or
base. The rate of general corrosion can be easily calculated by measuring mass loss [23].
2.4.2 Galvanic Corrosion
When two dissimilar metals are immersed in an electrolyte solution and come into contact (e.g., direct
contact or wire connection) an electrical loop is created. The voltage difference will cause the flow of
electrons and, consequently, one of the metals will become the cathode and the other one the anode
(Section 2.3). The rate of attack can be controlled by the difference of potential between the metals, the
formation of passive films and the surface area of the anode compared to the cathode (the larger the
cathode compared to the anode the greater the galvanic current and corrosion velocity) [22].
Aluminum has a low standard-state reduction potential value which means that it is a very active
11
metal, more anodic, than most. When in contact with more cathodic metals, and an electrolyte, aluminum
corrodes faster than it would by itself on the same conditions.
2.4.3 Localized Corrosion
When corrosion is not distributed uniformly is considered localized corrosion. This type of corrosion is
less predictable and can have serious consequences causing fatal accidents in some circumstances.
Pitting Corrosion
Pitting is the most common form of corrosion found on aluminum and it is usually manifested by the
random formation of pits [21]. Depending on their composition and environmental conditions the different
alloys are affected differently. In general, the purer the alloy the higher pitting resistance it contains
because pitting occurs as a galvanic reaction between different elements on the alloy.
It is often the most damaging form of corrosion due to the ability of perforation through the depth of
the metal which results in loss of strength properties through increased stress concentrations at these
locations.
When the oxide layer breaks down the exposed metal gives up electrons easily and the reaction
initiates tiny pits with localized chemistry supporting a rapid attack. This attack will cause the pit
propagation (Figure 2.3).
Figure 2.3: Mechanism of Pitting Corrosion of Aluminum (Source: [23])
Outside the cavity the reduction of water ( 23O2+3H2O+6e− → 6OH−) and ofH+ (6H++6e− → 3H2)
occurrs, which will locally lead to an alkaline pH (basic environment) [23].
At the pit’s bottom aluminum oxidation will occur (Equation 2.1), this creates an electrical field that
shifts Cl− ion towards the pit bottom forming aluminum chlorides. The aluminum chlorides then suffer
hydrolysis forming aluminum hydroxides (Al(OH)3) plus excess of hydrogen and chloride ions (H+,
Cl−). This will lead to the acidification of the pit’s bottom (acid environment) and the media will become
very aggressive making the pit to autopropagate [23]. Because of this, pitting is considered to be an
autocatalytic process which means that once it has initiated, it alters the local conditions to promote
further pit growth.
12
Al(OH)3 will precipitate and pushed to the opening of the pit where it forms a deposit of white spots
[23].
The pitting potential states the value, in a particular solution, above which pits will initiate and
propagate and below which they may form but will not propagate [22]. It is frequently called the
breakdown potential of the oxide film.
This type of corrosion will occur in the pH range where the metal is passive. If there is an increase in
acidity or alkalinity beyond this passive range corrosion attack becomes more uniform and at least the
pitting potential is reached.
Crevice Corrosion
Crevice corrosion occurs when a wetted metallic surface is in close proximity to another surface, with
a small separation gap (50 − 200µm), where oxygen cannot penetrate [22]. It is an anaerobic reaction
where corrosion is accelerated due to the gap being less aired, with a weak surface, and containing a
stagnant solution often rich in salt. Also, hydrolysis reactions within the crevices can produce changes
in pH and chloride concentration in the crevice environment aggravating the corrosion.
2.4.4 Metallurgically Influenced Corrosion
The metallurgical heterogeneities (defects present in every metal, the impurities and alloying elements,...)
that exist even in pure single crystals can become a concern in terms of corrosion.
One example of this corrosion type is called dealloying in which one or more elements of the alloy
are selectively dissolved [22].
Intergranular corrosion is also a metallurgically influenced corrosion type in which there is a ”selective
attack of the grain boundary zone, with no appreciable attack of the grain body or matrix.” ([22], p.218).
The boundaries of the constituent grains of the material are more susceptible to corrosion than their
insides which cause failure on paths around the grains and compromises the mechanical properties of
the material.
A form of severe intergranular corrosion is the exfoliation corrosion that occurs when the corrosion
on the boundaries of the grains propagates on the rolling direction. This causes thin sheets that are not
attacked, but gradually pushed away, to peel off like pages in a book [23].
2.4.5 Stress Cracking Corrosion
Stress corrosion is an interaction between corrosive attack and sustained tension stress. It can be very
difficult to identify because there is little or no corrosion products and its only manifestation is cracks.
It can result in premature brittleness of a ductile material and if it is not recognized can lead to sudden
failure of a component [21].
13
2.4.6 Mechanically Assisted Corrosion
This topic is divided into two categories that can interact: erosion and fatigue.
Corrosion by erosion occurs in moving media and is related to the flow speed of the fluid. It can lead
to other forms of corrosion or even to failure by fatigue or stress cracking corrosion. It wears down the
material creating scratches and undulations orientated in the flow direction [22].
Corrosion fatigue is ”the sequential stages of metal damage that evolve with accumulated load cycling
in an aggressive medium and result from the interaction of irresistible cyclic plastic deformation with
localized chemical or electrochemical reaction.” ([22], p.272). The combination of a corrosive environment
with cyclic stresses reduces the life of the components below what is expected with fatigue alone.
2.4.7 Microbiologically Influenced Corrosion
This type of corrosion consists on the acceleration of the material deterioration by various microorganisms
like bacteria, fungi or algae [22].
Especially in seawater applications, the presence of all sort of microorganisms results in localized
corrosion (e.g., pitting, crevice) due to the attachment of microorganisms to the surface of the materials,
where they colonize, proliferate, and produce a biofilm. Dissolved oxygen, chloride, gradients of pH
and sulfate existing in the biofilm create the conditions for localized corrosion [26]. This accumulation of
unwanted material on the metal surface, with detriment function, is called fouling.
2.5 Corrosion Mitigation Strategies in Aluminum and Aluminum
Alloys
The rate of attack and the type of corrosion mechanism depend on the nature of the material and the
environment. The major focus of experts has been to achieve an equilibrium, between the environment
and the material, trying, this way, to reduce corrosion exposure.
Some of the most used corrosion mitigation strategies are [26, 27]:
• Coatings - the most common method for retarding corrosion, the concept is to simply insulate the
material from the corrosive environment (e.g., preventing the electrolyte from coming into contact
with the metals);
• Cathodic and anodic protection - the difference between anodic and cathodic protection is the
electrode that is protected. The cathodic protection protects the metal by making it the cathode,
which is easily understood because reduction occurs on the cathode. Anodic protection is also an
electrochemical method but is based on the passivity phenomenon (Section 2.3.1) which only a
limited number of metals have. By controlling the potential of the anode one can maintain it on the
passivity zone, therefore protecting the anode;
14
• Inhibitors - substances or mixtures that affect the electrochemistry of the system and can be
classified as cathodic or anodic. The anodic ones can control the rate of oxidation and the cathodic
ones can reduce or prevent the rate of reduction;
• Control of the environment - if the environment is less corrosive there will be less corrosion, but
it is not always possible (e.g., seawater);
• Design - on the design phase, appropriate materials shall be chosen as well as suitable control
measures. Also, some geometrical features must be paid attention to avoid crevices and exclude
water and/or dirt traps.
2.5.1 Anodizing Aluminum
Since the aluminum has a natural protective layer, one of the methods to slow down the corrosion
process is thickening that layer. This is the principle of anodization. The layer is grown by passing a direct
current through an electrolytic solution, with the aluminum object serving as the anode, creating more
aluminum oxide and, therefore, thickening the layer. This increases abrasion and corrosion resistance
and also provides better adhesion for paints and primers [28].
The majority of anodizing processes are “soft”. The current soft type of anodizing is done in chromic
or sulfuric acid baths and is used in almost 90% of the production with oxide layers of 5–18 µm produced.
The hard type of anodizing (sulfuric acid bath alone or with some additives) produces a thickness on the
order of 51 µm [22].
2.6 Aluminum in Seawater
The corrosion in seawater is quite different from the atmospheric corrosion and even from other aqueous
corrosion. In an aqueous system, the environmental variables that affect corrosion are mainly the pH
(acidity), the dissolved oxygen, the temperature and heat transfer, the velocity (fluid movement) and the
solution components and their concentration [29]. In seawater, in addition to all these factors, there is
also the salinity and the biological activity to enhance the process.
The typical forms of corrosion that occur in seawater are the pitting corrosion (the most common due
to the pH of seawater (8-8.5) which is the range where aluminum and its alloys are prone to this form of
corrosion) and galvanic corrosion (due to the very high electrical conductivity of seawater) [23].
2.7 Typical Tests for Corrosion Evaluation
The different types of corrosion tests can be categorized as laboratory, field and service tests [26].
The laboratory tests can be standardized or special tests for a specific purpose, inside a laboratory.
These tests are mostly used when there is lack of time or budget constraints and they are mainly
immersion tests, cabinet-controlled or different electrochemical procedures [26]. They are called accelerated
tests because of the rapid breakdown of the specimens in comparison to what would occur in service.
15
Field tests are carried out in the atmosphere or under immersed conditions at test sites specially
prepared for this purpose, they are simulations of the service conditions.
Service trials are a special form of field test carried out in an environment where the material will
actually be used.
The main purpose of corrosion testing is the evaluation of the materials corrosion behavior and their
selection for a given application. This allows the development of a reference database information of
what already exists. It can also provide a determination of the aggressiveness of the environment in
certain situations and the causes of certain corrosion failures [22].
2.7.1 Testing Modes
Depending on the final application of the material, different test modes can be chosen. Some examples
are wetting the metallic surface in a medium with saturated vapor (e.g., to simulate acid rain), spraying
aggressive media (e.g., to simulate ocean climate) or testing the impact of liquid and/or solid particles
(e.g., to simulate erosion) [22].
After choosing the testing mode it should be adapted for total, partial or alternate (wet/dry cycles)
immersion/exposure [22].
Salt Spray Corrosion Test
The salt spray corrosion test is a cabinet-controlled corrosion test in which the specimens are sprayed
with a salt solution. It is the most commonly used cabinet corrosion tests and there are three different
types: the neutral salt spray (NSS), the acetic acid salt spray (AASS) and the copper-accelerated acetic
acid salt spray (CASS). The difference between the types is the characteristics of the atomized solution
to which the specimens will be subjected [30].
Today, the NSS is standardized by ASTM B117, the Practice for Operating Salt Spray (Fog) Apparatus
[31]. On this standard, the test must be performed at a temperature of 35oC ± 2oC and must have a salt
solution with 5 wt.% NaCl and a pH between 6.5 and 7.2.
As a laboratory test, the salt spray is an accelerated process, in comparison with real environments,
and does not simulate real service conditions. Therefore, the test can be misleading due to the lack of
knowledge of the role and limitation of accelerated corrosion testing.
The test was originally designed for the quality control of a specific material or coating, and it is used
successfully by some industries for this purpose. It’s serious misuse is to compare different materials or
coatings that have different characteristics [32].
The test limitations begin with the percentage of salt in the solution, which is not necessarily a
realistic representation of the service conditions. A more serious limitation is that the chamber provides
a continuous environment with no changes in conditions. Corrosion in a cyclic environment can be
very different from corrosion in a continuous environment. Nowadays, a combinations of wet/dry cycles,
temperature cycles, and even salt solution concentration cycles have been more and more tested and
some modified standards have been created (ASTM G85 - Standard Practice for Modified Salt Spray
16
(Fog) Testing, ASTM D5894 - Standard Practice for Cyclic Salt Fog/UV Exposure of Painted Metal,
among others).
There are several reasons why tensile tests are performed [33]:
• To ensure quality, tensile properties are frequently included in material specifications;
• To select materials for engineering applications;
• During the development of new materials and processes so that different materials and processes
can be compared;
• To predict the behavior of a material under different forms of loading.
The main concern of these tests is to find the strength of the material, either the strength necessary
to enter plastic deformation (σy) or the maximum stress the material can withstand (σUTS). These
parameters are fundamental in engineering design, where appropriate safety factors are applied.
Other parameters can be measured like the elongation to fracture or the reduction area which gives
us information about the tensile ductility of the specimen. It represents how much the material can
deform before it fractures.
The primary use of the tensile machine is to create the stress-strain diagram which is a graphical
description of the amount of deflection under load for a given material [33]. The stress, σ, is obtained by
dividing the load at a given time, P , by the cross section of the specimen, A0 (Equation 2.4).
σ = P
A0 (2.4)
The strain, ε, is obtained by dividing the elongation of the gauge length, L, by the original gauge
length of the specimen, L0 (Equation 2.5).
ε = L
L0 (2.5)
2.9 Fatigue Tests
Fatigue is a process that causes premature failure or damage of a component subjected to repeated
loading [34]. The cyclic loading applied makes the material fail below its yield strength and can be
described by the stress ratio SR (SR = σmin/σmax), the maximum stress applied (σmax), the minimum
stress applied (σmin), the stress amplitude (σa = σmax−σmin
2 ), the stress range (σ = σmax − σmin) and
the mean stress (σm = σmax+σmin
2 ) at any combination of two plus the frequency of application.
The life of a material suffering the fatigue process can be divided into two phases - initiation life
and propagation life. The initiation includes the development and early growth of a small crack, the
propagation includes the total life spent growing a crack to failure [34].
17
• The stress-life approach;
• The strain-life approach;
• The fracture mechanics approach.
When a low load is applied the material lasts for a high number of cycles. This is called high cycle
fatigue (HCF) and the approach used to measure fatigue life is the stress-life approach due to the fact
that there is no significant plastic deformation. The basis of the stress-life method is a plot with stress,
σ (average, maximum or amplitude) vs cycles to failure, N. Certain materials, such as steel, have an
endurance or fatigue limit which is the stress value below which the material has an infinite life (for
engineering purposes, infinite life is 1 million cycles [34]).
Depending on the materials, when a high load is applied the number of cycles to failure is low.
This is called low cycle fatigue (LCF) and the approach used to measure fatigue life is the strain-life
method. In opposition to the method described earlier, the low cycle fatigue has an appreciable plastic
component which makes the stress-life method inadequate. In LCF the cyclic stress-strain response
and the material behavior are best modeled under strain-controlled conditions because the damage is
dependent on plastic deformation or strain [34]. Crack growth is not accounted for which makes the
strain-life method an initiation life estimate.
The fracture mechanics approach is used to estimate the propagation life.
The dividing line between low and high cycle depends on the material but usually falls between 10
and 105 cycles [34].
The fatigue test has the purpose of obtaining the expected fatigue life of the material under different
loading conditions. It can be used to obtain the maximum cycles that the material can withstand for a
specific maximum load or the maximum load that could be applied for a required number of cycles.
18
Chapter 3
Experimental Procedure
In this chapter, it will be presented the specimen description including the material characteristics, shape
and dimensions, manufacturing, quantity, and preparation.
Furthermore, the setup description, evaluated parameters and procedure of the corrosion test, tensile
test and fatigue test were explained in detail.
The goal of this tests is to quantify corrosion in terms of loss of mechanical properties and fatigue life
of the 7075-T651 aluminum alloy.
3.1 Specimen Description
3.1.1 Material
The material used in this experiment was the Aluminum Alloy 7075-T651. It belongs to the 7xxx
aluminum alloy series (Section 2.2.1) which means its main alloying element is zinc. The composition of
the material is presented in Table 3.1.
Table 3.1: Chemical Composition of 7075-T651 Aluminum Alloy (wt.%) [Source [35]]
Zn Mg Cu Cr Fe Si Mn Ti V Ga Ni 5.58 2.43 1.43 0.2 0.16 0.08 0.065 0.03 0.022 0.007 0.006
On the material designation, the four first digits (7075) represent the material composition. The rest
of the digits (T651) represents the treatment that the material has been subjected to. ’T’ means that the
alloy is heat treatable, ’6’ means that after the heat treatment and cooling the material was subjected to
artificial aging and finally, ’51’ means that after the artificial aging a stress-relieve by stretching (1-3%
permanent set) was made [22]. The heat treatment was made between 460-565oC to dissolve soluble
alloying elements, then, quenching (rapid cooling normally using water) is applied to retain the alloying
elements in solid solution. After, the material is artificially aged at 115-195oC to precipitate these
elements in an optimum size and distribution. The stress-relieve treatment after the aging has the
purpose of eliminating the residual stresses caused by the quenching since there is a surface-to-center
19
cooling gradient. This process enhances the resistance to stress cracking corrosion (Section 2.4.5) of
the material.
3.1.2 Shape and Dimensions
The specimens were chosen in accordance with the ASTM B557M, Standard Test Method for Tension
Wrought and Cast Aluminum and Magnesium Alloy Products (Metric) [36], as the Rectangular Tension
Test Specimens, Standard Sheet-Type 12.5mm Wide. This specimen type was chosen due to the
ease of fabrication and also because it is adequate for both tensile and fatigue testing (present in both
standards). The shape of the specimen is presented on Figure 3.1 and its dimensions in Table 3.2.
Figure 3.1: Rectangular Tension Test Specimen (Source: [36])
Table 3.2: Test Specimen Dimensions (mm) (Source: [36])
G - gage length 50.0 W - width 12.5 T - thickness 3.0 R - radius of the fillet 50.0 L - overall length 200.0 A - length of the reduced section 57.0 B - length of the grip section 50.0 C - width of the grip section 20.0
3.1.3 Manufacturing
The specimens were machined by an AWEA 3 axis CNC BM 1020F from 600×400mm aluminum sheets
with 3 mm in thickness. To help with vibrational issues a 600 × 400mm vacuum table was used to hold
down the sheet during machining (Figure 3.2).
After machining, half of the specimens were anodized according to AMS2469 which is a standard for
hard anodic coating on aluminum and aluminum alloys.
20
Figure 3.2: Half Cuted Aluminum Sheet on top of the Vaccuum Table [Courtesy of CEiiA]
3.1.4 Quantity
The test has a total duration of 20 wet/dry cycles that will be further explained in detail in Section 3.2.
Five exposure times were considered (1 cycle, 3 cycles, 9 cycles, 15 cycles and 20 cycles) to obtain a
more accurate tendency line of the loss of mechanical properties and fatigue life. Since the total duration
of the test is 20 cycles the points should be equally distributed between 0 and 20, however, the corrosion
rate is higher in lower exposure times [8, 13] so three points were chosen until mid-way (10 cycles) and
two after mid-way.
For each time interval, thirty-six specimens were manufactured, eighteen for fatigue testing and
eighteen for tensile testing. The specimens were further divided into anodized and not anodized and on
the three washing methods. To get representative average values of all the parameters for each set of
conditions three specimens were tested (Table 3.2).
Table 3.3: Quantity and Variations of the Corroded Samples for Each Exposure Time
Quantity
Do Not wash Tensile 3 Fatigue 3
Anodized
Do Not wash Tensile 3 Fatigue 3 Total: 36
An addition of twelve specimens was also machined for the uncorroded tests, half for fatigue and half
for tensile studies further divided into anodized and bare specimens (Table 3.4).
21
Quantity
Fatigue 3
Anodized Tensile 3 Fatigue 3 Total: 12
With the goal of observing how S-N fatigue curves (stress-life approach in Section 2.9) would change
with increasing corrosion, an extra 15 samples were considered. This type of curve is very important
when a High Cycle Fatigue study is made. These specimens were divided into three groups of five
specimens where each group corresponds to a different corrosion exposure time. All fifteen specimens
were not anodized and washed with freshwater (Table 3.5).
Table 3.5: Quantity and Variations of the Samples used to built S-N curves
Exposure Time (cycles of corrosion) Quantity
Not Anodized Washed with Freshwater
0 5 3 5
15 5 Total: 15
3.1.5 Preparation
Before the beginning of the tests, the specimens were cleaned with light duty tissue wipers to remove
any oils or contaminants originated on the manufacturing and transportation process.
The specimens that will go inside the salt spray chamber were masked on the grip areas with a
3MTM Corrosion Resistant Duct Tape (50mm × 50m) in order to protect these areas from corrosion.
In the tensile and fatigue tests the grip areas are where the machine holds the specimens. Therefore,
the masking prevents these areas from being compromised with any residue or break due to corrosion
which will discard the tensile or fatigue test.
With a label printer, sticker labels were made and placed on top of the duct tape to identify every
specimen according to their characteristics and path variations. The labels and nomenclature used are
listed in Appendix A, Tables A.1, A.2, A.3, A.4 and A.5. The final result is showed in Figure 3.3.
The clean, masked and labeled specimens were then, in accordance to [37]:
• Measured in three points in width with a MITUTOYO Digimatic Caliper to the fourth significant
figure;
• Measured in three points in thickness with a MITUTOYO Digimatic Micrometer to the fourth significant
figure;
• Weighed three times with a KERN precision analytical balance to the sixth significant figure.
22
Figure 3.3: Masked and Labeled Specimens (Not Anodized - Top; Anodized - Bottom)[Courtesy of CEiiA]
3.2 Corrosion Test
The corrosion test consists of a wet/dry cycle of an 8h period on a salt spray chamber and a 15h period
on a humidity and temperature controlled chamber.
Between periods three washing methods were considered: washing with Nitric Acid, washing with
freshwater and to not wash.
The total duration of the test is 20 cycles which makes a total of 460h.
The test was chosen to simulate a device that goes in seawater for a few hours and then gets stored
in a warehouse during a certain period of time until it goes in seawater again. This process is done
cyclically during the lifetime of the device.
The duration of the wet and dry periods was chosen to be eight and fifteen hours, respectively, due
to the laboratory everyday work schedule.
All the equipment, software and support material used in this test is listed in Appendix B.
3.2.1 Wet Period
On each cycle, to simulate a corrosive environment, for 8h the specimens were inside the salt spray
chamber following the ASTM B117 - Standard Practice for Operating Salt Spray (Fog) Apparatus [31].
The test was performed at the temperature of 35oC ± 2oC, with a salt solution with 5wt.% NaCl and
pH between 6.5 and 7.2.
The machine used was the Dry Corrosion Test Cabinet (DCTC) 1200 by the brand ACS (Figure 3.4).
The DCTCTM is a system designed to recreate the corrosion process on all kinds of surfaces in the
laboratory. The shape, constituents, and dimensions are in accordance with the followed standard [31].
The machine’s technical features are presented in Table 3.6.
23
Figure 3.4: ACS Dry Corrosion Test Cabinet 1200 [Courtesy of CEiiA]
Internal Volume (lid included) (l) 1215 Useful Volume (l) 955
Internal dimensions (mm) Width = 1700 Depth = 650
Height = 820 (+280 at the cover top)
External dimensions (mm) Width = 2680 Depth = 850
Height = 1290 Temperature Range (oC) amb. ...+55 Main used Power (kW) 3.5 Supply voltage (Vac) 230V (10%) 50Hz 1+T
Table 3.6: Technical Features of the Dry Corrosion Test Cabinet (Source: [38])
The machine has on the left side (Figure 3.4) a 120-liter salt solution tank with a solution level
indicator. The salt solution is pumped from the tank and mixed with compressed air which causes the
solution to atomize. This creates a dense salt water fog which origins a corrosive environment for the
samples inside the chamber test room.
Supports for the specimens were made from glass-fiber reinforced epoxy resin (Figure 3.5).
Figure 3.5: Empty Support [Courtesy of CEiiA]
Each support has a maximum of 18 specimens placed at a 20o inclination [31]. On the front, it has a
24
label with a number (that represents the time of exposure) and a letter (F - Fatigue; T - Tensile) (Figure
3.6). Red tape divides the support into three parts according to the different washing methods. The first
6 will be washed with Nitric Acid, the middle six will be washed with freshwater and the last six will not
be washed.
Figure 3.6: Filled Support [Courtesy of CEiiA]
The supports were suspended by wooden bars inside the test room of the chamber parallel to the
principal direction of flow of fog through it, based upon the dominant surface being tested [31].
3.2.2 Between the Wet and the Dry Period
Between an 8h wet and a 15h dry period, three washing methods were tested:
1. Washing with Nitric Acid (HNO3) - following the ASTM G1 - Standard Practice for Preparing,
Cleaning and Evaluating Corrosion Test Specimens [37] - the specimens were immersed in a
Nitric Acid solution (specific gravity of 1.42), during 1 to 5 min at the temperature 20 to 25oC;
2. Washing with clean freshwater - gently washed or dipped in clean running water;
3. Not wash.
The Nitric Acid washing method was chosen because it removes extraneous deposits and bulky
corrosion products to avoid reactions that may result in excessive removal of the base metal of aluminum
and aluminum alloys [37].
A freshwater flush is a typical maintenance procedure of saltwater devices.
The non-wash method will evaluate how corrosion would evolve when a deposition of corrosion
products is present on the surface of the metal.
If the corrosion cycle is an extraction cycle, the specimens shall be weighed three times to the sixth
significant figure, between periods, after being washed. This decision was made to obtain results as
accurate as possible since the washing methods remove the corrosion products and a proper value of
25
weight loss can be calculated. The not washed specimens are not considered for weighing since the
results will include the salt and corrosion products deposition on the surface and will not be authentic.
3.2.3 Dry Period
On each cycle, for 15h the specimens were inside a humidity and temperature controlled chamber.
The temperature chosen was 25oC and the relative humidity (RH) was 80% to simulate a warehouse
near the sea [39].
The machine used was an Aralab Climatic Chamber Fitoclima 500 EP20. The chamber’s dimensions
and ranges are presented in Table 3.7.
Internal dimensions (mm) Width = 700 Depth = 780 Height = 920
External dimensions (mm) Width = 1000 Depth = 1500 Height = 1850
Temperature Range (oC) -20oC - +180oC Humidity Range (% RH) 10% - 98%
Table 3.7: Technical Features of the Aralab Climatic Chamber Fitoclima 500 EP20 (Source: [40])
Inside the chamber, the supports were placed on top of stainless steel shelves.
3.2.4 Parameters
A MITUTOYO Surface Roughness Measuring Tester was used to measure the roughness profile of a
chosen 10× 12.5mm2 square on every specimen.
A roughness profile is a measure that classifies the surface as rough or smooth. It consists of a
wave graph with points above zero and bellow zero, being zero the ideal surface. The more deviation
the points have from zero the rougher is the surface. The local maximum points above zero are called
peaks and the local minimum points under zero are called valleys [41].
Three profiles were measured, to obtain average values, and the evaluated parameters from the
results were:
• Maximum valley depth - the maximum value below zero of the profile;
• Average valley depth - the average value of the valley depths of the profile;
• Valley density - the number of valleys in 1 mm2.
After each corrosion exposure interval, each specimen was observed microscopically with a Nikon
ECLIPSE MA 100 compact-size inverted microscope. With the Nikon NIS Elements D software, some
photographs were taken of pits, cracks and other interesting microstructures.
From the weight loss of the specimens an assessment of corrosion damage can be calculated. The
average corrosion rate may be obtained by equation 3.1, [37]:
26
A× T ×D (3.1)
where K is a constant that depends on the pretended corrosion rate units (K = 3.45 × 106) for
corrosion rate unit millimeters per year (mpy)); W is the mass loss in grams; A is the exposed area in
cm2; T is the time of exposure in hours and D is the density of the material in g/cm3.
3.2.5 Procedure
The corrosion test was performed according to the following protocol:
1. Prepare the salt solution by dissolving 5± 1 parts by mass of sodium chloride in 95 parts of water
following the maximum allowable limits for impurity levels described in Table 1 of the ASTM B117,
[31];
2. Fill the tank with the salt solution;
3. Place the prepared specimens on the supports (the specimens shall be supported or suspended
between 15o and 30o from the vertical);
4. Place the supports inside the salt spray chamber;
5. Set and maintain the chamber temperature at 35± 2oC and the pH of the collected solution at this
chamber temperature shall be between 6.5 and 7.2;
6. After 8h remove the specimens from the salt spray chamber;
7. Immerse every specimen marked with a “B”, Appendix A, on a Nitric Acid solution (sp gr 1.42), 1
to 5 min at 20 to 25oC;
8. Gently wash or dipp in clean running water every specimen marked with a “C”, Appendix A ;
9. Do not wash every specimen marked with a “D”, Appendix A;
10. If the corrosion cycle is an extraction cycle, weigh the specimens that were washed (marked with
”B” and ”C”);
11. Place the specimens on their previous places on the supports;
12. Place the supports inside the environmental chamber;
13. After 15h switch the supports to the salt spray chamber.
Repeat this procedure (from step number 4 to 13) for 20 cycles adding specimen extractions on
pre-determined exposure times.
3.2.6 Scheme
The following figure, 3.7, represents a schematic of the procedure of the tests.
Figure 3.7: Scheme of the Test Procedure
28
3.3 Tensile Test
To evaluate the loss in mechanical properties, after the corrosion process, an uniaxial tensile test was
performed on half of the specimens from each exposure time.
The uniaxial tensile test is known as a basic and universal engineering test, [33], to obtain material
properties such as ultimate strength, yield strength, % elongation at fracture, % area of reduction,
Young’s modulus and others.
All the equipment, software and support material used in this test is listed in Appendix B.
3.3.1 Setup
The test was performed following the standard ASTM B557M - Standard Test Method for Tension
Wrought and Cast Aluminum and Magnesium Alloy Products (Metric) [36].
The machine used was an MTS Landmark Servohydraulic Testing Machine with a load cell of 100kN,
model 370.25, and an integrated Linear Variable Differential Transformer (LVDT), model 39-075-103,
used to measure linear displacement (Figure 3.8).
Figure 3.8: MTS Landmark Servohydraulic Testing Machine with a Load Cell of 100kN, Model 370.25 [Courtesy of CEiiA]
The machine has two grips showed in detail in Figure 3.9. The load cell is placed above the top grip
(black cylinder on top of Figure 3.9) and the LVDT is connected to the bottom grip. The grips hold the
specimen during the test and a predetermined longitudinal force (depending on the rate of displacement
requested) is applied. The top grip is static and the bottom grip moves at a predetermined speed until
fracture of the specimen.
29
Based on research of tensile tests on aluminum alloys, the chosen bottom grip’s rate of displacement
was 1mm/min ([8], among others).
Figure 3.9: MTS Landmark Servohydraulic Testing Machine’s Grips and Load Cell [Courtesy of CEiiA]
During the experiment, an MTS 632.85 Biaxial Extensometer is mounted on the specimens to
measure axial and transverse deflections (Figure 3.10). The extensometer has three sensor units, two
axial and one transverse.
Figure 3.10: Specimen during a tensile test with a biaxial extensometer [Courtesy of CEiiA]
The software used was the MTSTestSuiteTM TW Elite.
While the test is running the software records force values, displacement values, two axial deflection
values, one transverse deflection value and time values with a recording frequency of 10Hz. This values
will be further analyzed with an Excel Macro, designed purposely for the task, to get the curves and
parameter values needed.
3.3.2 Parameters
The typical stress-strain curve achieved with an uniaxial tensile test is represented in Figure 3.11.
Figure 3.11: Engineering Stress-Strain relationship under uniaxial tensile loading [Source: [42]]
The parameters obtained from this experiment were:
• Ultimate Tensile Strength, MPa (σUTS) - which is the maximum point reached by the stress-strain
curve (Figure 3.11);
• Yield Stress, MPa (σy) - which represents the transition point from the elastic domain to the plastic
domain. This parameter is calculated with the offset method described on the ASTM B557M
standard. The method consists on drawing a line parallel to the one on the elastic domain which
intersects the strain axis at 0.2%. The intersection of this line with the stress-strain curve is the
yielding point (Figure 3.12 with m = 0.2%, r = yielding point, R = yield stress and slope oA = slope
mn);
Figure 3.12: Stress-Strain Diagram for Determination of Yield Strength by the Offset Method [Source: [36]]
31
• Young’s Modulus (E) GPa - which is the slope of the stress-strain curve on the elastic area (Figure
3.11);
• Elongation at Fracture (e%) - which is the amount of deformation until fracture.
%Elongation = L
L0 × 100 (3.2)
This is an indicator of ductility and it corresponds to the strain value at the fracture moment in
percentage (Figure 3.11 - strain to fracture value);
• Reduction of Area (%RA) - which represents the reduction of the cross sectional area of the
fractured zone of the specimen when compared to the original area.
%RA = A
This is also a ductility measure to characterize the deformational characteristics of the material.
The final area is calculated by doing an average of three measurements of thickness and three
measurements of width on the necking area of each specimen;
The values were compared with the undamaged ones and a graph was made to see the evolution of
every property in function of the number of complete cycles of exposure.
3.3.3 Procedure
The tensile tests were carried out according to the following procedure:
1. Preparation of the test machine – upon startup or following a prolonged period of machine inactivity,
exercise or warm up the test machine to normal operating temperatures to minimize errors that may
result from transient conditions;
2. Open MTSTestSuiteTM TW Elite and open the test program;
3. Mark the middle of the specimen as well the grip zone distance;
4. Mount the extensometer at mid-section of the specimen;
5. Place the specimen in the wedge grips and close them;
6. Run the test program;
7. The test stops when a break is detected (complete separation);
8. Remove the extensometer;
9. Remove the specimen;
32
3.4 Fatigue Test
Fatigue is a phenomenon inseparable from corrosion [6]. As it has been said in Section 2.9, the
fatigue process includes an initiation stage and a propagation stage. Corrosion initiates cracks and also
propagates them which accelerates the initiation and propagation stages reducing the fatigue lifetime
exponentially.
To evaluate the fatigue life of the specimens, after the corrosion test, a fatigue test was performed on
half of the specimens at each time interval.
A fatigue test determines a material’s ability to withstand cyclic fatigue loading conditions. The fatigue
study is very relevant to several applications like aeronautical or anything that is subjected to dynamic
loads, that change through time, not necessarily cyclic [34].
All the equipment, software and support material used in this test is listed in Appendix B.
3.4.1 Setup
The test was performed following the ASTM E466 - Standard Practice for Conducting Force Controlled
Constant Amplitude Axial Fatigue Tests of Metallic Materials [43].
The machine used in this experiment was the MTS Landmark Servohydraulic Testing Machine with
a load cell of 50kN, model 370.10, and an integrated Linear Variable Differential Transformer (LVDT),
model 39-075-103, used to measure linear displacement (Figure 3.13). This machine has a similar
composition and operation to the tensile testing machine presented in Section 3.3.1.
Figure 3.13: MTS Landmark Servohydraulic Testing Machine with a Load Cell of 50kN, Model 370.10 [Courtesy of CEiiA]
33
The grips hold the specimen during the test and a predetermined cyclic force (depending on the
frequency, tension ratio and maximum tension chosen) is applied. The top grip is static and the bottom
grip moves up and down until fracture of the specimen.
The software used was the MTSTestSuiteTM Multipurpose Elite. The software records the cycle
counting until fracture (total separation) of the specimen or until it reaches 1 million cycles. Being
aluminum a non ferrous metal it does not have a true fatigue limit because the S-N curve never becomes
horizontal [42]. Due to practical purposes, 1 million cycles was considered the test limit.
The conditions imposed were:
• SR=0.1 which is a typical value used to test aircraft components since the tested material has a
lot of applications on the aircraft domain [44];
• f=10Hz. ”In the typical regime of 10−2 and 102 Hz over which most results are generated, fatigue
strength is generally unaffected for most metallic engineering materials” ([43], p.4);
• σmax = 165.5MPa. After several tests with different maximum loads the chosen one was 165.5MPa
due to the fact that the specimen would last near 1 million cycles;
Note: For the S-N curve specimens the several fatigues tests were carried out with variable maximum
stress so that the S-N curve could be obtained. These 15 test specimens were tested at the Mechanical
Testing Laboratory of Instituto Superior Tecnico with the same testing conditions and procedure.
3.4.2 Parameters
The parameter evaluated was the number of cycles to failure (N) with the application of the same
maximum load, frequency and SR. For the S-N curve samples the same parameter was evaluated,
N, with variable maximum load and constant frequency and SR. The uncorroded specimens’ fatigue life
was compared to the corroded samples with different exposure times and plot