Top Banner
Corrosion Behavior of 7075-T651 Aluminum Alloy under Different Environments Beatriz Almeida Ferreira Thesis to obtain the Master of Science Degree in 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
130

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

Jun 08, 2022

Download

Documents

dariahiddleston
Welcome message from author
This document is posted to help you gain knowledge. Please leave a comment to let me know what you think about it! Share it to your friends and learn new things together.
Transcript
Page 1: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Corrosion Behavior of 7075-T651 Aluminum Alloy underDifferent Environments

Beatriz Almeida Ferreira

Thesis to obtain the Master of Science Degree in

Aerospace Engineering

Supervisors: Prof. Luís Filipe Garlão ReisEng. Bruno Aires de Albuquerque e Concha de Almeida

Examination Committee

Chairperson: Prof. Filipe Szolnoky Ramos Pinto CunhaSupervisor: Prof. Luís Filipe Garlão Reis

Member of the Committee: Prof. Maria de Fátima Reis Vaz

November 2017

Page 2: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

ii

Page 3: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Dedicated to my parents

iii

Page 4: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

iv

Page 5: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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. Luıs 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

Page 6: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

vi

Page 7: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Resumo

As ligas de alumınio 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 alumınio 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 perıodo seco e o molhado tres metodos de lavagem foram considerados:

acido nıtrico, 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 superfıcie detetou-se que o principal mecanismo de corrosao foi o pitting. As covas

comecaram a aparecer em lugares aleatorios na superfıcie 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

corroıdo e gerar as suas propriedades na linguagem do programa Nastran. Estas propriedades podem

ser usadas para qualquer simulacao em qualquer peca feita de alumınio 7075-T651.

Palavras-chave: Liga de Alumınio 7075-T651, Teste em Camara de Nevoeiro Salino, Corrosao,

Pitting, Propriedades de Tracao, Vida a fadiga

vii

Page 8: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

viii

Page 9: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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.

Keywords: Aluminum Alloy 7075-T651, Salt Spray Test, Corrosion, Pitting, 7075-T651 Tensile

Properties, Fatigue Life

ix

Page 10: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

x

Page 11: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Contents

Acknowledgments . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . v

Resumo . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . vii

Abstract . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ix

List of Tables . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xv

List of Figures . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xvii

Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxi

Glossary . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . xxiii

1 Introduction 1

1.1 Motivation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.2 Topic Overview . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1

1.3 Framework . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 3

1.4 Objectives . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

1.5 Thesis Outline . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5

2 Background 7

2.1 Definition of Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 7

2.2 Historical Background of Aluminum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.2.1 Aluminum Alloys . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 8

2.3 Aluminum Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 9

2.3.1 Passivation of Aluminum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 10

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

2.4.4 Metallurgically Influenced Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.4.5 Stress Cracking Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 13

2.4.6 Mechanically Assisted Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . . . . 14

2.4.7 Microbiologically Influenced Corrosion . . . . . . . . . . . . . . . . . . . . . . . . . 14

2.5 Corrosion Mitigation Strategies in Aluminum and Aluminum Alloys . . . . . . . . . . . . . 14

2.5.1 Anodizing Aluminum . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 15

xi

Page 12: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

2.6 Aluminum in Seawater . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 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

3.1.1 Material . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 19

3.1.2 Shape and Dimensions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3.1.3 Manufacturing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 20

3.1.4 Quantity . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 21

3.1.5 Preparation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 22

3.2 Corrosion Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3.2.1 Wet Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 23

3.2.2 Between the Wet and the Dry Period . . . . . . . . . . . . . . . . . . . . . . . . . . 25

3.2.3 Dry Period . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.2.4 Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 26

3.2.5 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 27

3.2.6 Scheme . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 28

3.3 Tensile Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

3.3.1 Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 29

3.3.2 Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 31

3.3.3 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32

3.4 Fatigue Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3.4.1 Setup . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 33

3.4.2 Parameters . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

3.4.3 Procedure . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 34

4 Experimental Results and Discussion 35

4.1 Corrosion Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

4.1.1 Surface Evaluation . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 35

4.1.2 Corrosion Rate Determination . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 43

4.1.3 Maximum Valley Depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

4.1.4 Average Valley Depth . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 45

4.1.5 Valley Density . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 46

4.2 Tensile Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 47

4.2.1 Ultimate Tensile Strength . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 48

4.2.2 Yield Stress . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 49

4.2.3 Young’s Modulus . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 51

xii

Page 13: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

4.2.4 Elongation at Fracture . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 52

4.2.5 Reduction of Area . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 53

4.3 Fatigue Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 54

4.3.1 The S-N Curve . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 57

5 Finite Element Analysis 59

5.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 59

5.2 Software . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

5.3 Meshing . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 60

5.4 Load and Boundary Conditions . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 62

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

6 Conclusions 73

6.1 Achievements . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 74

6.2 Future Work . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 75

Bibliography 77

A Specimen Labeling & Tabulated Data 81

A.1 Nomenclature . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 81

A.2 Specimen’s Labels . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 82

A.3 Tensile Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84

A.4 Fatigue Test Results . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 90

B Equipment, Software and Support Material List 93

B.1 Corrosion Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 93

B.2 Tensile Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

B.3 Fatigue Test . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 94

C Output.txt Example File 97

D MATLAB R© Code 99

xiii

Page 14: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

xiv

Page 15: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

List of Tables

1.1 Medusa Deep Sea Main Characteristics . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

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

Page 16: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

xvi

Page 17: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

List of Figures

1.1 Medusa Deep Sea . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4

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

Page 18: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

4.12 AC - 9 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.13 AC - 15 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.14 AC - 20 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38

4.15 A - 0 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

4.16 AD - 1 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

4.17 AD - 3 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

4.18 AD - 9 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

4.19 AD - 15 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

4.20 AD - 20 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 40

4.21 NA - 0 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.22 NAB - 1 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.23 NAB - 3 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.24 NAB - 9 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.25 NAB - 15 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.26 NAB - 20 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 41

4.27 NA - 0 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

4.28 NAC - 1 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

4.29 NAC - 3 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

4.30 NAC - 9 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

4.31 NAC - 15 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

4.32 NAC - 20 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 42

4.33 NA - 0 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

4.34 NAD - 1 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

4.35 NAD - 3 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

4.36 NAD - 9 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

4.37 NAD - 15 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

4.38 NAD - 20 cycles . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 44

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

Page 19: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

xix

Page 20: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

xx

Page 21: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Nomenclature

Greek symbols

∆ Variaton.

σ Stress.

ε Strain.

Roman symbols

A Area.

b Y-intercept of the Plastic Line.

C Elasticity Matrix.

D Density.

d Divergence Rate.

E Young’s Modulus.

e Elongation at Fracture.

ET Tangential Modulus.

F Internal Loads.

f Frequency.

[K] Stiffness Matrix.

[KT ] Tangential Stiffness Matrix.

K Corrosion Rate Constant.

L Length.

m Slope of the Plastic Line.

N Number of Cycles to Failure.

P External Loads.

xxi

Page 22: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

R Load Error Vector.

RA Reduction Area.

SR Stress Ratio.

T Time.

t Thickness.

u Displacement.

W Mass loss.

Subscripts

0 Initial Value.

a Amplitude.

m Average.

max Maximum Value.

min Minimum Value.

n Number of increments.

p Pitting.

UTS Ultimate Tensile Strength.

y Yield Stress.

Superscripts

i Iteration number.

T Transpose.

xxii

Page 23: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 24: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

xxiv

Page 25: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Acronyms

AASS Acetic Acid Salt Spray. 16

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

SPC Single Point Constraint. 62

UTS Ultimate Tensile Strength. 48, 68, 70, 72

xxv

Page 26: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

xxvi

Page 27: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Chapter 1

Introduction

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

Page 28: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 29: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 30: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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 AUVSize 2.8m x 1.65m x 0.75m

Weight 350kgEndurance 7h

Range 30kmMaximum Operational Depth 3000m

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

Page 31: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

The present thesis is divided into six chapters:

• 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

Page 32: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

• 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

Page 33: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Chapter 2

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

Page 34: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 35: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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;

• 3xxx, manganese alloys, general-purpose alloys for architectural applications;

• 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

Page 36: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 37: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 38: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 39: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 40: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 41: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

• 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

Page 42: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 43: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

(Fog) Testing, ASTM D5894 - Standard Practice for Cyclic Salt Fog/UV Exposure of Painted Metal,

among others).

2.8 Tensile Tests

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

Page 44: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

There are three primary fatigue analysis methods [34]:

• 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

Page 45: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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 Ni5.58 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

Page 46: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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 50.0W - width 12.5T - thickness 3.0R - radius of the fillet 50.0L - overall length 200.0A - length of the reduced section 57.0B - length of the grip section 50.0C - 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

Page 47: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Each exposure time

Not Anodized

Wash HNO3Tensile 3Fatigue 3

Wash Freshwater Tensile 3Fatigue 3

Do Not wash Tensile 3Fatigue 3

Anodized

Wash HNO3Tensile 3Fatigue 3

Wash Freshwater Tensile 3Fatigue 3

Do Not wash Tensile 3Fatigue 3Total: 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

Page 48: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Table 3.4: Quantity and Variations of the Undamaged Samples

Quantity

UndamagedNot Anodized Tensile 3

Fatigue 3

Anodized Tensile 3Fatigue 3Total: 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 AnodizedWashed with Freshwater

0 53 5

15 5Total: 15

A total of 207 specimens were tested.

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

Page 49: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 50: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 3.4: ACS Dry Corrosion Test Cabinet 1200 [Courtesy of CEiiA]

Internal Volume (lid included) (l) 1215Useful Volume (l) 955

Internal dimensions (mm) Width = 1700Depth = 650

Height = 820 (+280 at the cover top)

External dimensions (mm) Width = 2680Depth = 850

Height = 1290Temperature Range (oC) amb. ...+55Main used Power (kW) 3.5Supply 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

Page 51: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 52: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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 = 700Depth = 780Height = 920

External dimensions (mm) Width = 1000Depth = 1500Height = 1850

Temperature Range (oC) -20oC - +180oCHumidity 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

Page 53: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

CorrosionRate =K ×W

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.

27

Page 54: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 55: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 56: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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.

30

Page 57: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 58: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

• 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

A0× 100 (3.3)

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;

10. Measure width and thickness at the fracture section.

32

Page 59: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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

Page 60: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

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 plots were made.

3.4.3 Procedure

The fatigue 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 Multipurpose Elite and open the test program;

3. Mount the specimen in the wedge grips and close them, checking the alignment;

4. Run the test program;

5. The test stops when a break is detected (complete separation) or when the defined maximum

number of cycles is reached;

6. Remove the specimen.

34

Page 61: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Chapter 4

Experimental Results and Discussion

In this chapter, results will be presented in order to demonstrate the influence of corrosion cycles in the

mechanical behavior of the aluminum alloy 7075-T651.

The results from the surface evaluation followed by a maximum valley depth, average valley depth,

and valley density analysis are given.

The tensile test results including properties like ultimate tensile strength, yield stress, Young’s modulus,

elongation at fracture and reduction area will also be discussed.

Lastly, the fatigue test results which include the number of cycles to failure of every tested specimen

along the corrosion exposure time in addition to the evolution of the S-N curve of the not anodized and

washed with freshwater specimens.

4.1 Corrosion Test Results

4.1.1 Surface Evaluation

Before starting this test campaign, on stage zero, microscopic images of uncorroded specimens, Figures 4.1

and 4.2, were recorded for further comparison.

Figure 4.1: Uncorroded and Anodized Figure 4.2: Uncorroded and Not Anodized

*The red line represents 100 µm.

35

Page 62: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

These figures were used as references to analyze the corrosion impact and types along the cycles

of exposure. There is a clear difference in color between the anodized and the not anodized figure (the

color present on the anodized sample is the color of the aluminum oxide) and some pores of the oxide

layer can also be seen on the anodized sample.

Examination of corroded surfaces revealed pitting corrosion as the predominant form of localized

corrosion. In the early stages of corrosion, the initiation of pits is confined to very few sites although they

started to appear already on the first cycle.

Pits can occur due to the breakdown of the passive film which is a rare occurrence (if nothing

influences it) that happens extremely rapidly on a very small scale [4]. The localized breakdown of

the passive film results in accelerated dissolution of the underlying metal.

On the aluminum alloys case, the alloy composition and microstructure can have strong effects on

the tendency for an alloy to pit due to the interaction of the aluminum (which is a very active metal) with

the other alloying elements generating an oxidation-reduction reaction (Section 2.3).

The constituent particles formed by the various alloying elements are also responsible for pitting [3].

Some particles acted as cathodes promoting the matrix dissolution around them and other acted as

anodes dissolving themselves.

The pitting mechanism can be divided into three main stages : nucleation, propagation and repassivation.

The nucleation stages is when pits are initiated, the propagation is when they grow and the repassivation

is when they cease to continue to grow [4]. All three stages occur simultaneously leading to a variation

in the location, depth, severity and density of pitting which is why it is such a complex phenomenon to

predict.

The next subsections demonstrate the degradation of the specimens along the corrosion process of

each condition variation tested. Inside each case, there are relevant figures of every extraction point.

Anodized Washed with Nitric Acid (AB)

For each exposure time considered, a microscopic image was selected representing the state of the

specimens, Figures 4.3, 4.4, 4.5, 4.6, 4.7 and 4.8.

Comparing Figure 4.3 with all the others there is a clear difference in color. Between periods, these

specimens were washed with nitric acid which dissolved the aluminum oxide that constitutes the oxide

layer, discoloring the specimens. This brutal discoloration happened after the very first wash and after

that, the color continued to go back to the specimen’s original color, before anodization.

In Figure 4.4 we can also observe some dispersed pits in a circular shape and from the first to the

third cycle (Figure 4.5) of exposure there were no significant changes in these specimens.

Between the third and the ninth cycle the density of pits increased significantly, as Figure 4.6 shows,

and pits started to lose their circular shape. The circular pits started to merge with close neighboring

pits preferably along the grain direction.

The samples were more corroded on the edges than on the middle part of the exposed surface.

Figure 4.7 was taken from an area near the edge of the specimen where it can be observed the

merging and connection of pre-formed pits. Increasing this coalescence the density of pits, even though

36

Page 63: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 4.3: A - 0 cycles Figure 4.4: AB - 1 cycles

Figure 4.5: AB - 3 cycles Figure 4.6: AB - 9 cycles

Figure 4.7: AB - 15 cycles Figure 4.8: AB - 20 cycles

*The red line represents 100 µm.

it increased the number of pits, decreases. Pits started to appear more in quantity, to be more irregular

in shape, bigger and deeper.

At cycle 20, to the naked eye, specimens were very damaged. The pits were visible all over the

surface of the specimen. Figure 4.8 shows a large pit found in one of the specimens with an approximate

area of 0.256 mm2. Around the perimeter of this pit, there are pit lines coming out meaning the

continuous propagation of the pit. These pit lines initiated on the edges of the bigger pit and are growing

outwards. This is one of the aspects that decreases fatigue life because it covers the crack initiation

37

Page 64: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

stage and even a little bit of the crack propagation period of the fatigue process.

Anodized Washed with Freshwater (AC)

For each exposure time considered, a microscopic image was selected representing the state of the

specimens, Figures 4.9, 4.10, 4.11, 4.12, 4.13 and 4.14.

Figure 4.9: A - 0 cycles Figure 4.10: AC - 1 cycles

Figure 4.11: AC - 3 cycles Figure 4.12: AC - 9 cycles

Figure 4.13: AC - 15 cycles Figure 4.14: AC - 20 cycles

*The red line represents 100 µm.

As it can be observed, figures 4.10 and 4.11 from cycle 1 and 3, respectively, are not very different

38

Page 65: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

from the uncorroded specimen (Figure 4.9). This means that, until then, the material was protected

effectively by the anodic coating.

When the samples reached cycle 9 (Figure 4.12) cracks on the oxide layer started to show. The

breakdown occurred in a very fine and uniform way. As the exposure time increased the cracks became

thicker and the color of the oxide layer started to fade (Figures 4.13 and 4.14).

Anodized Not Washed (AD)

For each exposure time considered, a microscopic image was selected representing the state of the

specimens, Figures 4.15, 4.16, 4.17, 4.18, 4.19 and 4.20.

These specimens had a very similar behavior to the Anodized washed with Freshwater (AC).

The anodization was almost unchanged until cycle 3 (Figure 4.15, 4.16 and 4.17). At cycle 9 (Figure

4.18) the crackle effect of the oxide layer started to show and, with increasing corrosion exposure,

revealed a growth in density of cracks (Figures 4.19 and 4.20). The specimens were also discolored

with increase in corrosion cycles.

Even though these specimens were not washed, salt depositions on the surface were not present.

Although some residues of salt were present in the specimens’ surface they were not comparable in

quantity and size with the salt crystals on the not anodized and not washed (NAD) samples.

Not Anodized Washed with Nitric Acid (NAB)

For each exposure time considered, a microscopic image was selected representing the state of the

specimens, Figures 4.21, 4.28, 4.29, 4.30, 4.31 and 4.32.

This group of specimens was the most damaged by corrosion from all the different tested conditions.

From the first cycle, noticeable changes were observed, at naked eye, on the entire surface of these

samples. All the specimens presented several cracks along the entire surface (Figure 4.28). These

cracks are the oxide layer (Section 2.3.1) breaking down due to corrosion.

As it has been seen before, on the anodized washed with nitric acid (NAB) subsection, the acid

removed the aluminum oxide. This corrosion product has a special feature that is protecting the surface

from further corrosion. When the nitric acid dissolves this film the surface gets more susceptible to

the corrosion phenomenon. On these specimens, the quantity of aluminum oxide is less than on the

anodized ones which means that the acid will dissolve it all faster.

After three cycles there were already several pits in random places that start small in size but big in

quantity. Around these pits, the oxide film appears more damaged (Figure 4.29). The pits start to merge

(Figure 4.30) after nine cycles forming bigger pits.

In Figure 4.31 we can observe three pits with approximate areas of 0.011 mm2, 0.055 mm2 and

0.233 mm2 from the smallest to the bigger pit, respectively.

On the last cycle, the pit in Figure 4.32 has an approximate area of 1.31mm2 being the scale different

due to the increse in pit size.

39

Page 66: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 4.15: A - 0 cycles Figure 4.16: AD - 1 cycles

Figure 4.17: AD - 3 cycles Figure 4.18: AD - 9 cycles

Figure 4.19: AD - 15 cycles Figure 4.20: AD - 20 cycles

*The red line represents 100 µm.

40

Page 67: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 4.21: NA - 0 cycles Figure 4.22: NAB - 1 cycles

Figure 4.23: NAB - 3 cycles Figure 4.24: NAB - 9 cycles

Figure 4.25: NAB - 15 cycles Figure 4.26: NAB - 20 cycles

*The red line represents 100 µm.

Not Anodized Washed with Freshwater (NAC)

For each exposure time considered, a microscopic image was selected representing the state of the

specimens, Figures 4.27, 4.22, 4.23, 4.24, 4.25 and 4.26.

Although these specimens did not appear as much corroded as the not anodized washed with nitric

acid (NAB), on the first cycle, visual signs of corrosion were still present. Small pits initiated all over the

41

Page 68: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 4.27: NA - 0 cycles Figure 4.28: NAC - 1 cycles

Figure 4.29: NAC - 3 cycles Figure 4.30: NAC - 9 cycles

Figure 4.31: NAC - 15 cycles Figure 4.32: NAC - 20 cycles

*The red line represents 100 µm.

exposed surface and some colors appeared (Figure 4.34).

After three cycles, the untreated specimens stayed colorful but cracks on the oxide film started to

come through in a fish scale pattern (Figure 4.35).

By the ninth cycle, cracks on the oxide film were still visible and the rainbow colors started to fade

(Figure 4.36).

Even though these specimens were washed with freshwater they were dirty, with significant salt

residues but not as much as the non-washed. This is represented by the darkness in some of the

42

Page 69: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

figures.

The density of cracks got bigger and pits appear more, in random places, with more area and depth

(Figure 4.37). The breaking of the oxide film provoked the localized corrosion. Some places had more

unprotected exposed surface than others which is why the pit sizes are different, in an initial phase.

The pits start to coalesce with neighboring pits and create bigger pits (preferably along the grain

direction) and the crackle effect of the oxide film becomes more and more evident (Figure 4.38).

In all specimens, corrosion was bigger near the edges of the exposed surface area.

Not Anodized Not Washed (NAD)

For each exposure time considered, a microscopic image was selected representing the state of the

specimens, Figures 4.33, 4.34, 4.35, 4.36, 4.37 and 4.38.

The behavior of these specimens was similar to the not anodized washed with freshwater (NAC).

The images of the non washed specimens appear a little darker than the others due to the salt

deposition on the surface which was very noticeable. The salt and other corrosion products that remained

on the surface appear black when seen with the microscope.

After the first cycle, signs of corrosion were already visible. Pits appear in irregular shapes and colors

come through on the uncorroded part of the exposed surface (Figure 4.34).

On the third cycle (Figure 4.35) there were not many changes when comparing to the first cycle

figure. Although, the pit density increased significantly.

By the ninth cycle, rainbow colors started to fade (Figure 4.36) and it can be seen already a considerably

large pit for this stage (approximate area of 0.343mm2).

The coalescence and growth of pits are evident in Figure 4.37 where it can also be seen that the

growth of the pits tends to be along the horizontal axis (which is the grain direction of the specimen).

With increase in exposure time, the pits continue to grow and new pits appear. The oxide film

continues to crack now in a more crackle effect type than fish scale pattern (Figure 4.38).

4.1.2 Corrosion Rate Determination

To get accurate weight losses the specimens were weighed between periods. Since the initial weight

was calculated taking into account the tape the only way of getting a comparison was to include the tape

in the final weighing. However, due to the water traps the tape got wet from within getting heavier and

invalidating the results. Some values were bigger than the ones measured for uncorroded specimens

which originated negative corrosion rates not physically possible.

For future work, it would be recommended a better way of weighing the specimens even if the final

weighing is made after the whole cycle with the unmasked and dry extracted specimen.

43

Page 70: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 4.33: NA - 0 cycles Figure 4.34: NAD - 1 cycles

Figure 4.35: NAD - 3 cycles Figure 4.36: NAD - 9 cycles

Figure 4.37: NAD - 15 cycles Figure 4.38: NAD - 20 cycles

*The red line represents 100 µm.

4.1.3 Maximum Valley Depth

The following graph represents the evolution of the maximum valley depth (Figure 4.39). It is a measure

of the depth of attack of corrosion along the cycles of exposure. The points are the mean values of the

maximum valley depth of the six tested replicates. The lines are the more approximate trending lines for

each case.

The different trending lines correspond to the anodized and bare specimens as well as the different

44

Page 71: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

washing methods they were subjected to according to the nomenclature in Tables A.1 and A.2.

Figure 4.39: Effect of corrosion on Maximum Valley Depth

As we can observe in Figure 4.39 the specimens with the most depth of attack were the ones washed

with nitric acid, either bare (NAB) or anodized (AB). The not anodized and not washed (NAD) and

washed with freshwater (NAC) were next and finally the anodized either not washed (AD) or washed

with freshwater (AC).

This result meets expectations.

The anodized specimens were less corroded and had the less maximum valley depth, which means

the anodization works.

The freshwater washing method reduced the corrosion on the bare and anodized specimens (NAC/AC)

when compared to the non-washed ones (NAD/AD). This difference was more accentuated in the not

anodized case.

The anodized specimens washed with nitric acid (AB) started similar to the other anodized samples,

however, since the nitric acid was constantly removing corrosion products, including the aluminum oxide

which protects the specimen, the evolution started to be more like the non-anodized that went through

the same washing method. In the end, it even got a maximum valley depth higher than the untreated

washed with nitric acid (NAB) (41.25 µm vs 32.81 µm).

4.1.4 Average Valley Depth

The following graph represents the evolution of the average valley depth (Figure 4.40). The trending

lines were created according to the mean values of average valley depth of the six tested replicates.

The different lines correspond to the anodized and bare specimens as well as the different washing

methods they were subjected to according to the nomenclature in Tables A.1 and A.2.

45

Page 72: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 4.40: Effect of corrosion on Average Valley Depth

The results for average valley depth are very similar to the maximum valley depth results, which is

accurate.

The anodized specimens washed with freshwater (AC) and not washed (AD) have the lowest average

valley density followed by the untreated washed with freshwater (NAC). The non-anodized and anodized

washed with nitric acid (NAB/AB) had a very close behavior and achieved the highest value of average

valley depth.

The most damaged specimens have the highest average valley depth (AB and NAB) due to the nitric

acid washing method that removed every corrosion product including the aluminum oxide. As for the

rest of the anodized specimens, the value is low due to the protection of the thicker passive layer.

The not anodized specimens present a large difference in evolution between the washed with freshwater

(NAC) and not washed (NAD) meaning that the washing with freshwater decreases the depth of attack,

of the bare specimens, originated by the salt spray fog.

4.1.5 Valley Density

The following graph represents the evolution of the valley density (Figure 4.41). The points are the mean

values of the valley density of the six tested replicates. The lines are the more approximate trending lines

for each case.

The different trending lines correspond to the anodized and bare specimens as well as the different

washing methods they were subjected to according to the nomenclature in Tables A.1 and A.2.

The variations of valley density were very small compared to the other surface roughness parameters.

Most of the lines are approximated by a polynomial and one by an exponential curve.

46

Page 73: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 4.41: Effect of corrosion on Valley Density

On the polynomial cases, there is a third order behavior. This kind of variation is because in the

beginning pits started to appear in a disperse way increasing the valley density. After this, they started

to get together making one larger pit which means that the density decreases.

In the end, both new pit appearances and pit merges started to happen at the same time varying

valley density up and down but not with a large amplitude.

The lower values of valley density are the bare and anodized washed with nitric acid (NAB/AB) that,

as seen in the previous subsections, had the highest value of maximum valley depth and average valley

depth. This means that pits were fewer but a lot bigger and deeper than in the other condition paths.

4.2 Tensile Test Results

The evaluated parameters from the tensile tests were the ultimate tensile strength, yield stress, Young’s

modulus, elongation at fracture and reduction area.

For each parameter different trend lines were deduced according to the several variations tested. A

graph was made comparing all the conditions for every tensile property. Every value obtained in this test

is included in Appendix A in Table A.6.

For reference values tensile tests were performed in anodized and bare uncorroded specimens, the

average results obtained for all the properties are presented in Table 4.1.

The main difference between anodized and bare specimens in terms of the mechanical properties

considered are the ductility measures (elongation at fracture and reduction area). Since the anodic film

is brittle the ductility is slightly less when comparing anodized with untreated specimens.

Figure 4.42 represents an overall view of the degradation of the mechanical properties of the tested

47

Page 74: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Table 4.1: Average Tensile Property Values of the Uncorroded Specimens

σUTS(MPa) σy(MPa) E(GPa) e(%) RA(%)Uncorroded & Anodized 587.04 519.33 70.02 5.44 16.78

Std Dev. 1.66 1.34 0.36 0.23 1.85Uncorroded & Not Anodized 586.54 522.44 69.62 6.00 18.17

Std Dev. 13.18 3.96 1.17 0.56 1.69

aluminum anodized and washed with nitric acid (AB). Each line corresponds to a representative sample

of each exposure time considered. All the curves are engineering stress-strain curves which means they

were constructed based entirely on the original dimensions of the specimens.

Almost all of the properties decreased with increasing corrosion exposure time as we can observe

in Figure 4.42, the lines get lower (lowering ultimate tensile strength and yield stress) and they also get

shorter (lowering the strain at fracture), the Young’s modulus is practically the same (the slope of the

elastic area is almost the same in all cases).

Figure 4.42: Engineering Stress-Strain Curve Evolution for Anodized Washed with Nitric Acid (AB)Specimens

4.2.1 Ultimate Tensile Strength

The Ultimate Tensile Strength (UTS) is the maximum amount of tensile stress that the specimen can

take before failure.

The evolution of this parameter is presented by average values in Figure 4.43. The different trending

lines correspond to the anodized and bare specimens as well as the different washing methods they

were subjected to according to the nomenclature in Tables A.1 and A.2.

All the mean ultimate tensile values used to achieve the trendlines presented in Figure 4.43 are

available in Table A.7. In this table, it can also be found a standard deviation for each case used to

48

Page 75: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

quantify the amount of variation or dispersion of the data values.

Figure 4.43: Effects of corrosion on Ultimate Tensile Strength of 7075-T651 Aluminum Alloy

As Figure 4.43 shows, in all cases, an overall decreasing of the ultimate tensile strength with increasing

exposure time was detected. This decrease is continuous and almost linear. It reaches values of

508.07MPa on the anodized specimens washed with nitric acid (AB) and 510.68MPa on bare specimens

washed with nitric acid (NAB). This corresponds to a 13.45% and 12.93% reduction from the initial value,

respectively. The specimens which presented a less degraded behavior were the anodized washed with

freshwater (AC) and the anodized not washed (AD), 4.90% and 1.99%, respectively.

The anodization protects the material from corrosion as it has been explained in Section 2.5.1 which

justifies why the degradation was less on these specimens besides the ones washed with nitric acid

(AB) due to the remotion of the anodic layer by the acid, reported on the previous subsection.

About the bare specimens, the ones washed with freshwater (NAC) ended stronger than the ones

not washed (NAD) and the washed with nitric acid (NAB) presented the most amount of degradation.

In this property, the freshwater washing method made a positive difference on the not anodized

specimens (NAC vs NAD). During the test campaign, it was verified that the anodized non-washed (AD)

finished a little bit stronger than the washed with freshwater (AC), however, the values were close.

4.2.2 Yield Stress

Yield stress is a material property that defines the transition value between the elastic and the plastic

region. If the stress value is greater than the yield stress value the material is in the plastic domain if it

is smaller the material is in the elastic domain. It represents the upper limit to forces that can be applied

without permanent deformation.

The evolution of this parameter is presented by average values in Figure 4.44. The different trending

lines correspond to the anodized and bare specimens as well as the different washing methods they

49

Page 76: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

were subjected to according to the nomenclature in Tables A.1 and A.2.

All the mean yield stress values used to achieve the trending lines presented in Figure 4.44 are

available in Table A.8. In this table, it can also be found a standard deviation for each case used to

quantify the amount of variation or dispersion of the data values.

Figure 4.44: Effects of corrosion on Yield Stress of 7075-T651 Aluminum Alloy

This property had a similar behavior to the ultimate tensile strength.

In all cases, there was a tendency for the yield stress to decrease with increasing cycles of exposure.

The minimum value achieved was 497.39MPa for anodized aluminum washed with nitric acid (AB) and

491.60MPa for not anodized washed with nitric acid. This corresponds to a loss of 4.22% and 5.90%,

respectively, in this property. The specimens with the lower yield stress values variation were the not

anodized washed with freshwater (NAC) with only 0.8% of loss.

The anodized specimens washed with freshwater (AC) and not washed (AD) had almost the same

behavior which means that the freshwater did not have a significant influence. As for the not anodize

samples the freshwater (NAC) case performed better than the others (NAB and NAD) which means that

it worked on this case scenario.

The main difference that it can be seen between Figure 4.43 and 4.44 is that the NAC line is higher

than the AC and AD lines. This means that washing the specimens with freshwater is better to prevent

yield stress loss than anodizing specimens. Another relevant observation is that the slopes are lower

than the ones of the ultimate tensile strength trending lines. The corrosion induced cracks made more

difference in the plastic region than the elastic. The probable reason is that the material’s plastic

deformation helps them propagate causing a higher degradation on ultimate tensile strength than on

yield stress.

50

Page 77: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

4.2.3 Young’s Modulus

The Young’s Modulus is a measure of the stiffness of a material in the linear elastic domain which means

it is the capability of the material to return to his original shape after applying the solicited loading.

The evolution of this parameter is presented by average values in Figure 4.45. The different trending

lines correspond to the anodized and bare specimens as well as the different washing methods they

were subjected to according to the nomenclature in Tables A.1 and A.2.

All the mean Young’s modulus values used to achieve the trending lines presented in Figure 4.45

are available in Table A.9. In this table, it can also be found a standard deviation for each case used to

quantify the amount of variation or dispersion of the data values.

Figure 4.45: Effects of corrosion on Young’s Modulus of of 7075-T651 Aluminum Alloy

This parameter has a maximum variation of 3.45% on the anodized washed with nitric acid (AB)

specimens and a minimum variation of 0.19% on the anodized not washed specimens (AD). All the

results have a similar average value which means that this parameter did not have a significant change

during the process. This happened because, unlike the failure behavior, the elastic behavior is not

related to maximum pitting but to the sum of the elastic strain all along the specimen.

The corrosion did not reach a point where it affected the elastic strain of the material as we can

observe in Figure 4.42.

Further tests with more corrosion cycles can be done to study if it will eventually change the Young’s

modulus or not.

51

Page 78: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

4.2.4 Elongation at Fracture

Elongation at fracture expresses the capability of a material to resist changes of shape without crack

formation.

The evolution of this parameter is presented by average values in Figure 4.46. The different trending

lines correspond to the anodized and bare specimens as well as the different washing methods they

were subjected to according to the nomenclature in Tables A.1 and A.2.

All the mean values of elongation used to achieve the trending lines presented in Figure 4.46 are

available in Table A.10. In this table, it can also be found a standard deviation for each case used to

quantify the amount of variation or dispersion of the data values.

Figure 4.46: Effects of corrosion on Elongation at Fracture of 7075-T651 Aluminum Alloy

The first point reveals an elongation at fracture higher on the not anodized specimens when compared

to the anodized. This is due to the brittle nature of the oxide layer that is present on the treated

specimens.

The anodized washed with freshwater (AC) and not washed (AD) had the least variation of this

parameter, 22.25% and 3.37% respectively, which is accurate because they are the least damaged

specimens.

The rest of the samples followed a decreasing exponential curve. The bigger variations correspond

to the specimens washed with nitric acid with 80.33% in the anodized case (AB) and 81.33% in the not

anodized case (NAB).

In this property, the non-washed specimens had lower losses than the washed with freshwater. On

the not anodized case, the lines are almost coincident (NAC and NAD) which means that the freshwater

did not make a difference. As for the anodized samples, the difference is very noticeable (AD and AC).

The order of the lines is in accordance with the other properties. In the end, the more damaged

specimens will present bigger losses.

52

Page 79: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

The loss in ductility is the most noticeable loss of all the tensile properties evaluated in this thesis.

4.2.5 Reduction of Area

Another ductility measure studied was the reduction area of the tested specimens. The more percentage

of reduction of area a specimen has the more ductile the material is.

The evolution of this parameter is presented by average values in Figure 4.47. The different trending

lines correspond to the anodized and bare specimens as well as the different washing methods they

were subjected to according to the nomenclature in Tables A.1 and A.2.

All the mean ultimate tensile values used to achieve the trendlines presented in Figure 4.47 are

available in Table A.11. In this table, it can also be found a standard deviation for each case used to

quantify the amount of variation or dispersion of the data values.

Figure 4.47: Effects of corrosion on Reduction Area of 7075-T651 Aluminum Alloy

In accordance with the previous section, in the beginning, the anodized specimens had a lower

reduction of area than the bare ones due to the brittleness of the anodic layer.

The specimens washed with nitric acid (NAB/AB) had a similar behavior, like in every tested property,

as did the not anodized washed with freshwater (NAC) and not washed (NAD).

The specimens with the least variation were the anodized not washed (AD), with 28.37%, being the

ones that represented the most ductile behavior (higher number of reduction area). This is coherent with

the elongation at fracture results.

Besides the anodized not washed (AD) all of the specimens had an almost exponential decrease,

being the not anodized and not washed (NAD) the curve that ended the most brittle with a reduction

area variation of 93.29%.

Even though this results should have been similar to the elongation at fracture results, since they

were both ductility measures, they were not. The AC and AD lines are the more coherent but the

53

Page 80: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

NAB and AB switched places with the NAD and NAC. This parameter has more experimental errors

associated.The initial and final area were both measured by hand with a micrometer and a caliper while

the elongation at fracture was retrieved from the engineering stress-strain curve obtained by the tensile

testing machine, which is more reliable. This is also perceptible by the standard deviation values when

comparing Table A.11 with A.10, the values are higher on the reduction area results.

4.3 Fatigue Test Results

The main objective of the fatigue test campaign was the analysis and evaluation of the maximum number

of fatigue cycles the specimens would withstand. This will quantify the influence of corrosion in the

material’s fatigue life.

The evaluated parameter was the number of cycles until failure. A limit of one million cycles was set

as the test program run out which means that the specimen either reaches one million cycles and the

test stops or it breaks before the cycle limit.

Different trend lines were deduced according to the several processes tested. A graph was made

comparing all the conditions.

For reference values fatigue tests were performed in anodized and bare uncorroded specimens, the

average results obtained for the properties are presented on Table 4.2.

Table 4.2: Average Fatigue Life of the Uncorroded Specimens

N (cycles) Std Dev.Uncorroded & Anodized 69030 15037

Uncorroded & Not Anodized 782938 375962

There was a 91.1% decrease in fatigue life just by anodizing the specimen before any corrosion

occurred.

It is generally accepted that reduction in fatigue performance of anodic oxide coated specimen is

directly related to the brittle and porous nature of the coating layer and tensile residual stress induced

during coating process [15]. The thicker the coating the more irregularities it contains which also

contributes to the decrease in fatigue life. Increasing the thickness also causes a larger area for micro

crack growth and coalescence.

For thicker oxide layers (for example hard anodization) the crack formation can also be attributed to

the difference in thermal expansion coefficients between the aluminum substrate and the coating layer

along with internal tensile residual stresses in the coating [15].

To illustrate the effects of corrosion on the reduction of fatigue life, a plot was made comparing failure

cycles with cycles of exposure for each variation considered on the experiment. The results are in Figure

4.48.

The mean fatigue values used to achieve the trending lines presented in Figure 4.48 are available

in Table A.13. In this table, it can also be found a standard deviation for each case used to quantify

the amount of variation or dispersion of the data values. In Appendix A Table A.12 there is all the data

54

Page 81: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

retrieved from the fatigue experiment.

Figure 4.48: Effects of corrosion on Fatigue Life of 7075-T651 Aluminum Alloy

The not anodized specimens had a significant loss in fatigue life with just one cycle of exposure

which means that the smallest amount of corrosion reduces a lot fatigue life. After the first cycle, the

differences were not so accentuated but it still decreased.

As for the anodized samples at first, since the whole corrosion process dissolves the anodic layer,

there is almost a maintenance of fatigue life or even a slight increase on the first cycle. After that between

the continuous dissolution of the film and simultaneous corrosion, fatigue life decreases. This is more

noticeable on the specimens washed with nitric acid (AB) (due to the almost complete dissolution of the

anodic layer) but it also occurs on the other anodized samples.

In all cases, the effect of corrosion appears to follow a decreasing exponential trend. The not

anodized cases show the largest decrease in fatigue life with a 96.63% for the washed with nitric acid

specimens (NAB), 94.67% for the freshwater washing method (NAC) and 97.33% for the not washed

samples (NAD).

The anodized samples presented a decrease of 63.42% for the washed with nitric acid case (AB),

41.16% for the freshwater washing method (AC) and 38.89% for the not washed (AD).

To analyze in more detail the graph below 300000 cycles to failure the plot in Figure 4.49 was made

representing a zoomed graph of the one in Figure 4.48.

55

Page 82: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 4.49: Effects of corrosion on Fatigue Life of 7075-T651 Aluminum Alloy

It can be observed that even though the not anodized started with higher fatigue lifetimes after twenty

cycles they ended up with less fatigue life than the anodized samples. This is because the anodization

prevented the pit formation where the fatigue cracks tend to initiate from. This means that pitting reduces

fatigue more than anodization on its own.

Pits are considered stress concentrations from where the failures occur. Fatigue cracks can nucleate

from these corrosion pits and grow at an accelerated rate in a corrosive environment.

It is also suspected that hydrogen embrittlement and acidification inside the pit makes the area at the

bottom (Section 2.4.3) of the pit brittle and more susceptible to cracking [11].

The reduction in fatigue life present in every test specimen is primary a result of premature crack

initiation caused by pitting. The pits create an accumulation of local irreversible plastic deformation and

microscopic flaws that grow and coalesce with other flaws from other microscopic cracks.

Fatigue failures can have one or more fatigue crack origins, a region of progressive fatigue crack

propagation, and a final fast overload fracture zone. (Figure 4.50).

Figure 4.50: Specimen’s Cross Section After Fatigue Test [Courtesy of CEiiA]

56

Page 83: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 4.50 shows a cross section of a selected specimen after a fatigue test. This specimen had

several initiation points. As the fatigue test runs cracks started to appear, preferably where pits were,

on both of the surfaces. Since the loading is cyclic striations or benchmarks are created. Each striation

corresponds to one loading cycle which is why fatigue crack growth rate can be estimated from the

spacing between benchmarks.

The circle in Figure 4.50 highlights two pits that got deeper along the cyclic loading. Typical fatigue

benchmarks are also visible on the lighter and softer areas and then the catastrophic failure is represented

by the rougher and darker area.

4.3.1 The S-N Curve

The S-N curve is the basic method of presenting engineering fatigue data. It is a plot of stress, S against

the number of cycles to failure, N. In this case the stress value is the maximum stress. As it has been

said in Section 2.9 this curve is used for HCF, which is the case.

These fatigue tests were carried out at the Instituto Superior Tecnico Mechanical Testing Laboratory.

The tested specimens were all not anodized and washed with freshwater between periods. The conditions

of this test were the same as the other fatigue tests performed in this master thesis (SR=0.1; f=10Hz)

besides the maximum tension which varied in order to have points for the curve.

The results of the fifteen fatigue tests are presented in Figure 4.51, with the respective logarithmic

trending lines, and the values used to achieve the graph are available in Table A.14.

Figure 4.51: Effect of Corrosion on the S-N Curve of 7075-T651 Aluminum Alloy

In Figure 4.51 there are three S-N curves for a different number of corrosion exposure cycles and

the first observation made is that corrosion has a tendency to decrease fatigue performance.

Comparing the blue line (uncorroded specimens) with the red line (specimens that were exposed

to three cycles of corrosion), it can be seen that for the same approximate interval of fatigue life the

maximum stress of the cyclic load applied on the material is considerably less. The maximum stress

57

Page 84: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

for the approximate same amount of cycles to failure decreased around 100MPa. This corroborates

the results presented in the previous subsection: a small amount of corrosion affects fatigue life very

significantly.

As for the green line (specimens that were exposed to fifteen corrosion cycles), when compared to

the red line, there was a decrease in fatigue cycles for the approximate same amount of maximum stress

applied. As we have seen before, as corrosion exposure increases fatigue life decreases exponentially.

This means that for small amounts of corrosion big variations will be observed but as the corrosion

exposure continues the variations start to decrease.

For an increase in corrosion time it is expected that the curve will move downwards and leftwards.

All the curves did not present an endurance limit (part where the curve gets horizontal meaning that

the material has an infinite fatigue life, Section 2.9). This is accurate since it has been proved before that

aluminum does not have this limit [42]. However, in the future, more specimens can be tested to obtain

a higher number of stress values in order to improve accuracy in the curves.

58

Page 85: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Chapter 5

Finite Element Analysis

In this section, a brief introduction to numerical approaches will be performed.

Information on the used software, the meshing decisions, the load and boundary conditions of the

problem, as well as on the material properties and analysis type will be provided.

Furthermore, the validation of the model and the creation of a program with practical applications will

be also explained in detail.

5.1 Introduction

There are three methods to solve any engineering problem, [45]:

• Analytical Method: which is the classical approach and gives, in general, accurate results.

However, it is only applicable for simple problems (e.g. simply supported beams);

• Numerical Method: which is a mathematical representation, with approximations and assumptions

made previously, that is used for solving real life complex problems (e.g. Finite Element Method

(FEM)). It is applicable even if the prototype is not available (initial design phase). However, the

results cannot be believed blindly, some of them must be validated by analytical or experimental

methods;

• Experimental Method: which is an actual measurement, it is time consuming and needs an

expensive set up, it needs a minimum of 3 to 5 prototypes to be considered accurate and it is not

applicable if the prototype is not available (e.g. Fatigue Test).

This chapter is focused on a numerical approach of the problem using the Finite Element Method.

This method reduces the degrees of freedom from infinite to finite with the help of discretization or

meshing (nodes or elements). All the calculations are made at a limited number of points (nodes) and

the entity joining nodes and forming a specific shape is known as an element. To get a value between

the calculation points an interpolation function is used [45].

The major advantages of the numerical method, when compared to the experimental method, are

the time and money costs. Hence, by way of comparison with the experimental results, a tensile test

59

Page 86: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

with a finite element model was simulated on a test specimen.

After achieving similar results in both methods a practical application was created that characterizes

the material based on the degradation equations found experimentally. This program uses the experimental

results to characterize the corroded material and generates the necessary material properties in the FEM

solver standard language. These properties can then be used for any simulation in any structure made

out of 7075-T651 aluminum alloy, within the test domain.

5.2 Software

The standard FEM process is divided into three steps: modeling/pre-processing; solution/analysis and

visualization/post-processing [45]. The software used for pre-processing was Altair HyperMesh, the

solver was MSC Nastran and for post-processing was Altair HyperView. The 3D design of the specimen

was made in CATIA by Dassault Systemes R©. The developed program uses a combination of MATLABR© and MSC Nastran.

5.3 Meshing

The basic idea of FEM is to only make calculations at a finite number of points called nodes. These

nodes, when connected, form entities in predetermined shapes called elements and then the group of

all elements is called mesh.

The meshing step is crucial to the finite element analysis as the quality of the mesh directly reflects

on the quality of the results generated.

Firstly, an element type needs to be defined. This decision is based on geometry size and shape,

type of analysis and time allotted for the project [45]. Since the thickness of the specimen is low when

compared to the width and length, 2D elements were chosen for this analysis [45]. 2D elements are

planar, just like paper, and the software knows two out of the three required dimensions. The third

dimension (thickness) is given by the user as input data.

The 2D meshing is carried out on a mid surface of the geometry due to the fact that the thickness

given by the user is assigned half on the element top and half on the element bottom side. Hence, to

represent the geometry as accurately as possible it is necessary to extract the mid surface and then

mesh on it.

Considering that the specimen’s geometry is a thin 3D structure, shell elements were chosen (PSHELL)

with a quadrilateral plate element type (CQUAD4). The Nastran’s PSHELL bulk entry is a shell element

property that defines the membrane, bending, transverse, shear, and coupling properties of thin shell

elements. The quadrilateral element CQUAD (Figure 5.1) was chosen because the can warp better

than triangular elements, which are stiffer. This element type has 4 nodes, each node with 6 degrees of

freedom 3 for translation and 3 for rotation.

60

Page 87: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 5.1: CQUAD4 Element Geometry and Coordinate Systems [Source:[46]]

To define the dimensions of the mesh, a convergence study was performed. With increasing number

of elements the values for maximum stress and strain were evaluated (Figure 5.2).

Figure 5.2: Mesh Convergence Study

Five multiples of the part’s thickness for the mesh dimension were considered (7t,4t,3t,2t,t), all

adapted, as well as it could be adapted, to the geometry in question.

In Figure 5.2 it can be seen that there is a convergence from the first to the fourth point. However,

the last point diverges because elements with dimension t × t × t no longer respect shell theory (there

is not a big difference between width and length in relation to thickness) which is why it is not a valid

mesh for this type of elements. There is no point in creating meshes bigger than 7t due to the fact that

loses the geometry accuracy and the differences, in the critical regions, in tension and extension are not

present.

The mesh was carried on so that the elements were as square as possible, and the chosen length

was 3t (in this case, t = 3mm), although, especially on the zone where the specimen experiences a

reduction area the dimensions were adapted to best suit the geometry and refined results (Figure 5.3).

61

Page 88: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 5.3: Meshed Specimen

The blue lines on the left side of Figure 5.3 represent an RBE2. An RBE2 is a 1D element used to

connect nodes. This connection is a rigid link that transfers motion from the independent node to the

dependent nodes [46]. In this case, a node in the center of the left grip of the specimen was created (the

independent node) and it is connected to all the nodes present in the grip rectangular area (dependent

nodes).

Since the final goal of this simulation is to recreate a tensile test, this element was added for the load

to be applied at just one point and be transmitted to the entire grip section equally. Applying the tensile

load in the independent node the RBE2 it will replicate accurately the tensile test.

The mesh has a total of 84 CQUAD4 elements and 1 RBE2 element. This gives a total of 117 nodes.

5.4 Load and Boundary Conditions

The tensile test, as explained in Section 3.3, immobilizes one grip of the specimen and pulls the other

with a predetermined load.

The representation of this in FEM is the creation of Single Point Constraint (SPC) in one grip and the

application of a force on the other (Figure 5.4).

Figure 5.4: Load and Boundary Conditions Applied on the Specimen

In Figure 5.4 the red arrow on the left represents the force applied and the gray triangles on the right

represent the SPC. The application of the SPC in the grip nodes constraints all the degrees of freedom

of the nodes fixing the model on those points.

62

Page 89: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

5.5 Analysis Type and Material Properties

5.5.1 Linear Static Approach

On a first approach, a linear static analysis was made in order to validate the model and the decisions

made while developing the mesh. The material properties defined represent the aluminum alloy 7075-T651

uncorroded and not anodized.

Linear static analysis means that the finite element solver will always follow a straight line, from base

to deformed state, and the force is static (does not vary with respect to time) [45]. This will only give

accurate results inside the material’s elastic domain (the linear part of the stress-strain curve).

The basic finite element equation to be solved for structures experiencing static loads can be expressed

as, [45]:

[K]{u} = [P ] (5.1)

where K is the stiffness matrix of the structure (an assemblage of individual element stiffness matrices);

the vector u is the displacement vector, and P is the vector of loads applied to the structure. Equation

5.1 represents the equilibrium of external (right side) and internal forces (left side).

Once the unknown displacements at the nodal points of the elements are calculated, the strains

can be achieved(εi = du

di

∣∣∣i=x,y,z

). With known strains, the stresses can be calculated by using the

constitutive relations for the material. When the deformations are in the elastic range the stresses are a

linear function of the strains (σ = Cε, where C is the elasticity matrix of the material).

Linear Model Validation

The point A (138.011;0.002012) from the linear part of the experimental (0-NA) stress-strain curve was

chosen. From the analytical formula for stress σ = PA , considering the area of the cross section constant

and equal to 37.5mm2 (12.5× 3mm2), the value of the applied force can be calculated (P = 5175.41N ).

This force value was inserted and a linear static analysis was simulated.

The results showed that maximum stress and strain occurred in the center of the gauge area (where

there is an induced reduction area so that the specimen will break in there), as predicted. The final

values are presented in Table 5.1 as well as the relative deviation between the two measures.

Table 5.1: Stress and Strain Value Comparison for Linear Static Analysis

Stress [MPa] Strain [mm/mm]Experimental 138.011 0.002012

Numerical 138.016 0.002012Relative Deviation 0.004% 0.00%

This result validates the numerical simulation.

63

Page 90: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

5.5.2 Nonlinear Static Approach

According to the fact that corrosion changed mostly the plastic part of the stress-strain curve, a material

nonlinearity must be considered to include in FEM the material’s plastic properties. Therefore, a non-linear

approach was also investigated.

The main differences between the linear static approach and the nonlinear static are [45]:

• Linear approach uses load and stiffness matrices to obtain displacement in a linear way ([K]{u} = [P ]).

In the nonlinear case, even though stiffness varies as a function of load it is not linearly;

• When a material is used beyond the elastic limit the stress-strain curve is nonlinear. The final state

after removing loads is different from the initial state if a tension greater that yield stress is applied

(which is the definition of plastic deformation);

• The solution scheme also changes, the load is split into small increments with iterations performed

to ensure that equilibrium is satisfied at every load increment. The computational time is bigger

and the software requires a lot of monitoring as it might fail to converge sometimes.

The major feature of the nonlinear analysis is the requirement for the incremental and iterative

processes to obtain a solution. The load is the variable that is incremented and the iterative process

iterates the variation of displacement. The incremental and iterative processes are complementary to

each other because the larger the increment size the more iterations the solution requires [47].

The increment size for load steps has the most significant effect on the efficiency and the accuracy

of the computation. While an excessively small increment reduces the computing efficiency without

any significant improvement in accuracy, a large increment may deteriorate the efficiency as well as the

accuracy; it may even cause divergence [47].

It is impossible to optimize the incremental step size in the absence of prior knowledge of the

structural response. The best engineering judgment should be exercised to determine the increment

size based on the severity of the nonlinearity,[47].

The iterative scheme used is the Newton-Raphson method with the error vector ({R}) being

{R} = {P} − {F} (5.2)

where P is the vector of the externally applied loads and F the element nodal forces. Based on

Newton’s method, a linearized system of equations is solved for incremental displacements by Gaussian

elimination (which is an algorithm for solving systems of linear equations) in succession. The Jacobian

of the error vector emerges as the tangential stiffness matrix [KT ] as follows, [47]:

[KT ]{∆ui} = {P} − {F (ui−1)} (5.3)

{ui} = {ui−1}+ {∆ui} (5.4)

Given the initial conditions of the displacement ({u0}) and externally applied forces ({P0}) the stiffness

matrix and the vector F (u) can be obtained ([KT ] and {F0}). Then, through the equilibrium function

64

Page 91: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

(Equation 5.3), a new {∆u} is achieved. With this new parameter, the new displacement ({ui}) is

calculated and convergence is evaluated. If the solution respects the convergence tolerances the

analysis proceeds with a load increment, if not, the iteration continues with the same process . The

number of load increments (NINC) is predefined by the user and the analysis stops when this value is

achieved (Figure 5.5).

Figure 5.5: Solution Scheme of a Nonlinear Problem in Nastran

After every iteration equilibrium conditions (external forces equal internal nodal element forces),

compatibility conditions (satisfy displacement boundary conditions) and the material’s stress-strain law

are checked and errors are calculated. From iteration to iteration these errors are what define the

different convergence criteria. In the pre-processing tool the values for this tolerances are predetermined,

based on tolerance levels the user decides to give, and also the combination of errors the user wants to

use (e.g. displacement error, load equilibrium error, work error, ...).

The stiffness matrix update operation is one of the most CPU consuming process in nonlinear

analysis [47]. From an efficiency point of view, the number of stiffness matrix updates and the number of

iterations should be minimized. Variations of Newton’s method are adopted in MSC Nastran to update

the matrix at every few iterations or even at every load increment. The user can choose which stiffness

updating algorithm is adequate to the problem in question.

The software also uses convergence acceleration techniques like the quasi-newton method, the line

search method, and the bisection method. These techniques are chosen, based on an examination of

the convergence rate, by the software. However, the user can establish limits and tolerances for all the

65

Page 92: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

methods.

By calculating the ratio of energy errors before and after the iteration a divergence rate (di) is defined

(Equation 5.5).

di ={∆ui}T {Ri}{∆ui}T {Ri−1}

(5.5)

Based on this rate, the number of diverging iterations (NDIV) is incremented in one unit (probable

divergence) or two units (absolute divergence).

• If di ≥ 1 or di ≤ −1012, then NDIV = NDIV + 2 - absolute divergence;

• If − 1012 < di < −1, then NDIV = NDIV + 1 - probable divergence.

When this number (NDIV) is higher than the maximum probable divergence conditions per iteration

(MAXDIV, defined by the user) the solution is assumed to diverge. If this occurs, the program does an

automatic bisection of the load increment. Anyhow, the number of bisections has also a limit, inserted

by the user, and if this limit is reached the stiffness matrix is updated and the analysis continues. If

probable divergence is further detected after the new [KT ] is formed the best solution is computed and

the analysis continues to the next increment. If an absolute divergence is recognized in two successive

iterations, in spite of the new stiffness, the computation will be terminated.

Nonlinear Model Validation

To validate the nonlinear approach two points, one in the elastic region and another in the plastic region,

of the experimental stress-strain curve of an uncorroded non anodized specimen (0-NA) were evaluated.

For the elastic region, the same point A (138.011;0.002012) from the linear static analysis was

considered. Even though the analysis is nonlinear it also has to work for the linear segment.

In order to run a nonlinear static analysis, some changes need to de made. The mesh, the applied

load, the applied constraints and the element property (PSHELL) are the same. However, in the material

property definitions the bulk entries MATS1 and TABLES1 were added, to characterize the plasticity of

the material. The nonlinear parameters (NLPARM) were also created and set to default.

The results of the simulation are given in Table 5.2.

Table 5.2: Stress and Strain Value Comparison of a Linear Point in Nonlinear Static Analysis

Stress [MPa] Strain [mm/mm]Experimental 138.011 0.002012

Numerical 138.016 0.002000Relative Deviation 0.004% 0.06%

These results validate the nonlinear approach for the linear region. Nonetheless, a nonlinear point

must also be tested.

From the same experimental curve, the plastic point B (541.019;0.02138) was chosen. The load that

corresponds to this stress is P = 20288.2N . The final results of the simulation are available in Table 5.3.

66

Page 93: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Table 5.3: Stress and Strain Value Comparison of a Nonlinear Point in Nonlinear Static Analysis

Stress [MPa] Strain [mm/mm]Experimental 541.019 0.02138

Numerical 540.995 0.02131Relative Deviation 0.004% 0.33%

The results from Table 5.2 and 5.3 validate the nonlinear numerical simulation.

5.6 Practical Application of Experimental Results

As stated previously, a part of the objectives of this research was to quantify the effects of corrosion on

the ultimate tensile strength and yield stress of 7075-T651 aluminum. This was investigated experimentally

and both properties had a general decrease with increasing exposure time.

When comparing the experimental tensile test curves along the exposure cycles (Figure 4.42) it can

be observed that the slope until yield stress (the Young’s modulus) is almost unchanged. The differences

between the curves rely on the plastic part of the graph. Because of this, the curve was divided into

two lines: one line from the origin to the yield stress (elastic line) and another from the yield stress until

fracture (plastic line). This plasticity model is the simplest one MSC Nastran uses (Figure 5.6).

Figure 5.6: Stress-Strain Curve Definition [Adapted from: [46]]

The slope of the uniaxial stress-strain curve in the plastic region is known as tangential modulus (ET )

and a linear approximation of this line can be made in the format y = mx+ b.

With the values present in the experimental tensile curves, a linear approximation of the plastic line

was made for the different curves. An evolution of the slope (m), Figure 5.7, and the y-intercept (b),

Figure 5.8, with increasing corrosion exposure cycle was built.

67

Page 94: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Figure 5.7: Effects of Corrosion on the Slope of the Plastic Line

Figure 5.8: Effects of Corrosion on the y-intercept of the Plastic Line

The effect of corrosion had an exponential trend on the slope (m) and a linearly decreasing trend

on the y-intercept (b) for all the different combinations of bare and anodized aluminum with the three

washing methods. The more corroded cases (AB and NAB) had the higher increase of m and the higher

decrease of b.

Having the knowledge of how the yield stress, ultimate stress and slope and y-intercept of the plastic

line degrade, the material properties for any corrosion exposure time can be achieved.

Note: The equations were made with a knowledge of 20 corrosion cycles. For a higher number of

corrosion exposure time, the results are extrapolations of experimental data and can, sometimes, being

physically impossible. This happens mostly due to the fact that yield stress has a lower negative slope

when compared to ultimate tensile stress, and from a certain number of corrosion cycles this number

gets higher than the UTS value.

68

Page 95: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Depending on the specimens condition case scenario the evolution of the parameters is different.

Hence, knowing the amount of corrosion cycles and the conditions the material nonlinearity can be fully

plotted according to the curve definition stated earlier (Figure 5.6).

For MSC Nastran to solve any finite element analysis a .bdf input file with all the details of the

simulation (element type, coordinates of each node, material properties, applied loads, applied constraints,

...) is created by the pre-processing program. This input file is written in the Nastran ”language” so that

the solver understands every detail of the simulation in question and it must also include the desired

outputs (e.g. stress values, strain values, displacement values, ...). After running the simulation, the

solver creates a .f06 output file in which all the previously asked results are presented.

Since MSC Nastran is the solver used by CEiiA’s engineers for finite element analysis, a program was

developed so that they can have the properties of the material for every corrosion time exposure directly

in the Nastran ”language”. In this way, for any structural part made out of aluminum alloy 7075-T651 the

effect of corrosion can be taken into account when doing the part’s structural analysis.

5.6.1 Program’s Inputs and Outputs

Using MATLAB R© and MSC Nastran an executable was made.

Program Inputs (Figure 5.9):

• The number of corrosion cycles the material has been exposed to;

• The adequate condition case (depending if the material was anodized or not and the washing

method it was subjected to, according to the nomenclature in Tables A.1 and A.2).

Figure 5.9: MATLAB R© Input Window - Example for 15 Corrosion Cycles and AB Case Specimen

69

Page 96: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Program Outputs:

• A file in the .txt format which includes (example file in Appendix C):

– The ultimate tensile value correspondent to the corroded material;

– The yield stress value correspondent to the corroded material;

– The code that characterizes the new material in the Nastran ”language” ready to be inserted

in the .bdf file;

– The instructions to change the Nastran input file from a linear static solution to a nonlinear

static solution that includes the code from the previous point.

• A graph with three stress-strain curves (example in Figure 5.10):

1. The uncorroded experimental curve, for a reference measure;

2. The plasticity model curve, based on the UTS, yield stress, m and b calculated for the

degraded material;

3. The curve generated after the FEM simulations.

Figure 5.10: MATLAB R© Graph Window - Example for 15 Corrosion Cycles and AB Case Specimen

70

Page 97: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

5.6.2 Program Scheme

The scheme of this program is presented in Figure 5.11.

Figure 5.11: Program Scheme

5.6.3 Code Development

MATLAB R© code number one (Figure 5.11) reads the desired condition code and the number of corrosion

exposure cycles and, with the formulas deduced from the experimental curves, obtains the values of

yield stress, ultimate tensile strength and m and b of the material’s plastic line. Inside this part of the

executable, there is a pre-made Nastran card of the specimen model during a tensile test with the

nonlinear analysis validated in Section 5.5.2. This .bdf file already contains all the mesh characteristics,

material properties, forces, and constraints applied and default nonlinear parameters so that a nonlinear

static analysis (SOL 106) can be performed.

The Nastran bulk data entries that define the material nonlinear properties and need to be changed

when the material is exposed to corrosion are :

• MAT1 - defines the material properties for linear isotropic materials and the parameters inserted

are the Young’s modulus and the Poisson’s ratio, this is the only material parameter needed for

linear analysis;

• MATS1 - specifies stress-dependent material properties for use in applications involving nonlinear

materials and it is used as a compliment of MAT1. The parameters inserted in MATS1 are the

plastic line definition of the stress-strain curve through a table entry (TABLES1), the type of material

nonlinearity (in this case ”PLASTIC”) and the yield stress;

• TABLES1 - it consists of 7 points (by option) with coordinates from the stress-strain curve. The

first point must be the origin, the second point the yield point. The next five stresses are obtained

by dividing the stresses between ultimate and yield into five equal parts. The value for the yield

71

Page 98: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

strain is calculated by the linear relationship σ = Eε; the strain that corresponds to the UTS value is

calculated with the σ = mε+b obtained from the inputs; the other values of strain are calculated by

an interpolation between this two points with a correction for m and b so that the curve is consistent

with the approximation in Figure 5.6.

Note: The bulk entries have more fields than the ones mentioned. Although, not all fields have to be

filled since Nastran has some default values and some fields are only necessary in specific cases. When

the entries are to be used please check Nastran Quick Reference Guide [46] for detailed information.

With the UTS and area values, the maximum force is calculated which is further divided into 20

force values equally spaced, so that a stress-strain curve can be constructed. After this, MATLAB R©

generates 20 .bdf files with the updated material properties (Nastran bulk data entries: MAT1, MATS1,

and TABLES1) and writes, in each one, a different force value (Nastran bulk data entry: FORCE).

From this first code are also created two .txt files. The first file (Output.txt) contains the UTS and

yield stresses of the corroded material as well as the Nastran code lines that need to be changed

concerning the material properties (MAT1, MATS1 and TABLES1). The second file (Data.txt) contains

the data from the uncorroded stress-strain curve (contained in a vector in code one depending if the user

chooses an anodized or a bare material) and the interpolated curve constructed based on the plasticity

theory of Figure 5.6.

The next step is to run the 20 .bdf files in Nastran which generates 20 .f06 files. Since it is only

needed one value for stress and strain, one element from the mesh had to be chosen. In tensile tests,

the ”dogbone” specimen shape is the most typical shape so that the deformation is confined to the

narrow center region to reduce the likelihood of fracture to occur at the ends of the specimen. Because

of this, in the input file, the requested outputs were only the stress and the strain present in the element

in the middle of the specimen, being this region the most probable for fracture to occur.

The second MATLAB R© code reads the results of the finite element analysis as well as the data from

the Data.txt file. With this values, it plots a graph comparing three stress-strain curves: the uncorroded

experimental, the interpolation based on the plasticity approach considered and the FEM result.

72

Page 99: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Chapter 6

Conclusions

The following conclusions can be drawn:

• Corrosion exposure to salt fog leads to extensive pitting of the not anodized material since the first

cycle, as expected;

• Increased corrosion was present on the edges of the exposed surface;

• Anodization protected the metal effectively from corrosion;

• Maximum and average valley depth both increased with exposure time. Valley density had a lot of

variations but with small amplitude;

• Reduced cross-section and stress concentrations caused by corrosion greatly influenced the load

carrying capacity of the specimen reducing, in general, ultimate and yield stress at a constant rate;

• Corrosion-induced degradation of mechanical properties occurs gradually for every tensile condition

tested. Young’s modulus did not have a significant variation. Elongation at fracture and reduction

area decreased exponentially to extremely low final values;

• From the tensile tests, the ductility measures were the more affected properties with higher degradation

percentage. Corrosion increased the material’s brittleness which in turn can change the failure

mode to a much more dangerous brittle failure;

• Corrosion had a greater degradation effect on the ultimate tensile strength than on the yield stress

due to the fact that corrosion affected mostly the plastic domain of the material;

• Based on experimental observations, fatigue cracks initiate from corrosion pits, as predicted;

• Fatigue performance of 7075-T651 alloy was significantly reduced by the anodic oxidation process.

The degrading effect was about 91.1% reduction. This reduction can be primarily ascribed to deep

micro cracks formed during the anodization process;

• Fatigue life appears to follow an exponential reduction with increasing exposure time;

73

Page 100: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

• Small amounts of corrosion decrease the not anodized specimens fatigue life significantly;

• The fatigue S-N curve moved down and left with increase in corrosion cycles. This means that

corrosion reduces the number of cycles to failure for the same maximum load applied;

• In a long term perspective fatigue life is more reduced by corrosion on an untreated specimen than

on an anodized, which makes the treatment an advantage for long term applications;

• The freshwater washing method made a more positive difference, in general, on the not anodized

than on the anodized samples;

• Specimens tend to fail on the more corroded areas either in tensile or fatigue tests as anticipated;

• In a general balance, the specimens that presented the best behavior were the anodized washed

with freshwater and not washed. If they resisted the salt spray exposure they will definitely

resist seawater and sea atmosphere since the test is more aggressive than reality (conservative

approach);

• This results correspond to the aluminum alloy 7075-T651 and cannot be extrapolated to other alloy

compositions nor other temper designations.

6.1 Achievements

With the work done in this MSc Thesis, the quantification of corrosion, under a wet/dry cycle, of the

aluminum alloy 7075-T651 was accomplished. This quantification was made based on loss of mechanical

properties and fatigue life. A practical application was also developed where the user inserts the

amount of corrosion and the condition case of the 7075-T651 alloy and the program returns the material

properties taking into account the corrosion degradation.

For this purpose, an investigation of corrosion, aluminum, and aluminum corrosion was made. All

types of corrosion were studied as well as its passivation, one of the main features of the material

corrosion-wise. Some mitigation strategies of corrosion were explored along with corrosion, fatigue and

tensile tests and standards.

CEiiA is developing a AUV with this aluminum alloy with hard anodization. The AUV has periods

underwater and periods stored in a warehouse thus the choice of a wet/dry cycle for the experiment.

The wet period was carried out in a salt spray chamber and the dry period in a humidity and temperature

controlled chamber. The total duration of the test was 20 cycles. Between periods three different

washing methods were tested: washing with nitric acid, washing with freshwater and to not wash.

Since anodization is a very expensive procedure, half of the specimens were anodized to have means

of comparison and check its worth. For mechanical and fatigue characterization specimens were

subjected to the tensile test and fatigue test after being exposed to different corrosion times. For each

condition case, three replicates were tested to obtain average, and more accurate, values. A total of 207

specimens were tested.

74

Page 101: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

From the corrosion test, a surface observation analysis, and surface roughness profile were achieved.

In this analysis pitting was confirmed to be the major type of corrosion present, as expected. The

anodization protected the material very effectively but the nitric acid dissolved the protective layer

thoroughly. As for the roughness of the specimens’ surface, the maximum valley depth got deeper

with increase corrosion exposure time as well as the average valley depth. Regarding the valley density,

it had ups and downs since the density of pits increased, however, it coalesced with neighboring pits

creating bigger pits and decreasing valley density.

With the tensile test, some mechanical properties were evaluated. Ultimate and yield stress decreased

in general at a decreasing rate. Young’s modulus did not have a significant variation. Elongation at

fracture and reduction area decreased exponentially to extremely low final values. Ductility properties

were the most affected by corrosion.

Upon the fatigue test, a fatigue life study was carried through. The loss of fatigue life with corrosion

exposure was very significant, especially on the more corroded samples. An S-N curve study was also

included presenting the same results: a low corrosion exposure has a very significant decrease in fatigue

life.

To accomplish a practical application of the results a program was developed. This program uses

the experimental trend lines’ equation to obtain the degradation of properties for any corrosion exposure

time, within a certain range. The software also outputs the material properties of the degraded material

in Nastran ”language” for an easy implementation of a nonlinear static analysis in FEM by any of CEiiA’s

engineers.

6.2 Future Work

The following recommendations are suggested for further investigation:

• The lack of a quantitative correlation between accelerated laboratory corrosion tests and in-service

corrosion attack or atmospheric corrosion tests calls for additional investigation related to corrosion

of aluminum structures;

• Investigating other types of corrosion tests, for example, immersion tests may produce less damaged

metal and more accurate results;

• Since the device, where the material will be used, works in high pressure environments an influence

of pressure in corrosion should be further evaluated;

• Some metallographic cross sections of the anodized specimens could be prepared to analyze the

loss in thickness of the oxide layer along the cycles of exposure;

• To investigate the influence of corrosion on the elastic domain the test can be repeated with more

corrosion cycles and more replicates for each case scenario;

75

Page 102: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

• The plasticity model used can be approached with more accuracy. Suggestion: In this thesis, just

two lines are defined to describe the stress-strain curve, however, this number could be increased,

especially in the plastic zone.

76

Page 103: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Bibliography

[1] Assessment of the Global Cost of Corrosion. URL http://impact.nace.org/economic-impact.

aspx . Acessed on 2017-03-06 .

[2] D. M. Ghiocel and E. J. Tuegel. Reliability Assessment of Aircraft Structure Joints under Corrosion-

Fatigue Damage. In Engineering Design Reliability Applications for the Aerospace, Automotive,

and Ship industries. CRC Press, 2004.

[3] G. S. Chen, M. Gao, and R. P. Wei. Microconstituent-Induced Pitting Corrosion in Aluminum Alloy

2024-T3. CORROSION, 52(1):8–15, jan 1996. ISSN 0010-9312. doi: 10.5006/1.3292099.

[4] G. S. Frankel. Pitting Corrosion of Metals A Review of the Critical Factors. Journal of the

Electrochemical Society, 145(6):2186–2198, 1998. doi: 10.1149/1.1838615.

[5] D. G. Harlow and R. P. Wei. A probability model for the growth o f corrosion pits in aluminum alloys

induced by constituent particles. Engineering Fracture Mechanics, 59(3):305–325, 1998.

[6] G. S. Chen, K.-C. Wan, M. Gao, R. P. Wei, and T. H. Flournoy. A Transition from pitting to fatigue

crack growth modeling of corrosion fatigue crack nucleation in a 2024-T3 aluminum alloy. Materials

Science and Engineering A219, pages 126–132, 1996.

[7] M. Du, F. Chiang, S. Kagwade, and C. Clayton. Damage of Al 2024 alloy due to sequential exposure

to fatigue, corrosion and fatigue. International Journal of Fatigue, 20(10):743–748, nov 1998. ISSN

01421123. doi: 10.1016/S0142-1123(98)00043-7.

[8] B. D. Obert. Quantification of corrosion in 7075-T6 aluminum alloy. Master’s thesis, Texas Tech

University, 2000.

[9] K. K. Sankaran, R. Perez, and K. V. Jata. Effects of pitting corrosion on the fatigue behavior of

aluminum alloy 7075-T6: modeling and experimental studies. Materials Science and Engineering

A297, pages 223–229, 2001.

[10] K. Jones and D. W. Hoeppner. Pit-to-crack transition in pre-corroded 7075-T6 aluminum alloy under

cyclic loading. Corrosion Science 47, 2004. doi: 10.1016/j.corsci.2004.10.004.

[11] R. M. Chlistovsky, P. J. Heffernan, and D. L. DuQuesnay. Corrosion-fatigue behaviour of 7075-T651

aluminum alloy subjected to periodic overloads. International Journal of Fatigue 29, pages

1941–1949, 2007. ISSN 01421123. doi: 10.1016/j.ijfatigue.2007.01.010.

77

Page 104: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

[12] M. K. Cavanaugh, R. G. Buchheit, and N. Birbilis. Modeling the environmental dependence of pit

growth using neural network approaches. Corrosion Science, 52:3070–3077, 2010. doi: 10.1016/

j.corsci.2010.05.027.

[13] H.-C. Liu, H.-P. Xu, L.-Q. Zhu, J.-Z. Liu, X.-B. Ye, and B.-R. Hu. Corrosion behavior of 2A12

aluminum alloy in neutral salt spray environment with different Al-clad removing processes. Metals

and Corrosion, 62(9999):1–5, 2011. doi: 10.1002/maco.201106121.

[14] C. A. Arriscorreta and D. W. Hoeppner. Effects of prior corrosion and stress in corrosion fatigue of

aluminum alloy 7075-T6. Corrosion, 68:950–960, 2012. ISSN 00109312. doi: 10.5006/0512.

[15] E. Cirik and K. Genel. Effect of anodic oxidation on fatigue performance of 7075-T6 alloy. Surface

and Coatings Technology, 202:5190–5201, 2008. doi: 10.1016/j.surfcoat.2008.06.049.

[16] L. Hemmouche, C. Fares, and M. A. Belouchrani. Influence of heat treatments and anodization on

fatigue life of 2017A alloy. Engineering Failure Analysis, 35:554–561, 2013. ISSN 13506307. doi:

10.1016/j.engfailanal.2013.05.003.

[17] K. Dejun and W. Jinchun. Salt spray corrosion and electrochemical corrosion properties of anodic

oxide film on 7475 aluminum alloy. Journal of Alloys and Compounds, 632:286–290, 2015. ISSN

09258388. doi: 10.1016/j.jallcom.2015.01.175.

[18] P. Rambabu, N. E. Prasad, and V. V. Kutumbarao. Aluminium Alloys for Aerospace Applications.

2017. doi: 10.1007/978-981-10-2134-3.

[19] V. R. Crispim and J. J. G. Silva. Detection of Corrosion in Aircraft Aluminum Alloys. Applied

Radiation and Isotopes, 49(7):779–782, 1998.

[20] B. A. Shaw and R. G. Kelly. What is Corrosion? The Electrochemical Society Interface, pages

24–26, 2006.

[21] J.R.Davis. Corrosion of Aluminums and Aluminum Alloys. ASM International R©, 1999.

[22] E. Ghali. Corrosion Resistance of Aluminum and Magnesium Alloys Understanding, Performance,

and Testing. Wiley, 2010.

[23] C. Vargel. Corrosion of Aluminum. Elsevier, 2004.

[24] R. Chang. Chemistry, 9th Edition. McGraw-Hill Education, 2007.

[25] AFSA. Corrosion Resistance of Aluminum and Protective Measures where appropriate. Aluminum

Federation of South Africa, 2011.

[26] K. A. Chandler. Marine and Offshore Corrosion: Marine Engineering Series. Butterworth & Co.

(Publishers) Lda., 1985.

[27] R. Javaherdashti. Technical Mitigation of Corrosion: Corrosion Management. In Microbiologically

Influenced Corrosion, pages 9–16. Springer, 2008.

78

Page 105: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

[28] J. Perryman. Corrosion Resistance of Aluminum. Crane Materials International, 2007.

[29] S. D. Cramer and B. S. CovinoJr. Corrosion: Fundamentals, Testing, and Protection ASM

INTERNATIONAL R© Publication Information and Contributors. 13A, 2003.

[30] R. Baboian. Corrosion Tests and Standards: Application and Interpretation. c©ASTM International,

2nd edition, 2005.

[31] ASTM B117. Standard Practice for Operating Salt Spray (Fog) Apparatus. Standard, American

Society for Testing and Materials, 2011.

[32] Salt Spray Testing. Why it should not be used to compare different types of coatings. Information

sheet, European General Galvanizers Association, June 2013.

[33] J. R. Davis. Tensile Testing. ASM International R©, 2nd edition, 2004.

[34] J. A. Bannantine, J. J. Comer, and J. L. Handrock. Fundamentals of Metal Fatigue Analysis.

Prentice Hall Inc. Englewood Cliffs, 1990.

[35] Aluminum Material Data Sheet. KMS.

[36] ASTM B557M. Standard Test Methods for Tension Testing Wrought and Cast Aluminum and

Magnesium Alloy Products (Metric). Standard, American Society for Testing and Materials, 2015.

[37] ASTM G1. Standard Practice for Preparing Cleaning and Evaluation Corrosion Test Specimens.

Standard, American Society for Testing and Materials, 2003.

[38] Dry Corrosion Test Cabinet Instruction Handbook. Angelantoni Test Technologies.

[39] Average Humidity in Matosinhos. URL https://weather-and-climate.com/

average-monthly-Humidity-perc,matosinhos-norte-region-pt,Portugal . Acessed on

2017-04-06 .

[40] Fitoterm & Fitoclima Reach-in Testing Chambers Catalogue. Aralab.

[41] Surface Roughness Measuring Tester Manual. Mitutoyo.

[42] A. International. ASM Handbook: Mechanical Testing and Evaluation, volume 8. ASM International,

2000.

[43] ASTM E466. Standard Practice for Conducting Force Controlled Constant Amplitude Axial Fatigue

Tests of Metallic Materials. Standard, American Society for Testing and Materials, 1996.

[44] G. J. Stephen, T. Pasang, and B. P. Withy. The effect of pitting corrosion on split sleeve cold

hole expanded, bare 7075-T651 aluminium alloy. Journal of Manufacturing Processes, 5:115–120,

2013. ISSN 15266125. doi: 10.1016/j.jmapro.2012.09.008.

[45] Practical Aspects of Finite Element Simulation - A Study Guide. Altair, 3rd edition, 2015.

[46] MD/MSC Nastran 2010 Quick Reference Guide. MSC Software, 2010.

[47] MSC/NASTRAN Nonlinear Analysis Handbook. The Macneal-Schwendler Corporation, 1992.

79

Page 106: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

80

Page 107: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Appendix A

Specimen Labeling & Tabulated Data

A.1 Nomenclature

Table A.1: Anodized Specimen Label’s Nomenclature

Exposure Time (1,3,9,15,20)Anodized (A)

Nitric Acid (B) Freshwater (C) Nothing (D)Fatigue (F) Tensile (T) Fatigue (F) Tensile (T) Fatigue (F) Tensile (T)X-A-B-F-0x X-A-C-T-0x X-A-C-F-0x X-A-C-T-0x X-A-D-F-0x X-A-D-T-0x

Table A.2: Not Anodized Specimen Label’s Nomenclature

Exposure Time (1,3,9,15,20)Not Anodized (NA)

Nitric Acid (B) Freshwater (C) Nothing (D)Fatigue (F) Tensile (T) Fatigue (F) Tensile (T) Fatigue (F) Tensile (T)

X-NA-B-F-0x X-NA-B-T-0x X-NA-C-F-0x X-NA-C-T-0x X-NA-D-F-0x X-NA-D-T-0x

Table A.3: S-N Curve Specimen Label’s Nomenclature

Exposure Time (0,3,5)Replicate Number (1,2,3,4,5)

x-0x

81

Page 108: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

A.2 Specimen’s Labels

Table A.4: Specimen’s Label

Specimen no Part Label Specimen no Part Label Specimen no Part Label

1 0-A-F-01 65 3-NA-D-F-02 129 15-A-D-F-03

2 0-A-F-02 66 3-NA-D-F-03 130 15-NA-B-F-01

3 0-A-F-03 67 3-A-B-T-01 131 15-NA-B-F-02

4 0-NA-F-01 68 3-A-B-T-02 132 15-NA-B-F-03

5 0-NA-F-02 69 3-A-B-T-03 133 15-NA-C-F-01

6 0-NA-F-03 70 3-A-C-T-01 134 15-NA-C-F-02

7 0-A-T-01 71 3-A-C-T-02 135 15-NA-C-F-03

8 0-A-T-02 72 3-A-C-T-03 136 15-NA-D-F-01

9 0-A-T-03 73 3-A-D-T-01 137 15-NA-D-F-02

10 0-NA-T-01 74 3-A-D-T-02 138 15-NA-D-F-03

11 0-NA-T-02 75 3-A-D-T-03 139 15-A-B-T-01

12 0-NA-T-03 76 3-NA-B-T-01 140 15-A-B-T-02

13 1-A-B-F-01 77 3-NA-B-T-02 141 15-A-B-T-03

14 1-A-B-F-02 78 3-NA-B-T-03 142 15-A-C-T-01

15 1-A-B-F-03 79 3-NA-C-T-01 143 15-A-C-T-02

16 1-A-C-F-01 80 3-NA-C-T-02 144 15-A-C-T-03

17 1-A-C-F-02 81 3-NA-C-T-03 145 15-A-D-T-01

18 1-A-C-F-03 82 3-NA-D-T-01 146 15-A-D-T-02

19 1-A-D-F-01 83 3-NA-D-T-02 147 15-A-D-T-03

20 1-A-D-F-02 84 3-NA-D-T-03 148 15-NA-B-T-01

21 1-A-D-F-03 85 9-A-B-F-01 149 15-NA-B-T-02

22 1-NA-B-F-01 86 9-A-B-F-02 150 15-NA-B-T-03

23 1-NA-B-F-02 87 9-A-B-F-03 151 15-NA-C-T-01

24 1-NA-B-F-03 88 9-A-C-F-01 152 15-NA-C-T-02

25 1-NA-C-F-01 89 9-A-C-F-02 153 15-NA-C-T-03

26 1-NA-C-F-02 90 9-A-C-F-03 154 15-NA-D-T-01

27 1-NA-C-F-03 91 9-A-D-F-01 155 15-NA-D-T-02

28 1-NA-D-F-01 92 9-A-D-F-02 156 15-NA-D-T-03

29 1-NA-D-F-02 93 9-A-D-F-03 157 20-A-B-F-01

30 1-NA-D-F-03 94 9-NA-B-F-01 158 20-A-B-F-02

31 1-A-B-T-01 95 9-NA-B-F-02 159 20-A-B-F-03

32 1-A-B-T-02 96 9-NA-B-F-03 160 20-A-C-F-01

33 1-A-B-T-03 97 9-NA-C-F-01 161 20-A-C-F-02

82

Page 109: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

34 1-A-C-T-01 98 9-NA-C-F-02 162 20-A-C-F-03

35 1-A-C-T-02 99 9-NA-C-F-03 163 20-A-D-F-01

36 1-A-C-T-03 100 9-NA-D-F-01 164 20-A-D-F-02

37 1-A-D-T-01 101 9-NA-D-F-02 165 20-A-D-F-03

38 1-A-D-T-02 102 9-NA-D-F-03 166 20-NA-B-F-01

39 1-A-D-T-03 103 9-A-B-T-01 167 20-NA-B-F-02

40 1-NA-B-T-01 104 9-A-B-T-02 168 20-NA-B-F-03

41 1-NA-B-T-02 105 9-A-B-T-03 169 20-NA-C-F-01

42 1-NA-B-T-03 106 9-A-C-T-01 170 20-NA-C-F-02

43 1-NA-C-T-01 107 9-A-C-T-02 171 20-NA-C-F-03

44 1-NA-C-T-02 108 9-A-C-T-03 172 20-NA-D-F-01

45 1-NA-C-T-03 109 9-A-D-T-01 173 20-NA-D-F-02

46 1-NA-D-T-01 110 9-A-D-T-02 174 20-NA-D-F-03

47 1-NA-D-T-02 111 9-A-D-T-03 175 20-A-B-T-01

48 1-NA-D-T-03 112 9-NA-B-T-01 176 20-A-B-T-02

49 3-A-B-F-01 113 9-NA-B-T-02 177 20-A-B-T-03

50 3-A-B-F-02 114 9-NA-B-T-03 178 20-A-C-T-01

51 3-A-B-F-03 115 9-NA-C-T-01 179 20-A-C-T-02

52 3-A-C-F-01 116 9-NA-C-T-02 180 20-A-C-T-03

53 3-A-C-F-02 117 9-NA-C-T-03 181 20-A-D-T-01

54 3-A-C-F-03 118 9-NA-D-T-01 182 20-A-D-T-02

55 3-A-D-F-01 119 9-NA-D-T-02 183 20-A-D-T-03

56 3-A-D-F-02 120 9-NA-D-T-03 184 20-NA-B-T-01

57 3-A-D-F-03 121 15-A-B-F-01 185 20-NA-B-T-02

58 3-NA-B-F-01 122 15-A-B-F-02 186 20-NA-B-T-03

59 3-NA-B-F-02 123 15-A-B-F-03 187 20-NA-C-T-01

60 3-NA-B-F-03 124 15-A-C-F-01 188 20-NA-C-T-02

61 3-NA-C-F-01 125 15-A-C-F-02 189 20-NA-C-T-03

62 3-NA-C-F-02 126 15-A-C-F-03 190 20-NA-D-T-01

63 3-NA-C-F-03 127 15-A-D-F-01 191 20-NA-D-T-02

64 3-NA-D-F-01 128 15-A-D-F-02 192 20-NA-D-T-03

83

Page 110: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Table A.5: S-N Curve Specimen’s Label

Specimen no Part Label Specimen no Part Label Specimen no Part Label193 0-01 198 3-01 203 15-01194 0-02 199 3-02 204 15-02195 0-03 200 3-03 205 15-03195 0-04 201 3-04 206 15-04197 0-05 202 3-05 207 15-05

A.3 Tensile Test Results

Table A.6: Tensile Test Results

Part NumberUTS

(MPa)

Yield

(MPa)

E

(GPa)

Elong. at Fracture

(%)

Reduction Area

(%)

0-A-T-01 586,34 520,8 70,38 5,70% 14,71%

0-A-T-02 588,94 518,18 70,02 5,30% 18,27%

0-A-T-03 585,86 519 69,67 5,32% 17,36%

0-NA-T-01 571,33 518,85 68,27 6,33% 20,09%

0-NA-T-02 593,88 521,79 70,35 6,32% 16,86%

0-NA-T-03 594,41 526,68 70,22 5,36% 17,58%

1-A-B-T-01 592,04 519,61 70,22 6,24% 16,28%

1-A-B-T-02 588,95 525,12 69,10 5,21% 21,16%

1-A-B-T-03 590,44 524,65 70,42 6,35% 16,07%

1-A-C-T-01 586,16 521,1 70,42 6,21% 18,20%

1-A-C-T-02 592,56 524,23 70,84 6,20% 17,65%

1-A-C-T-03 592,02 526,2 70,63 5,92% 21,79%

1-A-D-T-01 586,23 516,93 70,81 6,05% 18,91%

1-A-D-T-02 587,63 520,25 70,78 6,27% 18,22%

1-A-D-T-03 592,93 525,69 71,04 6,09% 19,02%

1-NA-B-T-01 592,82 523,07 70,21 6,28% 8,82%

1-NA-B-T-02 599,15 531,41 70,86 6,31% 13,97%

1-NA-B-T-03 596,05 530,4 70,72 6,31% 12,70%

1-NA-C-T-01 592,83 527,32 70,31 6,28% 8,49%

1-NA-C-T-02 588,32 530,95 70,79 6,30% 7,17%

1-NA-C-T-03 589,85 522,24 70,57 6,31% 9,00%

1-NA-D-T-01 573,96 517,23 69,90 6,09% 7,18%

1-NA-D-T-02 592,62 529,01 70,53 6,30% 9,06%

1-NA-D-T-03 585,50 525,54 70,07 6,32% 9,25%

3-A-B-T-01 577,71 514,83 69,57 5,94% 6,09%

3-A-B-T-02 572,30 516,4 69,78 5,56% 10,44%

84

Page 111: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

3-A-B-T-03 572,80 513,16 69,35 5,09% 5,37%

3-A-C-T-01 586,32 515,25 70,10 5,52% 9,75%

3-A-C-T-02 584,94 514,88 69,57 5,12% 19,18%

3-A-C-T-03 586,17 518,42 70,53 5,41% 17,85%

3-A-D-T-01 585,83 514,08 70,42 5,31% 18,31%

3-A-D-T-02 587,57 520,66 69,89 5,06% 6,91%

3-A-D-T-03 580,91 523,09 69,75 4,82% 17,78%

3-NA-B-T-01 581,55 520,81 71,34 6,04% 7,48%

3-NA-B-T-02 583,15 520,1 70,28 6,22% 10,63%

3-NA-B-T-03 584,55 525,56 70,07 5,51% 6,55%

3-NA-C-T-01 581,47 524,96 70,00 5,13% 5,41%

3-NA-C-T-02 581,09 520,33 70,32 4,19% 5,21%

3-NA-C-T-03 573,63 521,7 70,50 3,60% 2,75%

3-NA-D-T-01 578,24 519,76 69,77 5,71% 6,86%

3-NA-D-T-02 577,52 522,48 69,66 5,47% 6,47%

3-NA-D-T-03 580,41 518,29 69,11 5,29% 6,15%

9-A-B-T-01 550,92 512,44 69,25 2,37% 4,34%

9-A-B-T-02 545,41 505,61 69,26 2,21% 3,69%

9-A-B-T-03 556,67 499,92 69,43 2,96% 4,00%

9-A-C-T-01 580,16 513,8 70,39 6,31% 11,65%

9-A-C-T-02 582,76 511,61 71,06 6,23% 6,79%

9-A-C-T-03 581,86 516,017 70,69 6,06% 6,66%

9-A-D-T-01 578,23 512,16 70,16 6,25% 13,80%

9-A-D-T-02 576,72 510,53 69,94 6,26% 26,25%

9-A-D-T-03 579,79 512,8 70,25 5,90% 10,70%

9-NA-B-T-01 561,47 539,69 70,09 2,57% 5,64%

9-NA-B-T-02 554,37 505,94 69,31 2,45% 5,58%

9-NA-B-T-03 566,02 505,79 69,29 4,99% 7,36%

9-NA-C-T-01 557,77 512,98 70,27 2,75% 4,10%

9-NA-C-T-02 557,09 511,48 70,07 2,26% 2,40%

9-NA-C-T-03 559,09 516,37 70,74 2,44% 2,64%

9-NA-D-T-01 546,56 508,76 69,69 2,16% 2,19%

9-NA-D-T-02 565,26 514,41 70,47 4,12% 3,51%

9-NA-D-T-03 552,01 515,61 69,68 2,23% 2,53%

15-A-B-T-01 530,88 500,94 68,62 1,33% 4,15%

15-A-B-T-02 523,45 495,11 67,82 1,33% 5,68%

15-A-B-T-03 530,67 496,11 67,04 1,64% 5,01%

85

Page 112: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

15-A-C-T-01 523,12 508,2 69,62 1,00% -0,49%

15-A-C-T-02 568,30 513,18 70,30 3,41% 2,50%

15-A-C-T-03 583,76 510,92 69,99 5,37% 8,16%

15-A-D-T-01 575,51 514,1 70,43 5,62% 8,66%

15-A-D-T-02 580,22 515,49 70,62 6,23% 8,10%

15-A-D-T-03 579,20 514,3 69,50 5,19% 11,85%

15-NA-B-T-01 539,85 502,94 67,97 1,74% 4,68%

15-NA-B-T-02 526,98 504,17 68,13 1,29% 6,03%

15-NA-B-T-03 531,77 505,85 68,36 1,48% 6,09%

15-NA-C-T-01 549,17 521,61 70,49 1,65% 1,14%

15-NA-C-T-02 554,17 518,66 70,09 2,24% 3,03%

15-NA-C-T-03 564,44 524,26 70,85 2,66% 0,55%

15-NA-D-T-01 548,97 506,93 68,50 2,59% 0,35%

15-NA-D-T-02 549,69 505,16 68,26 3,53% 0,67%

15-NA-D-T-03 545,98 514,87 69,58 2,42% 0,12%

20-A-B-T-01 517,53 502,38 66,98 1,19% 3,32%

20-A-B-T-02 511,99 499,18 68,38 1,12% 6,59%

20-A-B-T-03 494,69 492,362 67,44 0,91% 4,73%

20-A-C-T-01 575,50 515,2 70,58 5,35% 4,74%

20-A-C-T-02 552,49 514,81 71,01 1,88% 1,23%

20-A-C-T-03 582,92 518,21 71,48 5,45% 3,38%

20-A-D-T-01 585,67 511,77 70,11 6,16% 10,64%

20-A-D-T-02 583,73 509,49 69,79 5,58% 18,63%

20-A-D-T-03 580,77 516,15 69,75 5,15% 6,78%

20-NA-B-T-01 519,49 493,26 67,57 1,28% 4,80%

20-NA-B-T-02 492,25 481 68,72 0,91% 7,29%

20-NA-B-T-03 520,30 500,55 68,57 1,18% 5,31%

20-NA-C-T-01 534,21 516,73 70,79 1,15% 1,62%

20-NA-C-T-02 554,95 514,26 69,49 1,94% 3,91%

20-NA-C-T-03 542,39 523,75 69,83 1,36% 4,43%

20-NA-D-T-01 521,11 508,59 69,67 1,06% 1,40%

20-NA-D-T-02 544,79 507,66 69,54 2,38% 1,46%

20-NA-D-T-03 505,03 503,71 69,00 0,90% 0,79%

86

Page 113: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Table A.7: Ultimate Tensile Strength Averages and Standard Deviations

UTS (MPa)Specimen Mean Std Dev. Specimen Mean Std Dev.

0-A-T 587.04 1.66 9-A-D-T 578.25 1.530-NA-T 586.54 13.18 9-NA-B-T 560.62 5.871-A-B-T 590.47 1.54 9-NA-C-T 557.98 1.011-A-C-T 590.25 3.55 9-NA-D-T 554.61 9.621-A-D-T 588.93 3.53 15-A-B-T 528.33 4.23

1-NA-B-T 596.01 3.17 15-A-C-T 558.39 31.511-NA-C-T 590.33 2.29 15-A-D-T 578.31 2.481-NA-D-T 584.03 9.41 15-NA-B-T 532.87 6.503-A-B-T 574.27 2.99 15-NA-C-T 555.93 7.793-A-C-T 585.81 0.76 15-NA-D-T 548.21 1.973-A-D-T 584.77 3.45 20-A-B-T 508.07 11.92

3-NA-B-T 583.08 1.50 20-A-C-T 570.30 15.873-NA-C-T 578.73 4.42 20-A-D-T 583.39 2.473-NA-D-T 578.72 1.50 20-NA-B-T 510.68 15.969-A-B-T 551.00 5.63 20-NA-C-T 543.85 10.459-A-C-T 581.59 1.32 20-NA-D-T 523.64 20.00

Table A.8: Yield Stress Averages and Standard Deviations

Yield Stress (MPa)Specimen Mean Std Dev. Specimen Mean Std Dev.

0-A-T 519.33 1.34 9-A-D-T 511.83 1.170-NA-T 522.44 3.96 9-NA-B-T 517.14 19.531-A-B-T 523.13 3.05 9-NA-C-T 513.61 2.511-A-C-T 523.84 2.57 9-NA-D-T 512.93 3.661-A-D-T 520.96 4.42 15-A-B-T 497.39 3.12

1-NA-B-T 528.29 4.55 15-A-C-T 510.77 2.491-NA-C-T 526.84 4.38 15-A-D-T 514.63 0.751-NA-D-T 523.93 6.05 15-NA-B-T 504.32 1.463-A-B-T 514.80 1.62 15-NA-C-T 521.51 2.803-A-C-T 516.18 1.95 15-NA-D-T 508.99 5.173-A-D-T 519.28 4.66 20-A-B-T 497.97 5.12

3-NA-B-T 522.16 2.97 20-A-C-T 516.07 1.863-NA-C-T 522.33 2.38 20-A-D-T 512.47 3.383-NA-D-T 520.18 2.13 20-NA-B-T 491.60 9.889-A-B-T 505.99 6.27 20-NA-C-T 518.25 4.929-A-C-T 513.81 2.20 20-NA-D-T 506.65 2.59

87

Page 114: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Table A.9: Young’s Modulus Averages and Standard Deviations

Young’s Moludus (GPa)Specimen Mean Std Dev. Specimen Mean Std Dev.

0-A-T 70.02 0.36 9-A-D-T 70.11 0.160-NA-T 69.62 1.17 9-NA-B-T 69.56 0.461-A-B-T 69.91 0.71 9-NA-C-T 70.36 0.341-A-C-T 70.63 0.21 9-NA-D-T 69.95 0.451-A-D-T 70.88 0.14 15-A-B-T 67.83 0.79

1-NA-B-T 70.60 0.34 15-A-C-T 69.97 0.341-NA-C-T 70.56 0.24 15-A-D-T 70.18 0.601-NA-D-T 70.17 0.33 15-NA-B-T 68.15 0.203-A-B-T 69.57 0.22 15-NA-C-T 70.48 0.383-A-C-T 70.07 0.48 15-NA-D-T 68.79 0.803-A-D-T 70.02 0.36 20-A-B-T 67.60 0.71

3-NA-B-T 70.57 0.68 20-A-C-T 71.02 0.453-NA-C-T 70.27 0.26 20-A-D-T 69.88 0.193-NA-D-T 69.51 0.36 20-NA-B-T 68.28 0.629-A-B-T 69.31 0.10 20-NA-C-T 70.04 0.679-A-C-T 70.71 0.33 20-NA-D-T 69.41 0.36

Table A.10: Elongation at Fracture Averages and Standard Deviations

Elongation at fracture (%)Specimen Mean Std Dev. Specimen Mean Std Dev.

0-A-T 5.44 0.23 9-A-D-T 6.14 0.210-NA-T 6.00 0.56 9-NA-B-T 3.34 1.441-A-B-T 5.93 0.63 9-NA-C-T 2.48 0.241-A-C-T 6.11 0.16 9-NA-D-T 2.84 1.111-A-D-T 6.14 0.12 15-A-B-T 1.44 0.18

1-NA-B-T 6.30 0.02 15-A-C-T 3.26 2.191-NA-C-T 6.30 0.02 15-A-D-T 5.68 0.521-NA-D-T 6.24 0.13 15-NA-B-T 1.51 0.233-A-B-T 5.53 0.42 15-NA-C-T 2.18 0.513-A-C-T 5.35 0.20 15-NA-D-T 2.85 0.603-A-D-T 5.06 0.25 20-A-B-T 1.07 0.15

3-NA-B-T 5.92 0.37 20-A-C-T 4.23 2.033-NA-C-T 4.31 0.77 20-A-D-T 5.63 0.513-NA-D-T 5.49 0.21 20-NA-B-T 1.12 0.199-A-B-T 2.51 0.39 20-NA-C-T 1.48 0.419-A-C-T 6.20 0.13 20-NA-D-T 1.45 0.82

88

Page 115: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Table A.11: Reduction Area Averages and Standard Deviations

Reduction Area (%)Specimen Mean Std Dev. Specimen Mean Std Dev.

0-A-T 16.78 1.85 9-A-D-T 16.92 8.230-NA-T 18.18 1.69 9-NA-B-T 6.19 1.011-A-B-T 17.84 2.88 9-NA-C-T 3.05 0.921-A-C-T 19.21 2.25 9-NA-D-T 2.74 0.681-A-D-T 18.71 0.43 15-A-B-T 4.95 0.77

1-NA-B-T 11.83 2.68 15-A-C-T 3.39 4.391-NA-C-T 8.22 0.95 15-A-D-T 9.54 2.031-NA-D-T 8.50 1.15 15-NA-B-T 5.60 0.803-A-B-T 7.30 2.75 15-NA-C-T 1.57 1.303-A-C-T 15.59 5.10 15-NA-D-T 0.38 0.283-A-D-T 14.33 6.44 20-A-B-T 4.88 1.64

3-NA-B-T 8.22 2.14 20-A-C-T 3.12 1.773-NA-C-T 4.46 1.48 20-A-D-T 12.02 6.053-NA-D-T 6.49 0.35 20-NA-B-T 5.80 1.329-A-B-T 4.01 0.33 20-NA-C-T 3.32 1.499-A-C-T 8.37 2.84 20-NA-D-T 1.22 0.37

89

Page 116: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

A.4 Fatigue Test Results

Table A.12: Fatigue Test Results

Part Number Fatigue Life(Cycles) Part Number Fatigue Life

(Cycles) Part Number Fatigue Life(Cycles)

0-A-F-01 54082 3-A-D-F-03 46470 15-A-C-F-02 306870-A-F-02 84155 3-NA-B-F-01 61192 15-A-C-F-03 837600-A-F-03 68854 3-NA-B-F-02 70135 15-A-D-F-01 41000

0-NA-F-01 1000000 3-NA-B-F-03 101966 15-A-D-F-02 377430-NA-F-02 1000000 3-NA-C-F-01 63315 15-A-D-F-03 364550-NA-F-03 348815 3-NA-C-F-02 40520 15-NA-B-F-01 242451-A-B-F-01 100218 3-NA-C-F-03 51854 15-NA-B-F-02 298211-A-B-F-02 93985 3-NA-D-F-01 59484 15-NA-B-F-03 218951-A-B-F-03 111115 3-NA-D-F-02 55936 15-NA-C-F-01 421531-A-C-F-01 47221 3-NA-D-F-03 63154 15-NA-C-F-02 404651-A-C-F-02 41924 9-A-B-F-01 43232 15-NA-C-F-03 375941-A-C-F-03 105400 9-A-B-F-02 38655 15-NA-D-F-01 246481-A-D-F-01 45495 9-A-B-F-03 41084 15-NA-D-F-02 283781-A-D-F-02 81574 9-A-C-F-01 34759 15-NA-D-F-03 273311-A-D-F-03 80000 9-A-C-F-02 60817 20-A-B-F-01 25657

1-NA-B-F-01 126222 9-A-C-F-03 47448 20-A-B-F-02 244991-NA-B-F-02 111551 9-A-D-F-01 40789 20-A-B-F-03 255931-NA-B-F-03 147080 9-A-D-F-02 37524 20-A-C-F-01 316141-NA-C-F-01 169144 9-A-D-F-03 40136 20-A-C-F-02 393311-NA-C-F-02 101426 9-NA-B-F-01 43964 20-A-C-F-03 509091-NA-C-F-03 107084 9-NA-B-F-02 46660 20-A-D-F-01 420991-NA-D-F-01 72104 9-NA-B-F-03 41931 20-A-D-F-02 534251-NA-D-F-02 223605 9-NA-C-F-01 45514 20-A-D-F-03 310211-NA-D-F-03 100000 9-NA-C-F-02 35100 20-NA-B-F-01 241453-A-B-F-01 70000 9-NA-C-F-03 27691 20-NA-B-F-02 279433-A-B-F-02 71284 9-NA-D-F-01 36896 20-NA-B-F-03 271123-A-B-F-03 64585 9-NA-D-F-02 30000 20-NA-C-F-01 360953-A-C-F-01 50335 9-NA-D-F-03 26741 20-NA-C-F-02 570273-A-C-F-02 67316 15-A-B-F-01 43968 20-NA-C-F-03 320983-A-C-F-03 67864 15-A-B-F-02 19289 20-NA-D-F-01 193013-A-D-F-01 45000 15-A-B-F-03 36211 20-NA-D-F-02 204993-A-D-F-02 47398 15-A-C-F-01 48133 20-NA-D-F-03 22912

90

Page 117: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Table A.13: Fatigue Life Averages and Standard Deviations

Fatigue Life (cycles)Specimen Mean Std Dev. Specimen Mean Std Dev.

0-A-F 69030 15037 9-A-D-F 39483 17280-NA-F 782938 375962 9-NA-B-F 44185 23721-A-B-F 101773 8670 9-NA-C-F 36102 89541-A-C-F 64848 35219 9-NA-D-F 31212 51851-A-D-F 69023 20391 15-A-B-F 33156 12620

1-NA-B-F 128284 17854 15-A-C-F 54193 270511-NA-C-F 125885 37570 15-A-D-F 38399 23431-NA-D-F 131903 80632 15-NA-B-F 25320 40713-A-B-F 68623 3555 15-NA-C-F 40071 23053-A-C-F 61838 9966 15-NA-D-F 26786 19243-A-D-F 46289 1209 20-A-B-F 25250 651

3-NA-B-F 77764 11398 20-A-C-F 40618 97123-NA-C-F 51896 11398 20-A-D-F 42182 112023-NA-D-F 59525 3609 20-NA-B-F 26400 19979-A-B-F 40990 2290 20-NA-C-F 41740 133899-A-C-F 47675 13030 20-NA-D-F 20904 1839

Table A.14: Fatigue Test Results for S-N Curve Specimens

Specimen Maximum Stress [MPa] Cycles to Failure0-01 270,0 420540-02 260,0 663330-03 250,0 1070010-04 235,0 3021150-05 205,0 9420563-01 165,5 502333-02 155,0 705183-03 145,0 859303-04 125,0 1810553-05 120,0 100000015-01 165,5 3345015-02 155,0 4911515-03 135,0 7198715-04 115,0 17309415-05 100,0 320044

91

Page 118: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

92

Page 119: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Appendix B

Equipment, Software and Support

Material List

B.1 Corrosion Test

Equipment:

• ACS Dry Corrosion Test Cabinet (DCTC) 1200;

• Aralab Climatic Chamber Fitoclima 500 EP20;

• Nikon Microscope Eclipse MA100;

• MITUTOYO Surface Roughness Measuring Tester.

Software:

• Nikon NIS Elements D.

Support Materials:

• 7075-T651 Aluminum Alloy Specimens;

• Glass-fiber reinforced epoxy resin specimens’ supports;

• NaCl;

• Water;

• 3MTM Corrosion Resistant Tape;

• Label Maker + Self-adhesive tape;

• Zip lock plastic bags;

• Nitric Acid (68%);

93

Page 120: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

• Acid resistant tweezers;

• Glass bowls;

• Gas Mask;

• Protection Gloves;

• Disposable Clothes.

• KERN precision analytical balance;

B.2 Tensile Test

Equipment:

• MTS Servohydraulic Testing Machine;

• Load cell of 100kN;

• Linear Variable Differential Transformer (LVDT);

• MTS 632.85 Biaxial Extensometer.

Software:

• Software MTSTestSuiteTM TW Elite.

Support Materials:

• Protective gloves;

• Goggles;

• MITUTOYO Digimatic Caliper;

• MITUTOYO Digimatic Micrometer.

B.3 Fatigue Test

Equipment:

• MTS Servohydraulic Testing Machine;

• Load cell of 50kN;

• Linear Variable Differential Transformer (LVDT).

Software:

• Software MTSTestSuiteTM Multipurpose Elite.

94

Page 121: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Support Materials:

• Protective gloves;

• Goggles;

• MITUTOYO Digimatic Caliper;

• MITUTOYO Digimatic Micrometer.

95

Page 122: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

96

Page 123: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Appendix C

Output.txt Example File

------------

OUTPUT FILE

------------

The inputs inserted by the user were:

Number of Corrosion Cycles = 15

Condition Code = AB

Condition Codes Legend

AB - Anodized washed with nitric acid

AC - Anodized washed with Freshwater

AD - Anodized not washed

NAB - Not anodized washed with nitric acid

NAC - Not anodized washed with freshwater

NAD - Not anodized not washed

------------

Allowables:

------------

Yield Stress = 500.764 MPa

Ultimate Tensile Strength = 527.965 MPa

WARNING: If the number of cycles inserted is greater than 20 the results

correspond to extrapolations and may not be phisically possible.

(ex. Yield Stres > Ultimate Tensile Stress)

----------------------------------------------------------

Lines to change in Nastran Bulk section according to Input:

----------------------------------------------------------

NOTE: This code corresponds to the properties of the corroded material ready

to copy to the nastran bulk section. See instructions below.

MAT1 170530.1 0.32

97

Page 124: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

MATS1 1 7 PLASTIC 500.764

TABLES1 7

+ 0.0 0.0 0.0071 500.764 0.0087 506.204 0.0103 511.644

+ 0.0120 517.085 0.0136 522.525 0.0152 527.965ENDT

NLPARM 3

ATTENTION: The non linear parameters, material and table identifications have

to be coherent along the card!

-------------

INSTRUCTIONS

-------------

To run a SOL 106 from a SOL 101 Card in Nastran the following changes need to be made:

-Change SOL 101 to SOL 106;

-Change ANALYSIS=STATICS to ANALYSIS=NLSTAT inside every SUBCASE created;

-Inside every SUBCASE create a new line identifying the non linear parameters

(ex. SUBCASE 1

LABEL= Case1

SPC = 1

LOAD = 4

NLPARM = 3

ANALYSIS = NLSTAT);

-If PSHELL is being used make sure MID3 is blank;

-In the material definitions add MATS1.

(ex MAT1 170541.9 0.32

MATS1 1 7 PLASTIC 522.010)

In MAT1 there is the material identification, young modulus and poisson ratio.

MATS1 has the material identification, the table of the plastic properties identification,

PLASTIC means the material has a plastic behaviour and Yield Stress value;

-A new line with the NLPARM needs to be created, inside bulk section, with the identification

coincident with the value inserted in the SUBCASES.

(ex.NLPARM 3 )

This defines the non linear parameters, when blank they are all set to DEFAULT.

For more information consult NASTRAN Quick Guide;

-A TABLES1 has to be created, inside bulk section, with points of stress and strain in

the plastic region. The first point of the table has to be the origin and the second point

the yield point. (ex. TABLES1 7

+ 0.0 0.0 0.0074 522.010 0.0188 534.342 0.0302 546.675

+ 0.0416 559.008 0.0530 571.340 0.0644 583.673ENDT );

NOTE: Make sure the identification matches the one in MATS1.

98

Page 125: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

Appendix D

MATLAB R© Code

%main .m

c lea r a l l

c l ea r v a r i a b l es

c l c

%Inpu t window

prompt = { ’ Enter number o f Corrosion Cycles : ’ , ’ Enter Cond i t ion Code : ’ } ;

d l g t i t l e = ’ Inpu t ’ ;

num l ines = 1;

de fau l tans = { ’ 1 ’ , ’NAC ’ } ;

answer = i n p u t d l g ( prompt , d l g t i t l e , num lines , de fau l tans ) ;

a=str2num ( answer {1} ) ;

c =( answer {2} ) ;

%Choose the appropr ia te equat ions

[ y i e l d , uts ,m, b , s t r a i n , s t ress ] = output ( a , answer ( 2 ) ) ;

%Create cards

[E, x , y , mcorr , bcor r ]= createcards ( y i e l d , uts ,m, b ) ;

%Create Output . t x t f i l e

readme ( s t ress , s t r a i n ,E, x , y , a , c , y i e l d , u ts ) ;

s t r a i nna =[0 ,0.002428534 ,0.00476622 ,0.007152668 ,0.011186372 ,0.016280468 ,0.021640334 ,

0.027080399 ,0.032465093 ,0.037778903 ,0.043036954 ,0.04824592 ,0.053386871 ,

0.058478238 ,0.063046455 ,0.063286394 ,0.063285746 ,0.063289363 ,0.063286491 ,

0.063277312 ,0.06327809];

s t ressna =[0 ,166.1947939 ,320.9182345 ,465.7445748 ,522.704083 ,534.2880456 ,541.2990989 ,

546.5870269 ,551.1725941 ,555.0602771 ,558.3422318 ,561.2526298 ,563.7515936 ,

564.8709973 ,566.6947171 ,569.5291675 ,568.8804743 ,568.9550326 ,568.0250862 ,

563.478985 ,547.5553072];

s t ressa =[0 ,194.9855615 ,372.9674711 ,509.9800842 ,542.7597344 ,554.996967 ,559.4110538 ,

99

Page 126: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

566.9920495 ,571.1266851 ,575.3377728 ,578.1160497 ,580.3276145 ,581.6884431 ,

583.0192662 ,584.4870148 ,585.1087027 ,585.774628 ,586.0303935 ,585.2051928 ,

582.522738 ,577.3729015];

s t r a i n a =[0 ,0.002771795 ,0.00543921 ,0.008716721 ,0.014281788 ,0.020396095 ,0.026400671 ,

0.032340389 ,0.038217589 ,0.044101387 ,0.049324702 ,0.053052343 ,0.055618996 ,

0.056723226 ,0.056820642 ,0.056825491 ,0.05682876 ,0.056829168 ,0.056462957 ,

0.056829443 ,0.056831436];

f u n c t i o n [ y i e l d , uts ,m, b , s t r a i n , s t ress ] = output (N, type )

i f strcmp ( type , ’NAC ’ ) ==1

y i e l d = (−0.2888)∗N + 523.14;

u ts = −2.2367∗N+586.79;

m = 757.36∗exp (0.0921∗N) ;

b= −3.2007∗N+544.16;

s t r a i n = s t r a i nna ;

s t ress =st ressna ;

e l s e i f strcmp ( type , ’NAB ’ ) ==1

y i e l d = (−1.6317)∗N + 527.58;

u ts = −4.0991∗N+594.43;

m = 632.71∗exp (0.1066∗N) ;

b= −4.8979∗N+522.83;

s t r a i n = s t r a i nna ;

s t ress =st ressna ;

e l s e i f strcmp ( type , ’NAD ’ ) ==1

y i e l d = (−0.8802)∗N + 522.89;

u ts = −3.0069∗N+586.68;

m = 658.41∗exp (0.0892∗N) ;

b= −3.2741∗N+540.58;

s t r a i n = s t r a i nna ;

s t ress =st ressna ;

e lse i f strcmp ( type , ’AB ’ ) ==1

y i e l d = (−1.2864)∗N + 520.06;

u ts = −4.081∗N+589.18;

m = 751.41∗exp (0.1197∗N) ;

b= −5.5492∗N+542.28;

s t ress =s t ressa ;

s t r a i n = s t r a i n a ;

e l s e i f strcmp ( type , ’AC ’ ) ==1

y i e l d = (−0.3626)∗N + 519.57;

u ts = −1.2746∗N+589.09;

100

Page 127: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

m = 743.92∗exp (0.063∗N) ;

b= −1.8988∗N+538.67;

s t ress =s t ressa ;

s t r a i n = s t r a i n a ;

e l s e i f strcmp ( type , ’AD ’ ) ==1

y i e l d = (−0.4031)∗N + 519.64;

u ts = −0.3482∗N+586.24;

m = 543.29∗exp (0.0373∗N) ;

b= −0.4934∗N+533.01;

s t ress =s t ressa ;

s t r a i n = s t r a i n a ;

e lse

h = msgbox ( ’ I n v a l i d Condi t ion ’ , ’ E r ro r ’ , ’ e r r o r ’ ) ;

r e t u r n

end

end

f u n c t i o n [E, x , y , mcorr , bcor r ]= createcards ( y i e l d , uts ,m, b )

i f y i e l d > uts

f p r i n t f ( ’WARNING Yie ld g rea te r than UTS ’ ) ;

end

%Area of the specimen

Area =12.5∗3;

% Read Base . bdf and d i v i d e i t i n t o l i n e s

f o r n = 1 : 20

f i d = fopen ( ’ base . bdf ’ , ’ r ’ ) ;

i = 1 ;

t l i n e = f g e t l ( f i d ) ;

A{ i } = t l i n e ;

wh i le i scha r ( t l i n e )

i = i +1;

t l i n e = f g e t l ( f i d ) ;

A{ i } = t l i n e ;

end

f c l o s e ( f i d ) ;

%Change Y ie ld i n MATS1

B1 = ’MATS1 1 7 PLASTIC ’ ;

B2 = y i e l d ;

A{239} = s p r i n t f ( ’%s%.3 f%s ’ ,B1 , B2) ;

101

Page 128: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

% Change Force

Fmax=uts ∗Area ;

A1 = ’FORCE 4 142 01.0 0.0 ’ ;

A2 = n∗Fmax/ 2 0 ;

A3= ’ 0.0 ’ ;

A{282} = s p r i n t f ( ’%s%.2 f %s ’ ,A1 , A2 , A3) ;

%Change TABLES1

E=71000.0;

y ( 1 ) =0; x ( 1 ) =0;

y ( 2 ) = y i e l d ; x ( 2 ) = y ( 2 ) /E ;

y ( 7 ) = uts ; x ( 7 ) = ( y ( 7 )−b ) /m;

mcorr = ( ( y ( 7 )−y ( 2 ) ) / ( x ( 7 )−x ( 2 ) ) ) ;

bcor r= y ( 7 )−mcorr∗x ( 7 ) ;

f o r j = 3 :1 :6

y ( j ) = y i e l d + ( ( uts−y i e l d ) / 5 ) ∗ ( j −2) ;

x ( j ) = ( y ( j )−bcor r ) / mcorr ;

end

C1= ’+ 0.0 0.0 ’ ;

A{280} = s p r i n t f ( ’%s%.4 f %.3 f %.4 f %.3 f %.4 f %.3 f ’ ,C1 , x ( 2 ) , y ( 2 ) , x ( 3 ) , y ( 3 ) , x ( 4 ) ,

y ( 4 ) ) ;

D1= ’+ ’ ;

D2= ’ENDT ’ ;

A{281} = s p r i n t f ( ’%s%.4 f %.3 f %.4 f %.3 f %.4 f %.3 f%s ’ ,D1, x ( 5 ) , y ( 5 ) , x ( 6 ) , y ( 6 ) , x ( 7 ) ,

y ( 7 ) ,D2) ;

%Change E i n MAT1

X=round ( x , 4 ) ;

Y=round ( y , 3 ) ;

%nastran gives an er ros i f E i s not the exact number o f the

%slope of the two f i r s t po in t s o f TABLES1

E=Y( 2 ) /X( 2 ) ;

E1= ’MAT1 1 ’ ;

E2= ’ 0.32 ’ ;

A{238} = s p r i n t f ( ’%s%.1 f%s ’ ,E1 ,E, E2) ;

% Wri te new c e l l s i n t o . bdf

f i lename = s p r i n t f ( ’%d . bdf ’ , n ) ;

f i d = fopen ( f i lename , ’w+ ’ ) ;

f o r i = 1 : numel (A)

i f A{ i +1} == −1

f p r i n t f ( f i d , ’%s ’ , A{ i } ) ;

102

Page 129: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

break

else

f p r i n t f ( f i d , ’%s\n ’ , A{ i } ) ;

end

end

end

end

%main2 .m

%Read the cards and e x t r a c t the d i s i r e d r e s u l t s

f o r n=1:1:20

f i lename = s p r i n t f ( ’%d . f06 ’ ,n ) ;

A= f i l e r e a d ( f i lename ) ;

C= s t r f i n d (A, ’ 0 64 ’ ) ;

s t ress =A(C+28:C+40) ;

s t r a i n =A(C+152:C+165) ;

s t ress =str2num ( s t ress ) ;

s t r a i n =str2num ( s t r a i n ) ;

y ( n ) =abs ( s t ress ) ;

x ( n ) =abs ( s t r a i n ) ;

end

%Create the output graph

[ a ( 1 ) ,a ( 2 ) ,a ( 3 ) ,a ( 4 ) ,a ( 5 ) ,a ( 6 ) ,a ( 7 ) ,b ( 1 ) ,b ( 2 ) ,b ( 3 ) ,b ( 4 ) ,b ( 5 ) ,b ( 6 ) ,b ( 7 ) ,

x0 ( 1 ) , x0 ( 2 ) , x0 ( 3 ) , x0 ( 4 ) , x0 ( 5 ) , x0 ( 6 ) , x0 ( 7 ) , x0 ( 8 ) , x0 ( 9 ) , x0 (10) , x0 (11) ,

x0 (12) , x0 (13) , x0 (14) , x0 (15) , x0 (16) , x0 (17) , x0 (18) , x0 (19) , x0 (20) , x0 (21) ,

y0 ( 1 ) , y0 ( 2 ) , y0 ( 3 ) , y0 ( 4 ) , y0 ( 5 ) , y0 ( 6 ) , y0 ( 7 ) , y0 ( 8 ) , y0 ( 9 ) , y0 (10) , y0 (11) ,

y0 (12) , y0 (13) , y0 (14) , y0 (15) , y0 (16) , y0 (17) , y0 (18) , y0 (19) , y0 (20) ,

y0 (21) ] = tex t read ( ’ data . t x t ’ ) , ’ %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.3 f ’

’ %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f ’

’ %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.4 f %.3 f %.3 f %.3 f ’

’ %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f %.3 f ’

’ %.3 f %.3 f %.3 f ’ ) ;

de le te ( ’ data . t x t ’ ) ;

p l o t ( a , b , x0 , y0 , x , y ) ;

t i t l e ( ’ Stress−S t r a i n Curve ’ ) ;

x l a b e l ( ’ S t r a i n [mm/mm] ’ ) % x−ax is l a b e l

y l a b e l ( ’ St ress [MPa] ’ ) % y−ax is l a b e l

legend ( ’ I n t e r p o l a t i o n ’ , ’ Uncorroded Exper imental ’ , ’FEM ’ ) ;

103

Page 130: Corrosion Behavior of 7075-T651 Aluminum Alloy under ...

104