Couplant Effect and Evaluation of FSW AA6061-T6 Butt Welded Joint

o Attribution — You must give appropriate credit, provide a link to the license, and indicate if changes were made. You may do so in any reasonable manner, but not in any way that suggests the licensor endorses you or your use.

o NonCommercial — You may not use the material for commercial purposes.

o ShareAlike — If you remix, transform, or build upon the material, you must distribute your contributions under the same license as the original.

How to cite this thesis

Surname, Initial(s). (2012) Title of the thesis or dissertation. PhD. (Chemistry)/ M.Sc. (Physics)/ M.A. (Philosophy)/M.Com. (Finance) etc. [Unpublished]: University of Johannesburg. Retrieved from: https://ujcontent.uj.ac.za/vital/access/manager/Index?site_name=Research%20Output (Accessed: Date).

Couplant Effect and Evaluation of FSW AA6061-T6 Butt Welded Joint

Itai Mumvenge

217093863

Submitted in partial fulfilment of the requirements for the degree of Master of Engineering in Mechanical Engineering

In the

Department of Mechanical Engineering Science

Of the

Faculty of Engineering and the Built Environment

At the

University of Johannesburg, South Africa

Supervisor: Dr Stephen A. Akinlabi

Co-Supervisor: Dr Peter M. Mashinini

November, 2017

1. DEDICATION

This dissertation is dedicated to my late grandmother Esnath Mvenge

2. COPYRIGHT STATEMENT

The copyright of this dissertation is owned by the University of Johannesburg, South Africa. No information derived from this publication may be published without the author’s prior consent, unless correctly referenced.

………………………………….. 25 November 2017

Author’s Signature Date:

iii 3. AUTHOR DECLARATION

I, Mumvenge Itai hereby declare that the research work documented in this dissertation is my own, and no portion of the work has been submitted in support of an application for another degree or qualification at this or any other university or institute of learning. All sources used or cited are documented and recognised.

………………………………….. 25 November 2017

Author’s Signature Date:

iv 4. ACKNOWLEDGMENT

I would like to acknowledge the following people for their help and support:

 Dr Stephen A. Akinlabi, for introducing me to the Friction Stir Welded process and his immense guidance in the early structuring of this study. His supervision and assistance with interpretation of results made this study a success.  Dr Peter Madindwa Mashinini, for his utmost guidance during this study. I am extremely grateful for his input and feedback. Without his valuable contribution, the study would never have been a success.  Prof Satish (India), for the preparation of the 9 friction stir welded AA6061 samples and clearly marking the process parameters on the samples. Without these samples the study would not have been conducted.  Mr. Shawn Crous and Ms. Ntombi Mthimunye, for helping me with performing the NDT test on the samples.  Mr. Brian Bakkes and Mikateko Shuma, for providing the training on Metallographic sample preparations.  Prof Esther T. Akinlabi and Phd Candidate Oladije, for their invaluable contribution during the laboratory testing on the samples and assistance in producing the experiment results.

Special thanks due to:

 The University of Johannesburg, Mechanical Engineering Material Science Department, for granting me this wonderful opportunity to further my studies.  Ms B.S. Mathobela for being there on every step providing emotional support and encouragement to overcome study challenges and assisting in creating a balance between my work, personal life and studies.

Personal thanks due to:

 My late grandmother Ms. Esnath Mvenge, to whom I dedicate this dissertation. Grandmother, thank you for taking care of me and showing me the direction to become independent no-matter what challenges arose during my upbringing.

5. ABSTRACT

Couplant Effect and Evaluation of FSW AA6061-T6 butt welded joint

Mumvenge I.

Aluminium is an alloy material vastly used for manufacturing components in aviation, transport and a host of commercial use. Friction stir welding is a novelty solid state welding technique that was invented at The Welding Institute (TWI) of UK in 1991.

This study presents the effects of process parameters on achieving sound welds of the friction stir butt welded joint of similar AA6061-T6 Aluminium alloys. The only parameters used and varied for this study were rotational speed and feed rate. The geometry of the tool was kept constant and the material used was tool steel, W302. Friction stir welds were evaluated both by visual inspection and non-destructive testing methods. Evaluation allowed for assessment of the weld integrity by examining for the presence of weld defects. The results indicated that the welds do not have any defects. The weld macrostructure and microstructure were examined and mechanical properties evaluated. The microhardness was also evaluated which showed that optimum speeds are required to achieve uniform hardness value across the weld traverse. The base metal showed higher hardness values when compared to the weld region. However, higher rotational speeds and higher feed rates result in increased hardness values, with the highest values recorded in the weld nugget and the least values recorded in the HAZ/TMAZ interfaces with a W-shaped hardness distribution.

Couplant attenuation effect was investigated using four different types of couplants, employing Ultrasonic Testing contact pulse-echo longitudinal wave to detect weld defects. The attenuation effect was minimal when the Ultrasonic gel, oil and water were used. This was attributed to the coupling conditions such as the acoustic impedance values and the viscous properties of the couplants. Conversely, grease provided poorer frequency gain than all the couplants though it had comparable acoustic impedance values. This was attributed to its poor wetting properties and development of air pockets on contact with weld sample to be examined. vi

6. TABLE OF CONTENTS

Copyright Statement...... iii Author Declaration...... iv Acknowledgment ...... v Abstract ...... vi Table of Contents ...... vii Abbreviations ...... x Nomenclature ...... xiii Units ...... xiv LIST Of Figures ...... xv List Of Tables ...... xxi Glossary of Terms ...... xxii 1. Background...... 1

1.1 Introduction ...... 1 1.2 Problem Statement ...... 4 1.3 Aim ...... 5 1.4 hypothesis ...... 5 1.5 Scope of research ...... 5 1.7 Research Methods ...... 6 1.8 Delimitations ...... 9 1.9 Significance of the Research...... 9 1.10 Study Layout ...... 9 2. Literature Review ...... 11

2 Introduction ...... 11 2.1 Welding ...... 11 2.1.1 Fusion Welding ...... 12

2.1.2 Solid-state Welding ...... 13

2.2 Friction Stir Welding ...... 13 2.3 Application of FSW ...... 16 vii

2.4 Friction Stir Weld Process Parameters ...... 18 2.5 Tool Design and Geometry ...... 19 2.6 Tool Tilt and Plunge Depth ...... 23 2.7 Transverse and Rotational Speeds ...... 25 2.8 Tool Axial Force ...... 26 2.9 Material Flow in FSW ...... 28 2.10 FSW Aluminium Microstructure ...... 30 2.10.1 Weld Nugget...... 32

2.10.2 Thermo-mechanical affected zone (TMAZ) ...... 34

2.10.3 Heat affected zone (HAZ) ...... 35

2.11 Mechanical Properties of FSW joints ...... 36 2.11.1 Microhardness ...... 36

2.11.2 Tensile Properties ...... 38

2.12 Defects ...... 40 2.13 Wave propagation and attenuation theory in Ultrasonic Testing ...... 41 2.13.1 Sound waves behaviour at boundaries ...... 41

2.13.2 Phased Array Ultrasonic Testing ...... 45

2.14 Radiographic testing application in FSW defect analysis ...... 48 2.15 Liquid Dye Penetrant testing application in FSW defect analysis ...... 49 2.16 Summary ...... 50 3 Experimental Procedure ...... 52

3.1 Introduction ...... 52 3.2 Parent Material ...... 52 3.3 FSW Process ...... 53 3.4 Sample Preparation ...... 55 3.5 Microscopy ...... 60 3.6 Vickers Hardness Testing ...... 61 3.7 Tensile Testing ...... 62 3.8 Liquid dye penetrant testing ...... 64 3.9 Digital radiography testing ...... 64 3.10 Ultrasonic Testing ...... 67 viii

3.10.1 Couplant attenuation effects Investigation ...... 67

3.10.2 Time of Flight Diffraction (ToFD) ...... 69

3.10.3 Phased Array UT inspection ...... 70

3.11 Summary ...... 73 4 Results and Discussion ...... 75

4.1 Introduction ...... 75 4.2 Physical appearance of the welds ...... 75 4.3 Macrostructural Characterization ...... 77 4.4 Microstructural Examination ...... 78 4.5 Tensile Properties ...... 91 4.6 Microhardness profiling ...... 95 4.7 WELD DEFECTS ...... 98 4.7.1. Liquid Dye Penetrant Weld Defect Analysis ...... 98

4.7.2. Ultrasonic Phased Array Weld Defect Analysis ...... 100

4.7.3. Ultrasonic ToFD Weld Defect Analysis ...... 105

4.7.4. Digital X-ray Radiographic Weld Defects Analysis ...... 110

4.7.5. Ultrasonic Testing pulse-echo Couplant Attenuation Effect ...... 113

4.8 Summary ...... 118 5 Conclusion and Future Work ...... 120

5.1 Introduction ...... 120 5.2 Conclusion ...... 120 5.3 Future Work ...... 121 References ...... 123

7. ABBREVIATIONS

AA – Aluminium Alloy

AFM – Atomic Force Geometry

AGG – Abnormal Grain Growth

AL – Aluminium

AS – Advancing Side

ASTM – American Society for Testing and Materials

BM – Base Material

CFSWC – China Friction Stir Welding Center

CPU – Central Processing Unit

EAM – Embedded-Atom Method

EDM – Electric Discharge Machine

EMMS – Energy Minimisation Multi-Scale

FCC – Face-Centered Cubic

FSP – Friction Stir Processing

FSPd – Friction Stir Processed

FSSW – Friction Stir Spot Welding

FSW – Friction Stir Welding

FPZ – Friction Processed Zone x

GNU – General Public License

HAZ – Heat Affected Zone

HDS – High-Density Steel

HV – Vickers Hardness

IISC – Indian Institute of Science

LAMMPS – Large Scale Atomic/Molecular Massively Parallel

Simulator

MD – Molecular Dynamics

MIG – Metal Inert Gas

MNM – Micro/Nano Machining

NASA – National Aeronautics and Space Administration

NGG – Normal Grain Growth

NZ – Nugget Zone

OM – Optical Microscope

OVITO – Open Visualisation Tool

PM – Parent Material

RAM – Random Access Memory

RS – Retreating Side

SEM – Scanning Electron Microscope

SZ – Stir Zone

TMAZ – Thermo-Mechanically Affected Zone

TWB – Tailored Welded Blanks

UTS – Ultimate Tensile Strength

xii

8. NOMENCLATURE

퐹푥 – Advancing force in the direction of welding (N, eV/Angs)

퐹푦 – Uniaxial force perpendicular to the Fx during the welding process

(N, eV/Angs)

퐹푧 – Vertical downward force on the tool (N, eV/Angs)

T – Temperature (K).

P – Pressure (Nm)

N – Number of Particles

μ – External chemical potential t – Time (Secs)

E – Energy (eV)

xiii

9. UNITS

μm – micrometer mm – millimeter mm/min – millimeter/minute in/min – inch/minute g – gram secs – seconds

N – newton kN – kilo newton

Nm – newton meter

MPa – mega pascal rpm – revolutions per minute kgf – kilogram-force eV – electron Volt

Angs – Angstrom dB/m - decibel per metre

Hz - Hertz

xiv

10. LIST OF FIGURES

Figure 1:1: A Schematic of friction stir welded aluminium specimen being inspected using UT Technique [10] ...... 3

Figure 1:2: Research study process flow ...... 8

Figure 2:1: Fusion welding performances of aluminium to other metals [32] ...... 13

Figure 2:2: Schematic of FSW tool movement on workpiece [36] ...... 14

Figure 2:3: Schematic representation of FSW process [39] ...... 15

Figure 2:4 : Fast Ferries Decking Rolled for transport [47]...... 18

Figure 2:5: Schematic of FSW Tool ...... 20

Figure 2:6: Typical FSW tool shoulder outer surfaces [59] ...... 21

Figure 2:7: Basic FSW/FSP tool profiles [64] ...... 23

Figure 2:8: Advanced FSW/FSP tools developed at TWI [65] ...... 23

Figure 2:9: Illustration of the FSW tool tilt angle and plunge depth [69] ...... 24

Figure 2:10 : Forces acting on to the tool and workpiece [82] ...... 27

Figure 2:11: Simulations of forces acting during FSW process stages [82]...... 28

Figure 2:12: Movement of FSW Tool plunged into material [94] ...... 30

Figure 2:13 : Typical Macrograph of the FSW various microstructure zones ...... 31

Figure 2:14: Typical microstructures at different regions of FSW AA6061 after etching with Keller’s reagent: (a) HAZ, (b) TMAZ, (c) BM, and (d) NZ [95] ...... 31

Figure 2:15: Graphical illustration of the microstructural evolution [97] ...... 32

Figure 2:16: Effects of nucleation site density [132] ...... 35

Figure 2:17: Hardness values for FSW AA6056-T4 [132] ...... 37

Figure 2:18 : Micro hardness distribution of a significant test [138]...... 38

Figure 2:19 : Engineering stress/strain curves for the AA 6016-T4 base material and for HW and CW weld samples [141]...... 39

Figure 2:20 : FSW typical defects...... 40

Figure 2:21: Boundaries wave propagation [154]...... 43

Figure 2:22 : Sound waves incident angles conversions [155]...... 44

Figure 2:23: Phased Array online scanning of weld using 60º probe [153]...... 46

Figure 2:24: Multiple Image displays [153]...... 46

Figure 2:25: POD of X-ray, UPA and 95% benchmark [156]...... 47

Figure 3:1: FSW platform ...... 53

Figure 3:2: Layout of Specimens ...... 56

Figure 3:3: Mecatome T300 Cut off discharge machine and typical cut specimen ... 57

Figure 3:4: Schematics of samples for microstructural and microhardness tests ..... 58

Figure 3:5: Struers hot mounting machine ...... 58

Figure 3:6: Struers polishing machine ...... 60

Figure 3:7: Setup of the Optical microscopy ...... 60

Figure 3:8 : TESCAN VEGA3 scanning electron microscope ...... 61

Figure 3:9 : ZwickRoell ZHµ Vickers microhardness tester ...... 62

Figure 3:10 : Instron tensile testing machine ...... 63

Figure 3:11: Schematics of tensile samples (All dimensions in mm) ...... 63

Figure 3:12: X-ray digital radiography inspection equipment set-up ...... 65

Figure 3:13: Figure 3.13: Epoch 600 UT equipment with probe without a membrane ...... 67

Figure 3:14: UT pulse-echo equipment set-up showing probe with a membrane ..... 68

Figure 3:15: UT ToFD Handyscan equipment with probe emitting shear wave mode inspection on FSW samples ...... 70

Figure 3:16: UT equipment showing ToFD image of the sample ...... 70

Figure 3:17: UT Phased Array inspection on FSW samples ...... 71

Figure 3:18: UT Phased Array probe longitudinal traverse directions relative to the weld ...... 72

Figure 3:19: UT Phased Array probe set-up ...... 73 xvi

Figure 4:1: Typical Macrograph images of FSW samples ...... 78

Figure 4:2 – AA6061-T6 Base Metal Microstructure; (a) OM micrograph (b) SEM .. 79

Figure 4:3 – Optical Microstructure micrographs of the welds produced at 700rpm and 80 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget ...... 80

Figure 4:4 – SEM Microstructure of the welds produced at 700rpm and 80 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget ...... 81

Figure 4:5 – SEM Microstructure of the welds produced at 700rpm and 60 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget ...... 82

Figure 4:6 – SEM Microstructure of the welds produced at 700rpm and 100 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget ...... 83

Figure 4:7 – SEM Microstructure of the welds produced at 900rpm and 60 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget ...... 84

Figure 4:8 – SEM Microstructure of the welds produced at 900rpm and 100 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget ...... 85

Figure 4:9 – SEM Microstructure of the welds produced at 1100rpm and 60 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget ...... 86

Figure 4:10 – SEM Microstructure of the welds produced at 1100rpm and 80 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget ...... 87

Figure 4:11 – SEM Microstructure of the welds produced at 1100rpm and 100 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget ...... 88

Figure 4:12 Load-Extension behaviour for the weld samples processed at different tool rotational and weld speeds ...... 93

Figure 4:13: Stress-strain curves obtained from tensile test results of the weld sample processed using tool rotational and weld speeds of 1100 rpm and 100mm/min...... 94

Figure 4:14: Effect of processing speeds on microhardness across the weld traverse...... 96

xvii

Figure 4:15 - UT Phased Array scan images for weld produced by 700 rpm and 60 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile ...... 101

Figure 4:16 - UT Phased Array scan images for weld produced by 700 rpm and 80 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile ...... 101

Figure 4:17 - UT Phased Array scan images for weld produced by 700 rpm and 100 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile ...... 102

Figure 4:18 - UT Phased Array scan images for weld produced by 900 rpm and 60 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile ...... 102

Figure 4:19 - UT Phased Array scan images for weld produced by 900 rpm and 80 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile ...... 103

Figure 4:20 - UT Phased Array scan images for weld produced by 900 rpm and 100 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile ...... 103

Figure 4:21 - UT Phased Array scan images for weld produced by 1100 rpm and 60 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile ...... 104

Figure 4:22: UT Phased Array scan images for weld produced by 1100 rpm and 80 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile ...... 104

Figure 4:23 - TOFD scan images obtained for the weld produced by 700rpm and 60 mm/min ...... 106

Figure 4:24 - TOFD scan images obtained for the weld produced by 700rpm and 80 mm/min ...... 107

Figure 4:25 - TOFD scan images obtained for the weld produced by 700rpm and 100 mm/min ...... 107

xviii

Figure 4:26 - TOFD scan images obtained for the weld produced by 900rpm and 60 mm/min ...... 108

Figure 4:27 -TOFD scan images obtained for the weld produced by 900rpm and 80 mm/min ...... 108

Figure 4:28 - TOFD scan images obtained for the weld produced by 900rpm and 100 mm/min ...... 109

Figure 4:29 - TOFD scan images obtained for the weld produced by 1100rpm and 60 mm/min ...... 109

Figure 4:30 - TOFD scan images obtained for the weld produced by 1100rpm and 80 mm/min ...... 110

Figure 4:31: TOFD scan images obtained for the weld produced by 1100rpm and 100 mm/min ...... 110

Figure 4:32 - UT frequency (y-axis) waveform in time (x-axis) domain scan done using Oil and Membrane attached to probe using and reference of 44.7 dB ...... 115

Figure 4:33 - UT frequency (y-axis) waveform in time (x-axis) domain scan done using Oil and no membrane attached at a reference of 44.7 and showed a frequency gain of – 11.4 dB ...... 115

Figure 4:34 - UT frequency (y-axis) waveform in time (x-axis) domain scan was done using Ultrasonic Gel with a membrane to the probe at a reference of 44.7 and showed frequency gain of – 5.6 dB ...... 116

Figure 4:35 - UT frequency (y-axis) waveform in time (x-axis) domain scan was done using Ultrasonic Gel with no membrane attached to the probe using a reference of 44.7 and obtained a frequency gain of –12.7 dB ...... 116

Figure 4:36 - UT frequency (y-axis) in time (x-axis) domain scan was done using Grease and a membrane attached to the probe using a reference of 44.7, results displayed a frequency gain of –1.5dB...... 117

Figure 4:37 - UT frequency (y-axis) waveform in time (x-axis) domain scan done using grease with no membrane attached to the probe using a reference of 44.7, results showed a frequency gain of –10.7dB ...... 117

xix

Figure 4:38 - UT frequency (y-axis) waveform in time (x-axis) domain scan done using Water with no membrane attached to the probe using a reference of 44.7, results showed a frequency gain of –14.4 dB ...... 118

11. LIST OF TABLES

Table 2.1: FSW key benefits [41] ...... 15

Table 2.2: Evolving properties obtained by localised modification in FSP [43] ...... 16

Table 2.3: Typical FSW Applications and their advantages [16] ...... 17

Table 2.4: Main process parameters in friction stir welding [52] ...... 19

Table 2.5: Grain size summary in nugget zone FSW aluminium alloys [41] ...... 32

Table 2.6: Couplants acoustic impedance values [151] ...... 42

Table 3.1: Chemical composition of AA6061-T6 ...... 52

Table 3.2: Mechanical Properties of the AA6061-T6 aluminum alloy ...... 52

Table 3.3: Process parameter matrix of the FSW process ...... 54

Table 3.4: Poly-fast Thermoplastic resin mounting process parameters ...... 59

Table 3.5: X-ray radiography exposure parameters for FSW joints ...... 66

Table 4.1: Physical appearances of the FSW samples ...... 76

Table 4.2: Average grain size in various region ...... 89

Table 4.3: Mechanical Properties and Weld efficiency obtained by tensile tests ...... 94

Table 4.4: Liquid dye penetrant tested weld joint photos...... 99

Table 4.5: X-ray radiograph scan images of the weld samples ...... 111

xxi

12. GLOSSARY OF TERMS

Advancing side – The advancing side is the side of the weld where the local direction of the rotating tool is in the same direction of traverse.

Alloy – A substance having metallic properties and being composed of two or more chemical elements, of which at least one is a metal.

Alloying element – An element added to, and remaining in metal, that changes its structure and properties.

Backing plate – A layer of material that is placed below the materials to be processed. It provides a surface to oppose the vertical downward force on the material, and protects the machine bed.

Clamping – Holding and preventing the workpiece from moving during the large forces involved in the FSP process.

Defect – A discontinuity or discontinuities that accumulate to render a part unable to meet minimum acceptance standards or criteria of the design specifications.

Deformation – Change in the form of a body due to stress, thermal, or other causes.

Ductility – The ability of a material to deform plastically before fracture.

xxii

Dwell time – The period after the rotating tool has been plunged into the work and for which it remains stationary, generating frictional heat and plasticising the materials, before commencing the traverse along the joint (seconds).

Elongation – The increase in gauge length of a body subjected to a tension force, referenced to a gauge length of a body. Usually expressed as a percentage of the original gauge length.

% Elongation – The total percent increase in the gauge length of a specimen after a tensile test.

Engineering strain – This is a dimensionless value that is the change in length (ΔL) per unit length of the original linear dimension (Lo) along the loading axis of the specimen; that is e = ΔL/Lo the amount that a material deforms per unit length in a tensile test.

Engineering stress – The normal stress, expressed in units of applied force, F, per unit of original cross-sectional area, Ao; that is, S = F/Ao.

Equilibrium – A state of dynamic balance between the opposing actions, reactions, or velocities of a reversible process.

Etchant – A chemical solution used to etch a metal to reveal structural details.

Etching – Subjecting the surface of a metal to preferential chemical or electrolytic attack to reveal structural details for metallographic examination.

Extrusion – The process where a material is shaped by force or squeezed through a die or nozzle.

Exit hole – A hole left at the end of the weld when the FSP tool is withdrawn, resulting from displacement of material during the plunge. Some special techniques are in-use to fill or prevent the occurrence of this hole.

xxiii

Face-centered cube – This is a crystal system where atoms are arranged at the corners and center of each cube face of the cell.

Filler Metal – Metal added in making a welded, brazed, or soldered joint.

Flash – A build-up of weld material, normally on the retreating side of the rotating tool, which has a ‘peel-like’ effect; this is termed side flash in FSW/FSP.

Force control – A mode in the Friction Stir Process in which a known force from previous welds is added to other input process parameters to produce a weld.

Fusion – The melting together of filler metal and base metal, or of base metal only, which results in coalescence.

Fusion welding – Any welding process that uses fusion of the base metal to make the weld.

Friction – The force required to cause one body in contact with another to begin to move.

Friction stir welding – A process developed at The Welding Institute (TWI) that utilises local friction heating to produce continuous solid-state seams. It allows butt and lap joints to be made without the use of filler metals. The solid-state low distortion welds produced are achieved with relatively low costs, using simple and energy- efficient mechanical equipment.

Friction stir processing – Friction Stir Processing (FSP) is a new solid state processing technique for microstructural modification. It is a method of changing the properties of a metal through intense, localised plastic deformation resulting in a significant evolution in the local microstructure.

Grain – An individual crystallite in metals. xxiv

Grain growth – This is a phenomenon which occurs when the temperature of a metal is raised, the grains begin to grow and their size may eventually exceed the original grain size.

Grain size – A measure of the areas or volumes of grains in a polycrystalline metal or alloy, usually expressed as an average when the individual sizes are uniform. Grain size is reported in terms of number of grains per unit area or volume, average diameter, or as a number derived from area measurements.

Grain boundary –An interface separating two grains, where the orientation of the lattice changes from that of one grain to that of the other. When the orientation change is very small, the boundary is sometimes referred to as a sub-boundary structure.

Grinding – Removing material from the surface of a workpiece by using a grinding wheel or abrasive grinding papers.

Hardness –A term used for describing the resistance of a material to plastic deformation.

Hardness test – A test to measures the resistance of a material to penetration by a sharp object.

Hardening – Increasing hardness by suitable treatment.

Heat affected zone - The portion of the base metal which has not been melted, but whose mechanical properties have been altered by the heat of welding or cutting.

Homogeneous – A chemical composition and physical state of any physical small portion, and that is the same as that of any other portion.

xxv

Indentation hardness – This is the hardness, as evaluated from the measurements of an area of an indentation made by pressing a specified indenter into the surface of a material under specified static loading conditions.

Joint – A point or edge where two or more pieces of metal or plastic are joined together.

Lap joint – A welded joint in which two overlapping metal parts are joined by means of a fillet, plug or slot weld.

Liquid penetrant testing – Is a non-destructive testing method that applies a dye in the sample to be analysed to detect flaws.

Macrograph – A graphic reproduction of a prepared surface of a specimen at a magnification not exceeding 25x.

Macrostructure – The structure of metals as revealed by macroscopic examination of the etched surface of a polished specimen.

Mechanical properties – The properties of a material that reveal its elastic or inelastic behaviour when force is applied, indicating the suitable mechanical applications.

Mechanical testing – The determination of mechanical properties.

Metallurgy – The science and technology of metals and their alloys including methods of extraction and use.

Microstructure – The structure of a prepared surface of a metal, as revealed by a microscope at a magnification.

xxvi

Non-destructive evaluation – Is a method used to measure or detect flaws into images using physical sciences without destroying the sample to be analysed.

Onion-ring like structure – A characteristic weld pattern featuring a cyclic ring or onion ring-like profile.

Parameter – The minimum and maximum parameters that will describe the operating range of a variable.

Parent material – This is the sheet-metal plate in its as manufactured form, as supplied.

Plastic deformation –This is the distortion of material continuously and permanently in any direction. The deformation that remains or will remain permanent after the release of the stress that caused it.

Plasticity– Capacity of a metal to deform non-elastically without rupturing.

Polished surface – A surface that reflects a large proportion of the incident light in a peculiar manner.

Position control – A mode in FSW in which the machine automatically adjusts the forces acting during the welding process.

Plunge depth – The plunge depth is the maximum depth that the tool shoulder penetrates the weld plates.

Plunge force – During the plunging stage of the tool pin in FSW, the vertical force in the direction of the Z-axis movement is normally referred to as the plunging force.

xxvii

Porosity – A rounded or elongated cavity formed by gas entrapment during cooldown or solidification.

Propagation – Is the growth of cracks and waves through materials

Radiography inspection – A non-destructive testing method that uses X-ray and isotopes to convert scanned information into images.

Recrystallization – A change from one crystal structure to another, such as that occurring upon heating and / or cooling through a critical temperature.

Residual stress – Stress in a body which is at rest, in equilibrium, and at uniform temperature in the absence of any external force.

Retreating side – The retreating side of the tool is where the local direction of the weld surface due to tool rotation and the direction of the traverse are in the opposite direction.

Rotational speed – The tool rotation speed is the rate of angular rotation (usually specified in rpm) of the tool about its rotational axis.

Spindle speed – This is also referred to as the rotational speed. It is the speed of the work holding device (chuck), measured in revolutions per minute.

Spindle torque – The spindle torque required to rotate the FSW tool when plunging into and traversing through the workpiece along the joint (Nm).

Stir zone – The recrystallised central area of the joint interface.

xxviii

Tensile strength – The maximum tensile stress which a material is capable of sustaining. Tensile strength is calculated from the maximum load during a tension test carried out to rupture, and the original cross-sectional area of the specimen.

Tensile test – This measures the response of a material to a slowly applied axial force. The yield strength, tensile strength, modulus of elasticity and ductility are obtained.

Tool shoulder – The part of the welding tool which rotates and is normally disk- shaped.

Tool pin – The part of the tool that rotates in contact with the surface of the workpiece.

Tool plunge – The process of forcing the tool into the material at the start of the weld.

Tool tilt angle – The angle at which the FSW tool is positioned relative to the workpiece surface; that is, zero tilt tools are positioned perpendicular to the workpiece surface (degrees).

Traverse speed – This is also referred to as feed rate; it is the speed at which the rotating FSW tool is translated along the joint line (mm/min).

Unaffected material – The bulk of material which is not affected by either heat or deformation during the welding process.

Ultrasonic Testing – A non-destructive method that uses high frequency acoustic waves to measure thicknesses and detect flaws in materials and weld joints.

xxix

Vickers hardness number – A number related to the applied load and the surface area of the permanent impression made by a square-based pyramid diamond indenter.

Void – The space that exist between particles or grains. Normally in welding, voids are associated with defects.

Welding – The process of joining, in which materials are enabled to form metallurgical bonds under the combined action of heat and pressure.

Workpiece – The component to be welded.

X x-axis – Relating to a specific axis (horizontal) or a fixed line determining the direction of movement or placement in a 2-Dimensional (2-D) or 3-Dimensional (3-D) co- ordinate system.

Y y-axis – Relating to a specific axis (perpendicular to x-axis) or a fixed line determining the direction of movement or placement in a 2-D or 3-D co-ordinate system.

Z z-axis – Relating to a specific axis (vertical) or a fixed line determining the direction of movement or placement in a 3-D co-ordinate system.

xxx

1. CHAPTER ONE - BACKGROUND

1.1 INTRODUCTION

Friction stir welding [1] is a novelty solid state welding technique that was invented at The Welding Institute (TWI) of UK in 1991 by TWI [2], and it was initially applied to aluminium alloys [3]. The welds are made by inserting/ plunging a rotating non- consumable FSW tool pin on the abutting edges of work-piece where the rotational speed causes friction to generate localized heat energy to achieve plastic deformation of the work-piece material, as the FSW tool traverses the frictional heat also traverses along the joint-line [4]. The predominantly concave shoulder with a pin attachment provides confinement for the heated volume of the material. The pin movement and geometry stirs the material facilitating the downward auger effect [5]. The fine microstructure generated, produces mechanical properties that are of high tensile strength and optimum fatigue and fracture resistant welds [6].

Like any new technology material flow complexities and weld processes and application can result in weld flaws. In the case of FSW, inappropriate processing conditions can cause formation of various visible or potential defects around weld nugget zone (WNZ), the thermo-mechanically affected zone (TMAZ), or sometimes at the WNZ/TMAZ interface [7]. This results in FSW weld-defect of different types, amongst them voids, defective tightness, surface groove, excessive flash, ‘kissing- bond’ [8] and crack-like root-flaw, which are quite different from conventional fusion welding flaws [9].

Some of these weld defects can be difficult to detect using visual inspection techniques or certain types of non-destructive evaluation techniques for the different types of weld defects. It is therefore imperative that correct selection of inspection method and its parameters are well defined in order to effectively evaluate the microstructure, defects and integrity of the friction stir weld.

Defects are inherent in most joint welding processes [10]. The formation of weld defects within the friction stir welds is mainly caused by the FSW process parameters. 1

Some of these process parameters that contribute to the formation of weld defects are mostly tool rotational speed [11], travel speed, tilt angle, shoulder target depth [12], applied load and pin configuration [13].

With recent developments in technology, weld joint integrity cab be evaluated without physical destruction of the weld by non-destruction. Non-destructive testing (NDT) relies on the application of physical principles to determine the characteristics of materials in detecting material flaws and critical defects [14]. The important characteristics of non-destructive testing are its ability to evaluate material and weld defects without changing the serviceability or usefulness of said materials that is, does not damage the material. NDT has great advantages which includes; cost effectiveness, high reliability and safer operation for a wide range of applications. Commercial use of NDT is widely spread in industrial fields, especially on railway, petrochemical pressure tanks, power generation, aviation, automotive and mining infrastructure inspections. Certainly, multiple NDT methods can also be applied in the quality control of FSW joints. NDT techniques cannot only make detection on a complete weld aiming to be suitable for use in its final application, but also favour real- time assessment of the weld as it is being done and eventually adjusting the weld parameters to correct any detected anomalies [15]; [16]. On the contrary, the weld bending test and the micro-optical observation by bending weld samples will both make the materials unusable, and they only give insight into the weld characteristics at the destroyed point where the fractured or image is obtained [17].

Quality control requires an effective and reliable non-destructive testing technique to detect small cracks and flaws in certain localities of the weld nugget material structure. It therefore becomes necessary to find the most effective NDT techniques for FSW because of its distinctly different characteristics of the welding zone and defects occurrence in welds [18]. Ultrasonic Testing (UT) method technique is the most commonly used commercially to evaluate weld flaws and defects. UT success in detecting weld flaws is highly dependent on effective transmission of the high acoustic waves from the transducer into the work-piece. Thus the reliability of UT depends not only on the intrinsic characteristic of the transducer probe, but on the way it is mounted on the work-piece under investigation. Figure 1.1 shows a typical transducer mounted on an aluminium work-piece to detect flaws in a FSW. It is therefore necessary to 2

correctly mount the transducer by ensuring that there is a good acoustic coupling between the probe and the surface in terms of the frequency content and amplitude.

The coupling of the transducer and the work-piece affects both the quality and quantity of data and a good acoustic coupling medium is essential in order to record good data. It is therefore imperative to investigate the couplant (coupling media) effect during the evaluation of friction stir welds.

Figure 1:1: A Schematic of friction stir welded aluminium specimen being inspected using UT Technique [10]

Friction stir welding is classified as a green technology [16]. FSW has been embraced by the industry and is fast becoming a well preferred welding technique over the conventional fusion welding processes due to its benefits in cost savings realized by a decrease in design, manufacturing, assembly and maintenance times, and improved corrosion performance by eliminating the fasteners as a source of dissimilar metal contact. In recent years the commonly used riveted joints in aircraft manufacturing has seen replacement of this fastening method with friction stir welded lap joints which has contributed to significant weight reduction. The friction stir weld has also been extended in many industries which has resulted in cost savings due weight savings due to the elimination of the fasteners [17].

Like any new technology, continuous improvements need to be achieved in defining the FSW process parameters that will assist in minimizing and eliminating a variety of 3

weld-defects. In an experimental investigation done, Rajakumara, et al., [19], conceded that establishing the most effective parameters on properties of friction stir welds as well as realizing their influence on the weld properties has become a major topic for researchers [20]-[22]. In their own findings they concluded that defect free fine grained microstructure and uniformly distributed finer particles [20] in the weld nugget are found to be the important factors responsible for the higher tensile strength and hardness of the weld. In recent years researches in non-destructive evaluation of FSW microstructures have made great strides in aiding not only to the quality control of the final product but to the monitoring of the process parameters during friction stir welding. These studies have made great strides in understanding the relationship between friction stir welding process parameters and the weld-defects formed in the heated affected regions of the weld microstructure.

Ultrasonic Testing (UT) uses simple wave propagation to detect and record defects without changing the material characteristics of the sample. The UT effectiveness and reliability depends on the physical contact between the transducer and the work piece. The transducer emits a wave that travels through the material that is reflected off its boundaries. An echo is produced as the wave hits the boundaries on reflection and the time elapsed is used to characterize the material using Time of Flight (TOF).

While defects are detected by monitoring TOF variations in the ultrasonic signal due to changes in the distance travelled by the wave, the TOF is also affected by the couplant viscous properties used to bond the transducer and the material. This is so because sound wave propagation speed in a solid is a function of acoustic impedance of the material. For this reason, poor flaws detection can result due to the absorption and scattering of waves during interface with different material boundaries.

1.2 PROBLEM STATEMENT

Most welds fail due to poor welding process parameters. To define correct welding specifications, it is important to have a strong understanding of the different types of defects that can lead to high probability of failures. It is therefore imperative to evaluate the microstructure inhomogeneity in the weld zone of a friction stir weld before the component are put in use for their respective applications.

Extensive research on NDT has been done to evaluate the microstructure of friction stir weld and their defects in the heated zones. However no studies have been conducted on the couplant effect of sound attenuation on friction stir weld joints.

It is therefore important to understand the couplant attenuation effects occurrence when ensuring some sort of bond exist between the probe and the work material. The sound reflection and material resistivity tends to affect the received quality of data that is used to characterize the material.

1.3 AIM

The aim of this study is to assess, evaluate and characterize the microstructural integrity within the weld zones of the friction stir welded AA6061-T6 butt joint and to establish the attenuation effect of sound waves during contact with ultrasonic testing to identify defects during the evaluation of the weld joint.

1.4 HYPOTHESIS

It is expected the material characterization of the selected FSW process parameter matrix in this study will establish the optimum process parameters to produce a quality weld on AA6061-T6 Aluminium Alloy. It is expected that correct optimum process parameters will result in improved material flow to produce refined grain structure to achieve greater mechanical properties. Furthermore, during FSW, it is envisaged that wrongly selected process parameters risks poor material flow, dynamic recrystallization, segregation and formation of brittle intermetallic phases resulting in poor welds and defects. Hence, this has been given the necessary attenttion to reduce any of the afore-mentioned development.

1.5 SCOPE OF RESEARCH

The scope of this research study covers the following:

 Identify the FSW process parameters that influence the outcome of a finished butt weld joint.

 Conduct an ultrasonic evaluation on a friction stir welded butt joint in order to identify defects and ascertain their positions relative to different heat affected regions and zones interface.  Investigate attenuation effect of sound wave propagation as a result of the couplant viscous properties during evaluation of FSW microstructure.  Conduct a weld integrity assessment by deriving the microstructure homogeneity of the welded joint aluminium alloy.

1.7 RESEARCH METHODS

FSW journal articles of previous weld integrity research studies were reviewed to assist in acquiring the behaviour and understanding of the FSW process, as well as to identify the knowledge gap on the critical areas to be investigated. A quantitative approach was adopted and metallographic sample pre-requisite training done to aid with laboratory experimental investigations.

A weld integrity assessment experimental investigation of the friction stir welded AA6016-T6 butt joint was conducted. Nine specimens with similar material characterization were prepared for etching to ensure uniformity of the surface. Three optimized categories of low, medium and high FSW process parameters were defined for the different specimens and samples marked in respect of their categories’ traverse speed, rotational speed and length of weld. A non-destructive testing evaluation of the weld along the joint to detect the flaws on the weld nugget, HAZ, TMAZ, BM regions and characterize the different flaw types in respect of their position in the weld zone.

The non-destructive evaluation experimental investigation of the different specimens and categories was to assist with generating set of statistical test data. Five different couplants of varying viscous properties with uniform thickness were applied on each specimen to ensure material boundaries were consistent for accurate results comparison. The UT non-destructive evaluation results of the experimental investigations formed the basis of the flaw detection and evaluation of the couplant (coupling medium) effects.

Microstructural analysis was conducted to ascertain the homogeneity and grain refinement in the friction stir welded alloy samples. The specimens were subjected to 6

a micro hardness test and tensile testing to determine their physical-mechanical properties and characterize the hardness of the different weld regions.

The summarized research methodology is as follows:

 Optimize different identified FSW process parameters using 9 samples with similar material properties constituency.  Material characterizations.  Visual inspection and geometrical properties profiling.  Detection of flaws implementing non-destructive evaluation techniques.  Microstructural characterizations.  Micro hardness profiling and tensile testing on the welds.

The summarised activities for this research study are demonstrated in Figure 1.2.

FSW Aluminium AA6061

Define FSW Process Parameters matrix & prepared samples

Experimental Investigation

Metallographical Qualitative Analysis Quantitative Analysis Analysis

Non-destructive Mechanical Microstructural Evaluation Properties Evaluation Evaluation Defects UT Scans, Microhardness [Hv] Micrographs & Radiography Scans, & Tensile Strength Macrographs Attenuation plots Tests

Optimum Process Conclusion

Figure 1:2: Research study process flow

1.8 DELIMITATIONS

This research study is limited to weld flaw detection using NDT focusing on contact UT to enable evaluation of the couplant effects during wave propagation. During assessment of the weld integrity, the research study was limited to the characterization profiling of microstructure, tensile strength and hardness. The study was also limited to the optimized FSW process parameters which were categorized in respect of rotational speed and traverse speed only. The following process parameters namely; plunge depth, dwell time, tilt angle were kept constant in this study. The study was further limited to one optimized size and type of concave shoulder and threaded pin with no other FSW tools considered.

1.9 SIGNIFICANCE OF THE RESEARCH

This research contributes to the academic research work currently done at the University of Johannesburg in as far as research undertakings relative to FSW of aluminium alloys is considered. The research will establish industrial interest in the application of FSW Aluminium AA6061 for mine structures conveyance fabrications within the South African mining industries. More-so in the field of man and material underground hoisting.

1.10 STUDY LAYOUT

Chapter One identifies the problem statement and explains the research aims and objectives. It also outlines the research hypothesis, the methods and technique used in the investigation, and defines the research battery limits or delimitations and clearly states the importance of this research project.

Chapter Two details the review of the related literature focusing on the FSW process, offering a brief overview of the defects detection using non-destructive evaluation techniques and highlighting the crucial role and significance of the essential process parameters in influencing the integrity of the weld joint.

Chapter Three describes the overall experimental set-up and preliminary investigation conducted. This includes tool designs, positions of the work pieces during welding,

optimization of the process parameters, resulting weld defects and the decision on the final weld matrix.

Chapter Four gives the investigation results and discussions of the research welds and;

Chapter Five concludes the observation benchmarking on the research FSW process parameters matrix and discusses possible future work and development.

2. CHAPTER TWO - LITERATURE REVIEW

2 INTRODUCTION

The purpose of this chapter is to review the factors affecting the FSW joint. The purpose of a welded joint is to allow two materials coalesced together to perform their application during their service life with no early dynamic failures. When two or more materials are metallurgically coalesced together to form a joint by an appropriate application of heat and/or pressure, the material joining process is known as ‘Welding’ [21, 22] . A filler material with mechanical properties matching that of the base metal can be used in fusion welding technique, commonly referred to as conventional welding. Alternatively, solid state techniques can be utilized more so with the advent advancement of Friction Stir Welding. FSW is now widespread commercially to join alloys achieving joint strength almost matching the base metal mechanical properties [23]. Thus, the two distinct types of welding techniques commercially available on large to small scale metal joining are the fusion and the solid-state welding [24, 25].

The review of related literature in this research work is focused on:

 The FSW process; Significance of the essential process parameters – tool design and geometry, process parameters, joint configuration and forces acting on the tool during FSW process.  FSW microstructure;  FSW microhardness;  FSW associated defects;  Non-destructive testing and defect analysis of the FSW joints;  Mechanical properties of friction stir welded joints;

2.1 WELDING

Welding is the solid bonding of materials to become one single piece. There are two different types of welding namely:-  Fusion welding 11

 Solid state welding

2.1.1 Fusion Welding

During fusion-welding, the application of heat is used to achieve molten state of the parent metals to be joined together and thereafter allowing it to solidify. To provide strength to the welded joint a filler metal can be applied to the molten pool during fusion welding process. Autogenous weld is also applicable where no filler metal is a not necessitate during the fusion operation. Different types of fusion welding are used for different applications and the most commonly used are categorized as Laser Beam Welding (LBW), Oxy-Fuel gas Welding (OFW), Resistance Welding (RW), Tungsten Inert Gas welding (TIG), Metal Inert Gas welding (MIG) amongst others. In most cases due to the solidification process transition from molten to solid state, thermal cyclic can result in changes to the joint mechanical properties amongst others tensile strength, ductility and fatigue strength [26]-[29].

Shivakumar, et al, [30], performed a microstructural evaluation on aluminium alloy fusion welded joints and concluded that the solidification process resulted in structure segregation, porosities as well as loss of alloying elements due to the varying thermal coefficients. In order to overcome these difficulties, a similar experimental evaluation was conducted using friction stir welded joints which showed FSW offered the best alternative through solid-state bonding.

Conventional welding has a variety of disadvantages amongst them:-

 The fumes and gases generated during welding can be harmful to the welder.  Slag and inclusions are quite common in fusion welding.  There is a requirement for high energy consumption.  Spatter and distortion of materials is quite dominant in most fusion welds.  Welding of dissimilar materials and complex shapes is quite difficult.

Akinlabi [31] observed that the molten state transition to solidification stage resulted in thermal cycles that has effects on material properties. The modification of material properties results in changes to the mechanical properties amongst the low tensile strength, ductility, toughness hardness and fatigue strength. While joining aluminium 12

to itself can be achievable the need to prepare the weld results it in distortion of mechanical properties and instances expensive to prepare welds. Increasingly aluminium is currently joined with disimilar materials as recently seen in marine, aerospace industries and automotive. It becomes quite challenging to apply fusion welding to dissimilar materials. Figure 2.1 illustrates the fusion welding performances of aluminium to other materials.

Figure 2:1: Fusion welding performances of aluminium to other metals [32]

The welding performance of aluminium alloys and dissimilar materials can be overcome by the application of solid state welding.

2.1.2 Solid-state Welding

The metallurgical joining of metals by application of heat below melting point and or application of pressure is referred to as Solid-state welding. Friction Stir Welding (FSW) amongst others such as cold pressure welding (CPW), Ultrasonic Welding (USW), electromagnetic pulse wedling (EPS) and Diffusion Welding (DFW) are some of the examples of solid-state welding. In all the solid-state welding processes no filler material is introduced to aid the mechanical properties of the weld [33].

The FSW process achieves metal coalescence at solidus temperature thus no melting occurs resulting in elimination of defects that would have otherwise occurred during the conventional fusion welding process.

2.2 FRICTION STIR WELDING

FSW is a continuous hot shear autogenous process where welds are made by inserting the non-consumable rotating probe on the abutting edges of work-piece. The rotational and traverse movements along the line joint cause friction energy to generate localized heating to achieve plastic deformation of the work-piece material 13

without reaching the melting point [34]. The usually concave shoulder attached to the pin provides confinement for the heated volume of the material. The pin movement stirs the material facilitating the downward auger effect and material flow, with the shoulder consolidating the plasticized material [35]. The fine microstructure generated produces mechanical properties that aid high strength, fatigue and fracture resistant welds. Figure 2.2 demonstrates the process definition of the tool and the work piece.

Figure 2:2: Schematic of FSW tool movement on workpiece [36]

In theory, FSW has four stages that consist of tool plunge, dwell time and weld and retraction. The rotating tool plunge (a), the rotating tool is forced into the material that are being joined until the shoulder makes contact with the work piece. Upon being plunged the next cycle is the dwell period (b); where the tool is rotated to generate frictional heat energy to cause the workpiece material to enter plastic deformation point without the tool being made to traverse along the weld joint line. The tool is then allowed to traverse (c) along the joint with a weld formed at the back of the tool. Once a full length weld is achieved the tool is the retracted (d) from the work piece and an exit hole remains at the point of retraction [37, 38]. The FSW process is illustrated in Figure 2.3.

Figure 2:3: Schematic representation of FSW process [39]

As a result of the tool action and influence on the work piece, when performed properly, a solid-state joint is produced. The various geometrical features on the tool may result in complex material movement around the pin with gradients in strain, temperature, and strain rate [40] However, the benefits of FSW in joining the aluminium alloys and other type alloys in different industrial applications have been realised as detailed in Table 2.1 [41]. More research is still being carried out to investigate and reduce the known limitations such as weld defects (wormholes, keyholes, cracks, tunnels, kiss- bond etc.) [42], and achieving a microstructure that is homogenous in the weld zone. This study will attempt to investigate further on the weld defects using non-destructive techniques and analyse the weld microstructure.

Table 2.1: FSW key benefits [41]

The above benefits have generated interest in researchers to study further on FSP modifications to achieve higher properties as detailed in Table 2.2 [43]. The research into the mechanical, electrical, fatigue and corrosion properties has improved the performance efficiency of various FSW components and generated wide commercial use in aviation, automotive, marine, pressure vessels. This application has resulted in reduced weights of components thereby reducing the cost of manufacture, use and maintenance.

Table 2.2: Evolving properties obtained by localised modification in FSP [43]

2.3 APPLICATION OF FSW

FSW application has been wide spread applications in commercial spaces such as automotive, marine, aviation, pressure vessels, sheet metalwork to mention a few. The automotive industry uses FSW for high volume production of components, for example, fuel tanks and light alloy wheels. Large satellite tanks from high strength aluminium alloys are now being made using FSW in aerospace. Commercial planes, light weight airframe structures for military are now widely produced using FSW applications [44, 45]. 16

Fuel tanks from difficult to weld aluminium alloys are now being made using FSW in USA. Boeing applied FSW to the inter-stage aluminium alloy modules of Delta II rocket and the first of these was launched in August 1999. FSW technology for the DSIV common booster core fuel tank increased the strength by 30% to 50% and lowered the cycle time by nearly 80%. A total of 2100m of defect free welds were processed with 60% cost saving and reducing manufacturing from 23 days to 6 days [46].

Airbus started showing interest in FSW since 1998 and invested several laboratory machines focusing on sheets and plates of aluminium alloys. 102 floor panels of the A400M military transport aircraft were made using FSW process. The FSW solution was selected because it fulfilled the lowest weight. Akinlabi, et al, [16] discussed FSW application listed Table 2.3.

Table 2.3: Typical FSW Applications and their advantages [16]

Industry Category Specific application Present process Advantages of using FSW Electrical Heat sinks – welded Gas Metal Arc Higher density of fins- laminations Welding (GMAW) better conductivity Electrical Cabinets and GMAW Reduced cost, weld enclosures through corrosion coatings Batteries Leads Solder Higher quality

Military Shipping pallets GMAW Reduced cost Extrusions Customized Not done today Can be customised to extrusions reduce need for large presses Boats and ship Keel, tanks and the Rivets & GMAW Stronger, less building hull distortions Golf cars and Chassis, suspension GMAW Less distortions, snowmobiles better fatigue properties Tanks, cylinders Fittings, Long & GMAW Higher quality – less circumferential seams leaks, higher uptime

Aerospace Floors, seams and Rivets Higher quality, fuselage cheaper (no rivets and holes) Automotive Wheel rims and GMAW, MIG Better joint integrity suspension arms Rail Industry Rail car body, window, GMAC Higher quality joints side wall and coupling gears

In 1995, Marine Aluminium, Hagesund, Norway initiated the application of friction stir welding of aluminium in the production of ship structures. This involved the FSW welding of 6 series aluminium extrusions to construct large panels for decking of fast ferries. One such example of ferry decking is shown in Figure 2.43. Since then, the use of friction stir welding has expanded into various other applications, with plenty involving joining of extrusions to form larger panels and more complex assemblies [47].

Figure 2:4 : Fast Ferries Decking Rolled for transport [47].

2.4 FRICTION STIR WELD PROCESS PARAMETERS

The FSW process parameters affects the mechanical and microstructure properties of the these joint primarily through heat generation and dissipation, so primary focus must be attributed to the effect of the welding process variables on heat generation and resultant outcomes. A lot of research work conducted has demonstrated that the

quality of the weld free of defect with sound mechanical and microstructure properties is attained as a direct result of efficiently controlled main process variables. Some of the main process variables [48]-[51] are as follows:

 Work piece material  Tool rotational speed  Welding traverse speed  Tilt angle  Shoulder diameter  Pin diameter and profile  Shoulder and pin material  Axial force

Some of the main process parameters and their effect in friction stir welding process are given in Table 2.4.

Table 2.4: Main process parameters in friction stir welding [52]

Parameter Effects Rotational Speed Frictional heat, oxide layer, breaking and mixing of materials Tilt angle The appearance of the weld, thinning and consolidation of plasticized material Welding speed Appearance Downward force Frictional heat, maintaining contact conditions

2.5 TOOL DESIGN AND GEOMETRY

The main goal of the FSW tool is to mix the material and consolidate the plasticized material in order to produce a quality weld. The tool serves three primary functions, namely: the heating of the work piece, the movement of material to produce the joint and the containment of the hot metal beneath the tool shoulder [53, 54].The mixing of material flow condition has a direct bearing on the quality of the weld that is driven by the dynamic stresses that occur at high temperature and abrasive wear of the material. As such the tool design and geometry must meet several crucial requirements to enable it to produce a quality weld. In addition, since FSW is a thermomechanical

process were the temperature of the tool approaches the base metal solidus temperature, it is desirable to select a tool material that does not lose dimensional stability, does not recrystallize during the welding and better still offers fracture resistance. Tool materials, besides having to meet the endurance of the welding process, affect friction coefficients and heat generation. Figure 2.4 shows a typical configuration of a tool consisting of the two major parts, a shoulder and a pin.

Figure 2:5: Schematic of FSW Tool

FSW tool design and configuration has a great effect of the joint size while the material selection and geometry also plays an important role in influencing heat generation, plastic flow, joint integrity, and the resultant mechanical and microstructure properties of the weld.

The role of the shoulder and the pin geometrical shape and size is important in the performance of their respective varying functions. In essence, the mixing and confinement of material has a correlation with tool size. The shoulder produces frictional heat to the surface and sub surface regions of the material while the pin produces the majority of heat in thicker materials. As such the shoulder diameter plays a vital role in generating the heat by application of a downward pressure on the workpiece and confining the plastic deformed material. Different types of shoulder profiles exist, namely, concave and convex. Different features can also be machined on the shoulder profile to aid with material deformation and mixing. Some of the typical shoulder features consist of scrolls, ridges or knurling, grooves and concentric circles that can be machined on any shoulder profile [55]-[58]. Figure 2.6 shows these different type of shoulder features which can influence the mechanical and microstructural properties and the quality of the weld.

Figure 2:6: Typical FSW tool shoulder outer surfaces [59]

Scialpi et al., [60], demonstrated the role of shoulder pattern. A study showed that the shoulder with a fillet and cavity configuration produced the best mechanical and microstructural properties demonstrating the influence of shoulder geometries on weld joints. This study was conducted on a friction stir welded joint of a 1.5 mm 6082 T6 aluminium alloy using three different shoulder geometries, namely, the scroll and fillet, cavity and fillet, and only fillet. The effect was analysed using optical macrograph, visual inspection, HV microhardness, bending test and longitudinal temperature tensile test. Due to its ability to provide the best crown by a combinations of the fillet and cavity that increased the longitudinal and transverse strength of the joint, the fillet and cavity geometry was considered the best tool geometry [60].

A study was conducted by Babu et al., [22] , on the influence of shoulder diameter to achieve defect free welds. FSW variant process parameters where applied using two different shoulder sizes of diameters 18 and 24 mm. Both shoulders had a similar 6mm diameter pin size that was used on a friction stir processed AZ31B magnesium alloy of thicknesses 6mm and 1.5mm respectively. Defects such as voids, tunnels and pin holes size were noticed on the processed zones when 18mm shoulder diameter was used for processing the 6mm thick workpiece. Increase in tool shoulder diameter to 24mm with varying process parameters produced defect free processed welds on the 6mm thick sample. Macroscopic evaluations also revealed the defects were reduced when the 18mm diameter shoulder was used to process the decreased thickness sample to 1.5mm. This study demonstrated that the tool shoulder diameter has an influence to refine and homogenize the grains at the selected region within the material [22].

The role of pin size in weld quality was illustrated in a study by Said et al., [61]. Mechanical properties were analysed for an aluminium A5083 of 5mm thickness that

had a lap joint FSW processed using a tapered cylindrical design pin tool with a variety of diameters (that is 5mm ,6mm, 7mm, 8mm and 9mm) and a fix 20 mm shoulder made of H13 Steel with varying rotational speeds (910 RPM, 1280 RPM and 1700 RPM). The evaluation showed that the optimum size of the tool was a 6mm diameter pin tool that produced the best weld in comparison to the other sizes. Similarly, Surekha and Akinlabi [62] also established that the pin length and size is determined by the thickness of the plate to be processed. It is important to note that the tool pin geometry plays an important role in influencing the material flow and thus the mechanical properties of the processed zone and selection of the pin tool size is relative to workpiece thickness [61].

Material flow is important in achieving a defect free weld. Ji et al., [63], performed a numerical analysis using ANSYS FLUENT software to demonstrate the effect of shoulder geometry and pin geometry on material flow behaviour. Three different types of shoulder geometries were studied, the plane-shoulder, the inner concave flute shoulder and the concentric flute shoulder. All tools showed effective mixing of material which was concluded as defect free. The geometrical influence of the different tools showed the flow velocity of material decreased with the increase of the distance away from the rotational axis of pin. The reduction of the cone angle and reduction of the screw groove achieved the increase in flow velocity of material inside weldment. The material flow direction was influenced by the rotational tool screw pin orientation indicating material near pin is downward whilst the flow direction near the TMAZ with different screw orientation varying the flow direction. It is important to note in this study the improvement of material flow was observed to be at best for the concentric-circles- flute shoulder compared to the other two. This illustrates that choosing the reasonable shoulder geometry and the pin geometry enhances material flow and mixing to avoid root defects during friction stir welding process. Some of the tool profiles are illustrated in Figure 2.7. Figure 2.8 shows some of the advanced tool profiles with features developed to improve material flow.

Figure 2:7: Basic FSW/FSP tool profiles [64]

Figure 2:8: Advanced FSW/FSP tools developed at TWI [65]

Studies in shoulder geometry effects on FSW has also been advanced in dissimilar materials. Akinlabi et al., [66], studied 3.175 mm thick dissimilar AA5754-C11000 FSW weld properties to investigate the effect of different shoulder diameters effect using sizes such as 15 mm, 18 mm and 25 mm. The 15mm tool was the optimum tool for the uniform material mixing between copper and aluminium with proper flow pattern of material. Research has also shown that that shoulder geometry strongly affects the material flow and the generation of intermetallic compounds for dissimilar copper aluminium FSW joint [67].

2.6 TOOL TILT AND PLUNGE DEPTH

Tool tilt is the angle to which the FSW processing tool is positioned relative to the surface of the work piece. At 0°, the tool will be positioned perpendicular to the surface work piece surface and can be increased to as much as 4° [68]. Tool tilt angle has an effect on the heat generation and the compressive force which effects influence on the 23

metallurgical bonding. Likewise, the position at which the tool is tilted or offset relative to the material has an effect on the stirring of materials from the front to the back of tool. Figure 2.9 illustrates a tilted tool.

Figure 2:9: Illustration of the FSW tool tilt angle and plunge depth [69]

Several research interests in the effects of tool tilt angle on the FSW process has been seen in recent years. Research investigations have focused on both similar and dissimilar joint materials. Kush et al., [70] conducted an FSW investigation on copper and AA6061-T651. Tool tilt angle were applied ranging from 0° and 4° varied in increamence of 1°. High mechanical properties were achieved at optimum 4° with a 181 HV nugget zone macro-hardness and 117MPa tensile strength. The tilt angles 2°, 3° and 4° were noticed to have produced defect free welds.

Recent developments has shown that with increase in tool tilt angle associated with lower feed rate and lower rotational speed, tensile strength of the FSW joint is reduced [71, 72]. Increasing the speeds achieved optimum results. Krishna et al., [73], investigated the effect of tool tilt angle on AA2014 aluminium by varying the tool angle between 0° and 3°, in increments of 0. 5°. Surface and internal defects were observed on lower tilt angles due to lack of filling. More so, the tool tilt angle has an effect on non-uniformity of the temperature distribution during FSW throughout the plate being welded. This was caused by the difference in temperature attained by regions in advancing side and retreating side. It was reported the higher temperature is attained in advancing side, than the retreating side. The optimum angle was observed at high

tool tilt angle of around 3° for welding AA 2014 aluminium alloy for the given value of feed 100 mm/min and speed 1000 rpm to get defect free welds.

Akinlabi [31], concurred in her review on effects of tool tilt angle of FSW A5005 aluminium and A1100 pure aluminium conducted by Shinoda [74]. Metallurgical observations revealed that the stir tool angle of tilt affects the metal flow patterns in two directions: bottom flow and surface flow. However, microhardness values across the weld revealed that the values were close to the base material. This was observed to have achieved a 95% efficient joint to base metal values.

2.7 TRANSVERSE AND ROTATIONAL SPEEDS

Optimum rotational speeds and traverse speeds provide control to the heat generation that is required to provide sufficient heat generation required for free flow of plasticized materials, adequate stirring and thereby assisting with a steady state cooling process. A FSW steady state process leads to sufficient recrystallization, finely bonded grain sizes with no defects and properly worked mechanical properties.

Celik et al., [75] investigated the optimum rotational speed for the 4 mm thick AA1050 aluminium alloy and pure copper friction stir welded sheets. Three different rotational speeds (630, 1330, 2440 rpm) and traverse speeds (20, 30, 50 mm/min) were implemented. The microstructure of the welds were examined using optical microscope and scanning electron microscope (SEM). The tensile test and bending test showed that the rotation speed of 1330 rpm at 20mm/min produced the maximum efficient weld at 89.5 and homogeneous mixture was obtained. At rotational speed of 2440 rpm with a traverse speed of 30 mm/min the maximum tensile strength of 92.91 MPa was obtained.

However increasing the traverse speed at the same rotational speed of 50 mm/min caused the tensile strength to reduce. They reported that that the welded part tensile strength increases at higher rotational speeds at optimum traverse speeds and optimum tool positioning. Kumar et al., [76], performed an experimental investigation by implementing rotational speeds of 900, 1120 and 1400 with feed rates of 25mm/min, 40mm/min and 80 mm/min. They reported that the tensile strength was lowest at 900 rpm and reached maximum at 1120 rpm. Higher rotation speed of 1400 25

rpm yielded excessive temperatures and slower cooling rates which caused excessive grain growth resulting in lower tensile strength. The lower tensile strength at 900 rpm was attributed to lesser heat energy which resulted in lack of stirring leading to lower tensile strength. Likewise the rotational speed of 1120 rpm yielded higher tensile strength which was due to optimum heat generation which was adequate to yield free flow of plasticized material and sufficient mechanical working.

Similar observations were also reported when aluminium was joined with a different material. Tehyo et al., [77], investigated the FS welded microstructure and mechanical properties of dissimilar materials SSM356-T6 grains and AA6061-T651. Two different tool rotation speeds (1,750 and 2,000 rpm) and six welding speeds (20, 50, 80, 120, 160, and 200 mm/min), where used as the two FSW prime joining parameters. The higher tool rotational speed were reported to have affected the weaker material’s maximum tensile strength lesser compared to the lower rotational speed. Higher rotational speed produced greater tensile elongation. Maximum tensile strength of 206.3MPa was achieved at 2000 rpm and traverse speed of 80mm/min. An increase in the welding speed associated with various tool rotation speeds affected lesser the base material’s tensile strength up to an optimum value thereafter which its effect increased. Tensile elongation was generally greater at greater tool rotation speed.

2.8 TOOL AXIAL FORCE

It is important to understand the forces applied during FSW process as it gives significant insight into the design and limitations of the machine and tooling.

The non-consumable rotating tool is plunged into the material by exerting a downward acting force,퐹푧, which is required to maintain the position of the tool at surface and or submerged into the material. The downward force causes severe deformations to the workpiece upon plunging into material. A dwell time period elapses upon plunging into the workpiece to generate optimum frictional heat to plastically deform the material local to the pin. Once optimum heat has been generated around tool and localized material, the rotating tool traverses along the joint line exerting a horizontal traverse force,퐹푥, at a controlled torque. The downward vertical force,퐹푧, is ramped up until an 26

positive opposing force acting towards the advancing tool, 퐹푦, acts to stabilize the forces to an equilibrium state to achieve a steady state FSW process [78] - [81]. Figure 2.10 shows a simplified illustration of the forces.

Figure 2:10 : Forces acting on to the tool and workpiece [82]

Figure 2.11 demonstrates the force simulations in time domain acting on the workpiece. A ramp up in the vertical forces is observed as the rotating tool is plunged into the workpiece. Upon achieving optimum temperature during the dwell time and plasticizing the local material, the forces steadies after the translation stage allowing the rotating tool to traverse along the workpiece at a steady state parameters to produce the weld. Trimble et al., [82] conducted a finite element method simulation on the FSW process to model forces acting on an AA2024-T3 weld joint. They demonstrated that maximum forces are experienced during the plunge stage reducing significantly by almost 35% during the translational stage. An experimental investigation yielded similar pattern. As such, the tool pin is most likely to be damaged during the plunge and shoulder damages most likely occurring during the process of translational stage.

Figure 2:11: Simulations of forces acting during FSW process stages [82]

Numerous FSW process studies have been conducted to investigate the effect of tool geometry on resultant force application. Researchers have reported that correct tool geometry choices influence the resultant forces that yields steady state processes with optimum recrystallization mechanisms, dynamic nucleation of grains to enable adequate plasticity, complete stirring and effective grain bonds [83, 84].

2.9 MATERIAL FLOW IN FSW

Material flow behaviour is a key factor in FSW process that influences the FSW joint quality. Different experimental techniques have been applied in the study of FSW material flow, for example, the stop-action technology [85], the marker material method [86], steel ball tracing technology [87] and the metallography method amongst others. However most of these investigative techniques have not been conclusive in determining the complex pattern of material flow.

Reynold et al., [88], used the marker insert technique to investigate material flow in FS butt-welded AA2195-T8. He reported that the technique provided a semi-quantitative, 3D view of the material transport in the welded zone. They observed that the material movement was not symmetrical about the weld centreline noting that the bulk of the material was transported to a position behind its original position. They concluded that the circulation increased with increasing weld energy.

Numerical analysis techniques have recently been applied to better understand the material flow in the stir region [89, 90] Chen et al., [91], formulated a thermomechanical finite element method derived from the arbitrary Lagrangian–Eulerian (ALE) implementing adaptive re-meshing technique to produce a 3D model to simulate FSW material flow characterising non-linear continuum mechanics behaviour. They reported that the plastic strain distribution correlated well with the weld microstructure zones distribution. It was observed the input shoulder axial forces influenced the variation to the equivalent plastic strain. This demonstrated that by increasing the torque and translational speeds material flow increased, observing that the swirl on the tool advancing side and the material flow in the swirl on the tool advancing side became faster with the relative increase of the translational speed.

Lee et al., [92], studied the FSW heterogeneous material of AA 5182-H111 and AA 6016-T4 aluminium 1 mm thick alloy sheets. They used two different types of tool shoulders namely the conical cavity shoulder and the scrolled shoulder. They observed the materials flow in the conical cavity shoulder was dominated by the Pin- driven flow that were characterized by an onion ring structure. Both the pin-driven and shoulder-driven flow was observed as being restricted to the crown of the weld from the retreating side of the tool, extending throughout the weld thickness, at the advancing side. The scrolled shoulder did not exhibit any onion ring structure and was characterised with extensive mixing of the parent materials that occurred in a plasticized layer flowing through the thickness around the rotating tool pin. It was observed that the shoulder-driven flow is the most intense and continuous behavioural around the tool.

In his research study, Oyindamola [93], reviewed that “the tool shoulder facilitates bulk material flow while the pin helps in the layer by layer material flow. The shape of the pin and shoulder has significant influence on the flow of the plasticised material. It has been reported that square tool geometry results in a more homogeneous distribution of particles than other tools. However, circular tools seem to show much less wear than flat-faced tools”.

2.10 FSW ALUMINIUM MICROSTRUCTURE

In FSW process, the welding zone is subjected to recrystallization of material during the intense plastic deformation stage, the solidification phase, effects of cyclic temperature variations behaviour and the growth of texture during the stirring of material caused by the tool movement motion illustrated in Fig 2.12. These process mechanisms results in varying granular patterns formation around the weld zone. The weld zones microstructure is also hugely influenced by the relative FSW process parameters yielding distinct weld zones from the point of tool plunge area to the furthest point where nominal heat and compressive forces causes nominal parent material modifications. Figure 2.13 shows the FSW microstructure characterized by four distinct zones with a typical microstructure grain structure shown in Figure 2.14. These four distinct microstructure weld zones are:-

 Base material (BM) also known as parent metal (PM)  Heat affected zone (HAZ)  Thermo-mechanically affected zone (TMAZ)  Weld nugget zone (NZ) also known as stir zone (SZ)

Figure 2:12: Movement of FSW Tool plunged into material [94]

Figure 2:13 : Typical Macrograph of the FSW various microstructure zones

Figure 2:14: Typical microstructures at different regions of FSW AA6061 after etching with Keller’s reagent: (a) HAZ, (b) TMAZ, (c) BM, and (d) NZ [95]

Figure 2.15 illustrates the microstructural evolution in various FSW microstructural zones. Static recovery, recrystallization mechanisms, dynamic nucleation, precipitate dissolution and coarsening of the grain structure are the main processes that drives formation of the different microstructural zones [96, 97]. The texture of the grains influences various properties such as the strength, ductility formability and corrosion resistance. Each of these zones has distinct thermo-mechanical history and the texture information is the used to establish the grain boundaries commonly identified using the orientation imaging microscopy [98].

Figure 2:15: Graphical illustration of the microstructural evolution [97]

2.10.1 Weld Nugget

The nugget zone is the stirred region that experiences intense plastic deformation and frictional heating which is heavily influenced by the FSW processing parameter, tool geometry, workpiece temperature and material thermal conductivity. The microstructure is characterized by recrystallized fine equiaxed grains that would have undergone dynamic recrystallization [99, 95]. Different nugget shapes have been observed in recent studies amongst them the basin-shaped [100], onion ring structure [101], elliptical [102] and rounded shape [103] . The nugget shape is greatly influenced by the tool geometry, tool rotation and traverse speed. In FSW process, special tool geometry can be used to influence the recrystallized grain size to ultra-fine grained microstructure. Ma, et al, [41] reported the grain size that were achieved in various studies.

Table 2.5: Grain size summary in nugget zone FSW aluminium alloys [41]

Material Plate Tool Geometry Rotatio Traverse Grain Ref thick nal Speed Size (mm) Rate (mm/min) (mm) (rpm) 7075Al-t6 6.35 - - 127 2 [104]

6061Al-T6 6.3 Cylindrical 300-100 90-150 10 [105]

Al-Li-Cu 7.6 - - - 9 [106]

7075Al-T651 6.35 Threaded, cylindrical 350,400 102,152 3.8-7.5 [107]

6063Al-T4, T5 4.0 - 360 800-2450 5.9-17.8 [108]

6013Al-T4, T6 4.0 - 1400 400-450 10-15 [109]

1100Al 6.0 Cylindrical 400 60 4 [110]

5054Al 6.0 - - - 6 [111]

1080Al-O 4.0 - - - 20 [112]

5083Al-O 6.0 - - - 4 [113]

2017Al-T6 3.0 Threaded, cylindrical 1250 60 9-10 [114]

2095Al 1.6 - 1000 126-152 1.6 [115]

Al-Cu-Mg-Ag- 4.0 - 850 75 5 [116] T6 2024Al-T351 6.0 - - 80 2-3 [117]

7010Al-T7651 6.35 - 180,450 95 1.7,6 [118]

7050-Al-T651 6.35 - 350 15 1-4

Al-4Mg-1Zr 10.0 Threaded, cylindrical 350 102 1.5

2024Al 6.35 Threaded, cylindrical 200-300 25.4 2.0-3.9 [119]

7475Al 6.35 - - - 2.2 [120]

5083Al 6.35 Threaded, cylindrical 400 25.4 6.0 [121]

2519Al-T87 25.4 - 275 101.6 2-12 [122]

Rhodes et al., [123], studied a joint on a 6.35 mm gauge 7075T6 Al plate which was friction stir welded, showing small variations of microhardness measurements across the joint from the base metal. The base metal exhibited an elongated matrix grain morphology whilst the weld nugget was characterized by concentric flow lines. Unlike the parent metal, the weld nugget had a recrystallized, fine equiaxed grain structure on the order of 2-4 µm in diameter. In contrast to the parent metal, the dislocation density in the weld nugget is quite low. Individual dislocations extended between particles, but no tangles. The transition zone between the parent metal and the weld

nugget is characterized by a highly deformed structure. TEM reveals that these grains have not recrystallized, as occurred in the weld nugget. The larger (50-75 nm) hardening precipitates appear not to have been significantly altered, either in size or morphology, during the joining process. The smaller precipitates, on the other hand, have coarsened during welding [123].

2.10.2 Thermo-mechanical affected zone (TMAZ)

The TMAZ undergoes both plastic and thermal deformations yielding a highly deformed structure. The grain size is influenced by the dissolution of some precipitates and subsequent grain growth. Akinlabi, [31], reported it is possible to obtain significant plastic strain without recrystallization with a distinct boundary between nugget recrystallized zone and the deformed TMAZ. The grain formation in the TMAZ is indicative of the strain, strain rates and the temperature conditions. The grain morphology is segregated containing high density refined and recovered grains. In most cases the advancing side of the tool is characterized with lower particle sizes and density [31].

Mcqueen et al., [124], investigated the microstructure of a FSW AA6065-T6 aluminium alloy plate. In the TMAZ, they reported bent and elongated grains of 350mm diameter with an aspect ratio rising from 3 to 11 on nearing the nugget zone. Higher temperatures caused dynamic recoveries which resulted in crystallites growth with some static recovery later. They also observed that there was no discontinuous recrystallization as characterized by cells of elongated grains with alternating layers of crystallites. The macrograph showed no evidence of growth by boundary migration from lower density nugget layers to more TMAZ cell layers that were deformed. Likewise, the TMAZ zone on the tool advancing side showed large fraction of high- angle boundaries attributed to both shear strain and temperature increase as the stirred material approached the tool pin. The retreating side showed wider density as the material ahead of pin was displaced due to tool rotation. Near the weld nugget in the TMAZ region, a mixture of fine recrystallized material and elongated grains was reported [124].

2.10.3 Heat affected zone (HAZ)

HAZ region is associated with thermal cycles which causes material modification without necessarily undergoing any form of plastic deformation. Studies have shown that FSW process has nominal effect on the subgrains size within the HAZ [125, 126]. Significant coarsening of precipitates is observed in this region and the widening of PFZ’s [127]-[129]. Predominantly the grain structure in the HAZ and the parental material has little difference and is only identifiable by their respective hardness properties [130, 131].

Lee et al., [131], used wide range friction stir weld conditions to achieve sound successful 220MPa relatively 85% of the tensile strength of the base metal on the joint using an aged hardenable 6005 Al alloy to evaluate the friction stir welded joint microstructural change related with the hardness profile. Fine and equiaxed grains in the stir zone (SZ) were created due to the frictional heat and plasticity, while in the TMAZ the recovered grains were elongated. The grain structure in the HAZ and the PM had no difference and were only identifiable by their respective hardness. During friction stir welding a softened region characterized by the dissolution and coarsening of the strengthening precipitate was formed near the weld zone [131].

Denquin et al., [132], investigated the microstructure of FSW AA6056-T4 aluminium alloy and modelled the effects nucleation density as shown in Figure 2.16. They observed the weakest point was the interface between TMAZ and HAZ. This was due to coarse and heterogeneous precipitation predominantly at this weakest point [132],

Figure 2:16: Effects of nucleation site density [132] 35

2.11 MECHANICAL PROPERTIES OF FSW JOINTS

In FSW, the base metal’s yield strength, hardness and ductility influence the plastic flow of the stirred material under action of the rotating non-consumable tool [133] In relation, the modified tensile strength, grain structure and microhardness are also influenced by the dynamic recrystallization, deformation, precipitation nucleation, thermal cycles and solidification effects during FSW process. As such the tool geometry and process parameters have a great bearing on the modified mechanical properties values of the weld joint. This section will discuss the essential FSW joint mechanical properties.

2.11.1 Microhardness

To investigate the modifications of mechanical properties encountered during FSW, measuring hardness values across the traverse of the weld joint aid in evaluating the mechanical properties variations from the initial base metal. Figure 2.17 shows typical hardness profile of a FSW AA6056-T4. It can be seen that the weld nugget values almost matches that of the base metal. However, the lowest hardness values are on the interface between the TMAZ and the HAZ. This can be attributed to the absence of solid solution in the weakest zone due to the effect of dynamic recrystallization and the effect of dynamic recovery processes [132].

Figure 2:17: Hardness values for FSW AA6056-T4 [132]

Zhao et al., [134], performed a numerical analysis using finite element method to predict the microhardness of a steady-state friction stir welding of 2014 Al alloy using Eulerian formulation. The model considered heat transfer and coupled visco-plastic flow near the tool. They reported that the hardness is essentially related to grain size through the Hall–Petch relationship in the stir zone. They validated the results by performing an experimental investigation whose results yielded similar hardness values as the numerical analysis [134].

Filippis et al., [135], developed an Artificial Neural Networks (ANNs) simulation model to analyse the correlation between the process parameters and the mechanical properties of the welded AA5754 H111 aluminium plates. Indentation Vickers hardness tests were simulated using a back-propagation neural network algorithm. The statistical hardness data curve showed a W-shape as illustrated in Figure 2.18. For all the samples the average weld nugget hardness was similar to that of the base material, was similar among all the samples. They reported that the hardness profile was greatly influenced by the precipitate distribution and only slightly by the grain and dislocation structures [135].

Figure 2:18 : Micro hardness distribution of a significant test [138].

Panella et al., [136], studied the effects of process parameters on mechanical properties using FSW AA6056 sample. The process matrix was based on rotational speeds of 500, 800, 1000 rev/min and traverse speeds of 50, 56 and 80 mm/min. Using microhardness plots in distance domain from the centre of the weld, they reported very uniform hardness for the material joined using the lowest rotating speed (500 rpm) and the lowest welding speeds, particularly for 40 and 56 mm/min. Higher rotating and traverse speeds produced higher material hardness values in all the conditions with hardness profiles becoming less uniform across the weld centre.

Likewise, it is therefore imperative in this research study to quantitatively analyse the weld hardness values across the weld microstructure for the different process parameters of rotating speeds 700, 900, 1100 rev/min and traverse speeds of 60, 80 100 mm/min with all other remaining process parameters having constant optimized value.

2.11.2 Tensile Properties

Tensile strength evaluations are critical to establish the ultimate strength and ductility of FSW joints which in essence is important to determine the weld integrity. The modified tensile strength values in the weld zones can then be compared to that of the 38

base metals. Variations in tensile strength are mostly attributed to recrystallization and reversion during the FSW process stages. It is quite possible that processed welds can be improved by application FSP post welding to have tensile strength and elongation values higher than that of the base metals [137].

Tensile tests can eventually be used to generate the stress–strain curves on the basis of engineering values calculated from the weld sample geometry. Vilaca et al., [138], produced an S-curve, as shown in Fig 2.19, using data obtained in the tensile tests of base material (AA 6016-T4) and transverse weld samples produced from a conical shoulder tool with a 8° inclination cavity and 10 mm in diameter (HW-1800 rpm; 80mm/min) and a scrolled shoulder with 14 mm in diameter (CW -320mm/min; 1120rpm). They reported that the yield stress values matched for both weld samples and for the base material and that the weld samples ductility was less than that of the base material.

Figure 2:19 : Engineering stress/strain curves for the AA 6016-T4 base material and for HW and CW weld samples [141].

They noted that the decrease in ductility of the CW weld samples relative to the base material was due to severe hardness under-match for these welds while the cross section variation caused loss of ductility in HW welds. They concluded that the HW and CW welds had, respectively, 30% and 70% lower ductility than the base material.

2.12 DEFECTS

Studies have shown that defect are caused by poor combinations of process parameters. It is therefore paramount to make a proper choice of the tool configuration, rotational speed, traverse speed, angle of tool tilt, plunge depth to suit the material specification and thickness [139, 140].

In FSW, inappropriate processing conditions can cause formation of various visible or potential defects around weld nugget zone (WNZ), the thermo-mechanically affected zone (TMAZ), or sometimes at the WNZ/TMAZ interface [141]. This results in FSW weld-defect of different types, amongst them pin hole, honey-comb, tunnel, void, porosity, defective tightness, surface groove, excessive flash, ‘Z-curve’, ‘kissing-bond’ and crack-like root-flaw [142, 143]. Some of the common defects associated with inappropriate FSW process parameters are shown in Figure 2.20 below.

Figure 2:20 : FSW typical defects. 40

2.13 WAVE PROPAGATION AND ATTENUATION THEORY IN ULTRASONIC TESTING

Ultrasonic testing (UT) is the most commercial commonly used technique to evaluate weld flaws and analyse defects. The application principle of wave propagation is used to detect and record defects without changing the material characteristics of the sample. The technique effectiveness and reliability depends on the physical contact between the transducer and the work piece. The removal of air from the interface between a measurement surface and transducer is crucial to the transmission of ultrasonic energy. Theobald et al., [144], reported that that use of a couplant greatly improves this transmission by around 2 times at 100 kHz and more than 10 times at 500 kHz; a typical gel-based couplant having a high acoustic impedance around 4 orders of magnitude higher than that of air. In his discussion paper, Larson, [145], established ten criterion in selecting the correct couplant specific to an application which are critical to both ensure you do no harm to the environment and to obtain optimal results. The criteria are acoustic impendence, temperature range, viscocity, toxicity, environmental sensitivity, surface wetting, dry time or evaporation rate, couplant removal and homogenous reliability.

A thin film of couplant is applied on the workpiece surface and should be constant throughout. This thin layer can cause an error in the measurement due to attenuation effects. Couplants of different forms, for example, oil, grease, pastes and many more have specific attenuation effects [146]. These couplants maybe liquid, semi-liquid or a paste. Very few studies have been done to demonstrate the effect of couplants in non- destructive testing [147].

2.13.1 Sound waves behaviour at boundaries

The higher the amount of sound energy that is absorbed by the material under investigation, the better the quality of the images that are resultantly obtained. The acoustic impedance of the boundary materials is critical to the intensity of energy reflected and transmitted through the media involved. Materials have different acoustic impedance that causes resistance to the propagation of sound waves. The acoustic impedance is mathematically given by Equation 1:-

푍 = 𝜌푐 (1) 41

Where, 𝜌 is the density of the material and c is its acoustic velocity. It will be important in this study to select couplants of varying acoustic impedances to effectively demonstrate the attenuation effects. Table 2.7 shows the selected couplants and their properties, to be used in this study.

Table 2.6: Couplants acoustic impedance values [151]

Couplant Acoustic Sound Velocity (c) Impedance (Z) (푚/푠)

푘푔 ( ) ∗ 106 푚2푠 Oil 1.28 1460 Ultrasonic gel - 2170 Water 1.49 1480 Grease - 1160

At the media boundary interface, the acoustic impedances of the material determines what amount of the incident sound wave will be reflected or transmitted into the second medium. In a research investigation on couplants effects, Drury [148], demonstrated that acoustic impedance is a critical factor in determining the reflection and transmission coefficients in a material under ultrasonic testing.

As the sound waves makes contact with the media, they are either reflected back or transmitted into the material. The reflection and transmission coefficients determines the amount of the energy to be reflected or absorbed at the interface. Kinsler et al., [149], expressed the reflection and transmission coefficients mathematically in Equations 2 and 3 below.

2 푍1−푍2 푃푟 = ( ) (2) 푍1+푍2

2 푍1−푍2 푃푡 = ( ) (3) 푍1+푍2

 Where 푃푟 and 푃푡 stands for reflection and transmission coefficient respectively

 푍1 is the impedance of the material in which the incident and reflected waves are travelling. 42

 푍2 is the impedance of the material in which the transmitted wave is propagating.

Figure 2.21 illustrates the wave propagating in different materials.

Figure 2:21: Boundaries wave propagation [154].

Refraction of waves is likely to occur when sound waves incident angles are influenced when a travelling wave encounters a boundary with a different material at an angle other than 0º. The wave energy transmitted into the next material is refracted in accordance with Snell’ law. Snell’s law relates the sines on the incident and refracted angle to the wave of velocity in each material as given mathematically in Equation 4 and illustrated in Figure 2.22.

sin 휃푖 sin 휃푟푙 sin 휃푟푠 = = (4) 푐푖 푐푟푙 푐푟푠

Where :-

휃푖 = incident angle of the wedge

휃푟푙 = angle of the refracted longitudinal wave

휃푟푠 = angle of the refracted longitudinal wave

푐푖 = velocity of the incident material (longitudinal)

푐푟푙 = material sound of velocity (longitudinal)

푐푟푠 = velocity of the test material (shear)

Figure 2:22 : Sound waves incident angles conversions [155].

As the acoustic waves propagate through a material, the travel distance affects their intensities thus reducing the amplitude of the echoes from the back wall. This effect is due to scattering and absorption of the wave. The phenomenon of reduction in amplitude due to scattering and absorption is referred to as attenuation. The resultant amplitude, A, of the original amplitude, 퐴표, that would have travelled a distance y is determined by using Equation 4 given below.

−∝푦 퐴 = 퐴표푒 (5)

Where α stands for attenuation coefficient of the material.

Attenuation coefficient in contact ultrasonic testing is given by the ratio of the first back echo and the second back echo wall and is referred to as apparent attenuation coefficient. The attenuation coefficient is expressed mathematically as given in Equation 5;

퐴 푚 = 20푙표푔 표 퐴 (6)

In pulse echo, a certain proportion of the acoustic wave is picked up by the transducer when they are reflected by the material due to the pulse trip. The apparent attenuation coefficient will then be given by Equation 6.

푚 ∝= 푑 (7)

Where, d is the thickness of the material.

2.13.2 Phased Array Ultrasonic Testing

Young et al., [150], used two point source light to create interference patterns and derived the principle of constructive and destructive interaction of waves. They observed that waves that combine in phase reinforce each other, while the waves that combine out of phase cancel each other. The phased array technique was largely confined in the medical field, due to the predictable nature of the human anatomy which made interpretation quite easy. Applications to the Industrial field was limited by the varying metals acoustic properties, complex geometries and varying thicknesses [151].

The phased array technique allows for inspections using linear array, 2D matrix, annular array, circular matrix and more complex shapes. The sophisticated integration with computer based instrument enables the phased array system to sweep sound beams through a range of refracted angles or along a linear path, or dynamically focus at a number of different depth increasing both flexibility and capability in inspection setups [152].

Phased array imaging provides real-time visualisation of volumetric data through the electronic scanning system and increase the probability of detection. Phased array scanning as illustrated in Figure 2.23, provides the ability to display multiple image types and store raw waveform information for the entire inspection allowing post- scanning analysis of the inspections results. Post analysis can then enables reconstruction of sectorial scans, C-scans and or B-scans with corresponding A-scan information at any inspection locations. Figure 2.24 shows a simultaneous imaging of the A-scan, the sector scan and the planar C-scan of the weld profile [153].

Figure 2:23: Phased Array online scanning of weld using 60º probe [153].

Figure 2:24: Multiple Image displays [153].

Britos, [154], performed physical non-destructive testing on friction stir welded AA2024-T351 samples, using X-Ray radiographic and phased array ultrasonic (UPA) testing techniques to collect data used to determine the probability of detection (POD). The experimental results data were input into a Gaussian (normal distribution) algorithm to determine the probability of detection as a function of the defect 46

discontinuity size. He observed that the phased array ultrasonic technique had a POD of above 90% as shown in Figure 2.24.

Figure 2:25: POD of X-ray, UPA and 95% benchmark [156].

Tabatabaeipour et al., [155], characterized the weld defects of friction stir welded AA7XXX aluminium alloy samples. They used an oblique incident C-scan in pulse echo mode to improve on the defect sizes detection. They concluded the C-scan image displaying the backscatter response is efficient to inspect nugget zone and obtain the root surface flaw status of the weld.

Bird et al., [156], analyzed the weld defect on friction stir welded 7075-T351 aluminum alloy with deliberate incorrect FSW process and defects to investigate phased array and eddy current applications on thin aluminum sheets. The samples containing different types of defects including entrapped oxide were successfully inspected. They concluded that the phased array technique demonstrated that this non-destructive testing technique provides accurate indication of the presence of very tight kissing bonds [156].

In alloys such as copper, phased array inspections are more reliable on detection of non-volumetric flaws, and more accurate than X-ray radiography inspection. Discontinuities in joint weld line are also more detectable using ultrasound, while radiography are not able to detect same [157].

2.14 RADIOGRAPHIC TESTING APPLICATION IN FSW DEFECT ANALYSIS

Radiographic testing (RT), is a non-destructive testing technique used to analyse hidden flaws by using the ability of short wavelength electromagnetic radiation to penetrate materials. A radioactive source uses photons such as Ir-192, Co-60 and Cs- 137. Neutrons can also be used as a replacement of photons to penetrate the material. X-ray radiography can also be used to analyse weld defects [158].

“Digital image processing involves manipulating one or more digital images Operations of practical importance on a single image involve artifact suppression, grey-scale manipulations, distortion corrections, and edge enhancement. Important techniques involving two or more images are (time difference) subtraction, dual energy cancellation of bone or soft tissue structures, and the extraction of flow or organ function parameters. The processing of a single image can be classified as a point operation, a local operation or a global operation. A point operation uses a single point (pixel) of the initial or input image to obtain the corresponding point of the final or output image. A local or neighbourhood operation uses several pixels in a limited area of the input image to obtain a point in the processed or final image” [153].

Gaurav et al., [153] performed a digital radiography test for GTAW and FSW weldments of AA7075-T651 aluminium alloy and observed the defect of porosity, cluster porosity and small tunnel defect.

Edwards et al., [159], used X-ray radiography and validated with metallographic images to examine friction stir welded 6 mm thick Ti-6Al-4V butt joints processed embedded with a tracer material. A variety of rotational speed and traversing speed were implemented. They concluded that X-ray radiography scans were able to depict the characterization of the material flow behaviour [159].

Patil et al., [160], applied X-ray radiography testing technique to analyse weld defects in similar and dissimilar friction stir welded aluminium alloy butt weld joints. They used AA7075 T651 to process similar material butt weld joints. AA77075 T651 and AA6061 T6 aluminium alloys were used in processing the dissimilar weld joints. Different tool rotational speeds (800, 900, 1000 rpm) and feed rates (30, 35, 40mm/min) were employed with all the other process parameters optimised at constant. The visual inspection were only able to identify flashes on all the samples. The x-ray radiographic testing techniques revealed the presence of lack of penetrations flaws in all weld samples and cracks, voids, wormhole defects in some of the welds. They concluded that the x-ray radiography was successful in evaluating the weld defects of the friction stir weld sample for similar and dissimilar materials and was observed that increasing the transverse speed increases the occurrence of weld defects [160].

2.15 LIQUID DYE PENETRANT TESTING APPLICATION IN FSW DEFECT ANALYSIS

Liquid dye penetrant testing is a technique that uses a penetrant that is sprayed on the surface of the material to be inspected for near surface defects. The penetrant is allowed to soak into the material for a period of time referred to as the “dwell time”. Excess penetrant is wiped off the material surface before a developer is sprayed to allow the penetrant to float to the surface the penetrant through flaws. The visible colour is then used to give indication and orientation of the defects [161].

Liquid dye penetrant is limited to surface defects and root defects and is unable to detect and quantify typical defects [162].

Kichen et al., [163], studied the resultant anomalies caused by operating outside the optimum friction stir welding process parameters. They used numerous techniques such as visual inspection, liquid penetrant amongst others to inspect weld defect of friction stir welded AA 2195 aluminium alloy. P135E and P6F4 liquid penetrants were applied on single etched and double etched test samples. Inspections on etched test samples revealed lack of penetration. They observed the liquid penetrants P135E and P6F4 had different sensitivity detection levels, which resulted in dissimilar results on lack of penetration. The P135E detected defects much deeper that the results obtained for the P6F4. Compared to the single etching, double etching enhanced the detection of lack of penetration with better results. They further studied the effects of dwell time and the use of developer and observed both did not affect or improve the detection of lack of penetration flaws. They concluded that effective etching improved the detection of the flaws and the technique is limited to defects of depth of 1.6mm extending from surface.

2.16 SUMMARY

An overview on the evolvement of FSW process was highlighted in the literature review. A comparison of fusion welding and solid state welding was drawn. The superiority of solid state welding in FSW of alloys has generated interest of researchers to wide range of FSW of aluminium alloys varying series.

A wide range of aluminium alloys with distinct metallurgical properties have aroused interest to FSW research. These include the 1xxx, 2xx, 3xxx, 4xxx, 5xxx, 6xxx and 7xxx alloys, as well as the newer Al-Li alloys. Furthermore, these different class alloys have inherently different forging characteristics. As such, processing for each alloy class may vary. However, high-integrity joints can be achieved in all classes.

The AA6061 aluminium alloy, belonging to the 6000 series was selected in this research work. It offers distinct characteristics such as retaining ductility, good

workability, moderate strength, weldability and remarkable corrosion resistance. AA6061 aluminium is commercially available for applications in aviation, mining, marine and sheet metal work. The literature review covered the applications of FSW aluminium in several economic sectors offering a distinct advantage to reduce equipment masses and cost savings.

The quality of welds is influenced by the process parameters. As such to effectively investigate the effect of processes parameters, the process matrix was developed. The main process parameters such as rotational speed, traverse speed will be varied while maintaining constant optimum plunge depth, pin displacement, tool tilt, tool geometry and the dwell time.

In order to validate the integrity of the welds, the welds zones mechanical properties will be compared to that of the parent metal. Hardness values and microstructural evaluation will be taken at the midpoint of the weld samples. The midpoint is where aggressive varying temperature compromises the weld integrity. Microstructural evaluation of the welded specimens will be characterized to understand the different heat affected zones of the FSW joint, determine and compare granular particles precipitate formation.

The weld defects will be investigated to find the effect of FSW intermetallic phases that lower the toughness leading to flaws formations. The ability of the acoustic waves to propagate in aluminium alloys will be investigated. The basic background theory on wave attenuation effects has been discussed in the literature review.

Recent papers have focused on detection of FSW aluminium defects with no investigation on the energy intensity and amplitude reduction due to attenuations effects resulting from waves scattering and absorption. The literature review has given a background on the behaviour of acoustic waves when propagating in materials to detect flaws.

This chapter offered an in depth discussion into the literature and theory in respect to this research study.

The next chapter will discuss the experimental procedures used in this research study.

3. CHAPTER THREE - EXPERIMENTAL PROCEDURE

3.1 INTRODUCTION

This chapter outlines the experimental techniques used in the evaluation of the FSW microstructure, hardness, tensile strength and weld defects analysis of the aluminium butt welded sample. The non-destructive procedure to determine the couplant effects using contact UTS pulse-echo is also discussed. The chapter addresses in detail the materials, equipment and methods employed for the various investigations. These investigations range from the sample chemical composition, the FSW processing and its defined process parameter matrix, the welded sample and the preparation of samples for laboratory analysis. The laboratory experimental procedures and standards used during the investigations are discussed in detail.

3.2 PARENT MATERIAL

The parent material used in this study was an AA6061-T6 aluminium alloy. The dimensions of each aluminium specimen were 200 x 150 x 4.09 mm3 sheet having a friction stir welded butt joint length of 120mm. A spectrometer was used to analyse the elementary composition of the parent material as shown in Appendix A1, which was found to be in accordance to AA6061-T6 standard specifications [164]. The chemical composition of the samples is shown in Table 3.1. The mechanical properties of the base metal can be seen in Table 3.2.

Table 3.1: Chemical composition of AA6061-T6

Element Si Fe Cu Mn Mg Cr Ni Zn Ti Al Ag

Wt. % 0.68 0.49 0.21 0.08 0.84 0.06 0.01 0.07 0.07 97.40 Balance

Table 3.2: Mechanical Properties of the AA6061-T6 aluminum alloy

Yield Strength (MPa) Ultimate Strength (MPa) Elongation (%) Hardness(HV) 269 ± 2 300 ± 1 15 ± 1 105 ± 1

3.3 FSW PROCESS

The friction stir weld process was conducted using a computer controlled 2-Axis FSW machine at the Indian Institute of Science (IISC), Bangalore, India. It is a custom designed computer controlled 2-Axis FSW machine which has its tool traversing in a horizontal position to process the weld on a specimen that is held in a vertically positioned clamp as shown in Figure 3.1. This machine was developed with the help of ETA technologies in Bangalore and has the capability to vary tool rotational speed, traverse speed and plunge depth during the friction stir welding process.

Figure 3:1: FSW platform

As concave shoulder tool with a cylindrical pin length of 3.8mm constructed from tool steel material W302 was used to produce the welds. The tool material properties chosen offered high strength, hardness, availability and low cost. A vertical clamping 53

system was used to secure the workpiece laid on a rigid, smooth, and mild steel backing plate. The orientation of the back plate was configured to enable withstanding of the significant perpendicular and lateral forces applied during the friction stir weld process as the tool mounted on a FSW horizontal platform approached the workpiece located in the vertical position with joint positioned at 180º. Tool tilt angle was kept constant at 0º as the tool traverse along the joint line, and all FSW were performed in a position control mode, with a plunge depth of 3.5 mm. The plunge depth was optimized to achieve a downward pressure which was required to ensure the tool fully penetrated the weld. The weld matrix implemented consists of nine weld samples with optimum parameters carefully selected for the research study outcomes. A criteria of low, medium and high settings was used to categorise the rotational speeds and feed rates. The rotational speeds used were 700, 900, and 1100 rpm (L, M and H respectively) and the feed rates were 60, 80, and 100 mm/min (l, m and h respectively).

Table 3.3: Process parameter matrix of the FSW process

Weld Designation Rotational speed Feed rate Code (rpm) (mm/min

BW1 700 60 Ll

BW2 700 80 Lm

BW3 700 100 Lh

BW4 900 60 Ml

BW5 900 80 Mm

BW6 900 100 Mh

BW7 1100 60 Hl

BW8 1100 80 Hm

BW9 1100 100 Hh

Where,

L = low rotational speed, l = low feed rate M = medium rotational speed, m = medium feed rate H = high rotational speed, h = high feed rate

The FSW machine has a user interface monitor with a keypad were input parameter are entered to enable the customised software programme to output the FSW preferred control system process. The tool is configured to traverse in the x and z directions, rotating and engaging all the defined process parameters and derivatives using the customised programme. Upon plunging into the workpiece, a dwell time of 3 seconds was provisioned to allow intense heat generation sufficient to cause plastic deformation to the tool localized workpiece material. The traverse speed of the tool was controlled along the weld line in accordance to the feed rate process weld matrix while maintaining the rotational speed for low, medium and high for the respective matrix category. The resulted forces from the welding process were recorded using the data retrieval system in the machine’s coded software programme. Once all the nine samples were produced they were then sent for further test namely, visual inspections, non-destructive tests, microstructural characterization, hardness and tensile tests to determine the integrity of the weld.

3.4 SAMPLE PREPARATION

Figure 3.3 shows a Mecatome T300 cut off machine that was used to section the samples into specimens extracted from the plunge zone, stabilized region and the retraction zone as illustrated by the test specimen layout in Figure 3.2. A diameter 230mm x 1.5mm thick cutting off wheel was mounted on the cut off machine at a pre- set rotational speed of 3800 rpm. A precise manual force was applied on the cut-off machine lever to provide an optimum feed rate to produce the 25 x 5 x 4.09mm specimens used for microhardness, tensile and microstructure analysis. A wet abrasive cutting system was employed in order to achieve a sample that was representative of maximum planeness, minimum deformation and structural changes to achieve minimum cutting post-processing. The mounting of the sample in the cut- off machine was by means of an adjustable jaw that secured the workpiece from to eliminate lateral movements. This was important to ensure the cut sample was free from burrs and structural deformations. The cutting parameters initially discussed that is, the rotational speed, feed rate, cut-off wheel material and applied manual force were carefully selected and controlled in relation to the sample hardness and cut-off wheel hardness for easy of cutting. The hardness of the cut-off wheel selected was four times the hardness of the sample being cut. 55

Figure 3:2: Layout of Specimens

Figure 3:3: Mecatome T300 Cut off discharge machine and typical cut specimen

Figure 3:4: Schematics of samples for microstructural and microhardness tests

A total of 54 samples of size 25 x 5 x 4.09 mm were produced. 3 samples for microstructure and the other 3 samples for microhardness were sectioned from each of the initial 9 weld samples. The sectioned samples were representative of the plunge zone, the stabilised zone and retraction zone from the joint interface and each mounted as shown in Figure 3.4. A Struers hot mounting machine shown Figure 3.5 was used to mount the specimens using a poly-fast thermoplastic hot mounting resin. Each of the mounted sample from the plunge zone, the stabilized region, the retraction zone was clearly identified for microhardness test and microstructure analysis by using an engraving tool to indent a mark number in respect of process parameter and purpose of test.

Figure 3:5: Struers hot mounting machine

The samples were mounted such that the retreating side of the weld was always to the left. The mounting of the samples was produced using process parameters shown in Table 3.4. Standard metallographic procedures were followed during the mounting process [165] , as presented in Appendix B1.1.

Table 3.4: Poly-fast Thermoplastic resin mounting process parameters

Parameter Value Pressure 250 Bar Temperature 180ºC Heating Time 3.5mim Cooling Time 1.5min

A Struers polishing machine shown in Figure 3.6 was used for grinding and polishing the mounted test samples surface with the specimen to remove scratches, deformation and to ensure the specimen retained all of its structure elements. The samples were mounted, ground and polished following the standard metallographic procedures [166], shown in Appendix B1.2. Distilled water was used to remove any foreign matters on the polished specimen surface and wiped with a smooth clean cloth to achieve a plane reflective surface. Samples were then chemically etched to reveal their microstructures. At the initial stage of the procedure, the samples were etched with the Keller’s reagent with a solution of 190 ml distilled H2O, 5 ml HNO3, 3 ml HCL and 2 ml HF. The initial etching did not reveal the grains clearly. Subsequently, Weck’s reagent of solution 100 ml distilled H2O, 4 g KMnO4, 1 g NaOH was used to achieve grains that were clearly visible. The samples were then stored safe, ready for microstructure and hardness analysis.

Figure 3:6: Struers polishing machine

3.5 MICROSCOPY

Figure 3.7 shows the optical microscopy workstation with the equipment, Olympus BX51M and Olympus SZX16 optical microscopes. Microstructure observations were conducted using the Olympus BX51M and macrographs for the specimens implemented using the Olympus SZX16. Olympus Stream Essential software program was used to process the digital output. The plunged region, stabilization region and the retraction point cross traverse weld microstructures for all the nine test samples were produced for the WN, TMAZ, HAZ, and the BM zones for comparison.

ASTM E112-12 [167] standard test specification was used to determine the average grain size of the friction stir weld microstructure.

Figure 3:7: Setup of the Optical microscopy 60

The analysis of the chemical composition of the phases at the joint interphase and quantification of chemical composition of the intermetallic compounds was carries out using the Scanning Electron Microscope. Figure 3.8 shows the TESCAN VEGA3 scanning electron microscope (SEM) equipment set-up used. The system is processed by implementing Vega TC software to output the image on the SEM.

Figure 3:8 : TESCAN VEGA3 scanning electron microscope

3.6 VICKERS HARDNESS TESTING

Vickers microhardness values were measured using a ZHµ Vickers microhardness tester, by ZwickRoell (Figure 3.9) according to ASTM 384-16 [168]. Hardness profiles were obtained across the weld traverse of FSW samples at a depth of 1.5mm from the top surface. The measurements were taken in as-polished conditions by indenting with a HV0.5 load of 4.903 N and a dwell time of 10 secs. The indentations were taken at 2.2 mm intervals on the sample with focusing done manually for the indentations. The hardness values were digitally displayed.

Figure 3:9 : ZwickRoell ZHµ Vickers microhardness tester

3.7 TENSILE TESTING

Figure 3.10 shows a servo-hydraulic Instron tensile testing machine model 1195 which was used to carry out the tensile strength tests and the percentage elongation values were obtained.

Figure 3.11 shows a typical tensile test specimen schematic that was used to produce 27 tensile samples with specification dimensions in accordance to ASTM B557-14 standard [169].The tensile testing procedure was conducted in accordance to the ASTM E8M-13 standard [170]. Before starting the test, the computer system connected to the machine was set up by inputting the necessary information of each sample’s gauge length and width. The computer system was then prepared to record the data and to output the necessary load-deflection graphs. The test was conducted at room temperature by gripping the ends of the samples in the tensile test machine and then loading at a constant cross head speed until failure. An extensometer was used to measure the strain of the samples during the experiment at an extension rate

of 5 mm/min and a gauge length of 25 mm with a 100 kN load cell. The load-deflection curve was shown on the computer screen as a visual representation, and the data collected using a customised Instron Bluehill2 software. The tests were repeated three times to check for consistency and accuracy. After testing, the final length of the failed samples was measured to determine the ductility of the samples.

Figure 3:10 : Instron tensile testing machine

Figure 3:11: Schematics of tensile samples (All dimensions in mm)

3.8 LIQUID DYE PENETRANT TESTING

Nine specimens made of aluminium AA6061-T6 were tested for surface root defects using liquid dye penetrant testing technique. Inspections were done on 100% along the root of the weld for any defects according to ASTM E1417-16 standard [171]. Each specimen was cleaned to remove any dirt, oil and any loose scale by use of a solvent to ensure the area was free from extraneous matter. A penetrant of type Adrox 996 PB was sprayed over the area to be examined using an aerosol can. A dwell time of 40 minutes was allowed for penetrant to soak into the flaws. A solvent dampened lint free towel was used to wipe any excessive penetrant. A thin layer of Ardox 9D1 B developer was applied using an aerosol can and 30 minutes allowed to draw the developer from defects. Any areas that bled out were examined to give an indication of orientation and type of defects on the surface. Ardox 9PR 5 was then applied to provide post testing cleaning.

3.9 DIGITAL RADIOGRAPHY TESTING

X-ray digital radiography was conducted on the test samples to inspect weld defects using the equipment Vidico fox-Raysor, 80kV, 0.6A and Balteau X-ray source as presented in Figure 3.12.

The radiography procedure was conducted according to ASTM E155-15 standard specification [172]. The foX-Rayzor FPD was used in combination with a Vidisco supplied 270 kV pulsed x-ray source or almost any x-ray source / isotope. The recommended x-ray sources must be up to 160 kV, otherwise shielding is required for higher voltages.

The Imager detector equipment consisted of X-ray scintillate, photodiode and thin film transistor (TFT). The TFT is an electronic switch made from amorphous silicon (a-Si) on a flat panel detector / imager, allowing the charge collected at each pixel to be independently transferred to electronic circuits where it is amplified by photomultiplier to be then converted from analogue to digital into 512 gray shades in order to produce the digital image data. The digital image is processed to produce the suitable image for display and diagnosis on the control and display unit (CDU - laptop computer). The

imager control unit (ICU) serves as the interface between the flat panel detector / imager and the CDU.

Figure 3:12: X-ray digital radiography inspection equipment set-up

The ability to detect defects is dependent on the quality of the image. The quality of a computed radiography image includes the geometric un-sharpness, the signal / noise ratio, the scatter and the contrast sensitivity. To achieve and capture all the characteristics an amorphous silicon flat panel imager of 13 mm deep was used. The panels have a wide dynamic range of 16,384 gray scale levels (14 bit) and high resolution of 3.5 line pair / mm and imaging area of 482 cm2. Super sensitive panels

reveal layers of information that enable increased detection capabilities with more details for improved analysis.

A Balteau NDT Baltospot – Ceram 235 generator was used. The generator uses a pulsed wave high voltage that can drive a higher current at full power and will still keep an extremely broad spectrum to guarantee that soft X-rays will be present even at full kV giving a contrast that is not comparable with any other technique at the same kV. The parameter for the x-ray radiography testing used are shown in Table 3.5.

Table 3.5: X-ray radiography exposure parameters for FSW joints

Process Parameter Value

Electrical voltage range 70 – 235 [kV]

Tube current range 0.1 – 5 [mA]

Exposure Time 30 [sec]

X-ray tube beam angle 40º

Focal spot size 2 x 2 [mm]

X-ray Inherent filtration 1 Be mm Source to Object 664 [mm] Distance (SOB)

In conducting the x-ray radiography examination, the machine was switched on and left for to warm up for a period of about 20-30 minutes. The power was set at 80kV and a tube current of 0.6 mA. An exposure time of 30 seconds for the aluminium test sample was determined from the exposure chart referred in Addendum B. The samples were taped with a lead number that corresponded to the process parameters. The sample were placed on the imager and the source to detector distance was kept at 664mm. Once the X-ray had shot the sample, immediately the radiography image was displayed on the computer for any interpretation and density adjustments.

3.10 ULTRASONIC TESTING

This section details experimental procedures using the UT contact pulse-echo technique to investigate the couplant attenuation effect applying the longitudinal wave, the UT ToFD principle to analyse defects using the acoustic waves of the shear mode and lastly, the UT phased array technique to analyse sub surface defects using multiple focuses beams.

3.10.1 Couplant attenuation effects Investigation

At initial stage, all friction stir welded specimens of aluminium AA6061-T6 with dimensions 200x150x4.09 mm were subjected to defect inspections using Ultrasonic testing to analyse the couplants attenuation effects in contact pulse-echo mode using the normal longitudinal wave form. Figure 3.13 show the UT equipment, an Olympus Epoch 600 with a 4 MHz single crystal Olympus Panametrics of 24 mm diameter longitudinal wave transducer.

Probe without a membrane

Figure 3:13: Figure 3.13: Epoch 600 UT equipment with probe without a membrane

Probe without a membrane

Figure 3:14: UT pulse-echo equipment set-up showing probe with a membrane

Four different types of couplants namely oil, grease, ultrasonic gel and water were selected for the investigation and comparison. Ultrasonic gel and oil were chosen as these are high-performance ultrasonic couplants. In particular, ultrasonic gel, is routinely used in commercial applications as it provides good transmission, however it is only suitable for relatively short measurement periods. Grease was also selected for the comparison as it is widely used in industrial applications due its non-drip consistency and stability with time. Water was used as it is readily available, predominantly used and provides high impedance properties. Four sets of measurements were carried out.

The probe was covered with a membrane and a layer (thickness) of couplant applied on the surface of the sample which was kept constant by applying the same pressure on the transducer throughout all the experiments. The pulse echo technique in accordance to ASTM E164-13 standard specification [173] was used for the attenuation effects investigations. After determining the near fields of aluminium and steel which showed negligent variation, a V1 calibration block [174] was used and the frequency recorded at 80% full screen height of the first echo. Observations of the behaviour for the first back wall echoes, the amplitudes of the back walls were recorded. The same experiment procedure was repeated for all the couplants without application of a membrane on the probe.

The single crystal transducer was replaced with a 4 MHz twin crystal transducer without a membrane. Only three couplants namely water, ultrasonic gel and grease 68

were applied. Oil was not applied during the tests using the twin crystal transducer. The acoustic barrier (usually cork) absorbs the oil causing interference with the sound waves. Care had to be taken to ensure that the acoustic isolation between the transmitter and the receiver crystals are not impaired.

The ultrasonic pulse generator Olympus Epoch 600 was used to excite the probe and to receive the pulse echo signals. The signal was measured and a back-echo wall wave displayed on the screen in frequency-time domain.

3.10.2 Time of Flight Diffraction (ToFD)

Defects in the welds were examined using the time of flight diffraction principle of the shear wave form. Figure 3.15 shows the ToFD equipment unit Handy Scan with a 60º angle beam with a 5 MHz piezoelectric transducer. The testing procedure was done in accordance to specification ASTM E2373-04 standard [175].

ToFD inspection procedure was carried out as follows:-

1. The probe was calibrated before inspection using a 12.7mm reference plate. 2. A dedicated channel 1 was used to transmit waves through the probe and set for ToFD configuration. 3. Channel 2 was set-up as the receiver connection. 4. The probes were placed so that they straddle exactly on the centreline of the weld. 5. A smooth scan was carried out along the full length of the weld, ensuring that no scanning data is missing

The scan was saved as shown in Figure 3.16 with all indications found.

Figure 3:15: UT ToFD Handyscan equipment with probe emitting shear wave mode inspection on FSW samples

Figure 3:16: UT equipment showing ToFD image of the sample

3.10.3 Phased Array UT inspection

The phased array technique was implemented to analyse presence of sub surface weld defects. Figure 3.17 shows the equipment, an Olympus Omniscan MX2 equipment with a 10L16-4.96X5-A00-P-2.5-OM and SA00-N45S Wedge.

Figure 3:17: UT Phased Array inspection on FSW samples

The phased array calibration for aluminium was performed in accordance to ASTM E2491 [176] standard specification, using a reference from the side drilled hole of 1.5mm set at 80% full screen height.

Figure 3.18 shows the phased array block and probe set-up. The beam angle was set electronically to scan the weld longitudinally. This allowed a full coverage of the weld with focused beams. The refraction angles from 30º shear waves to 70º shear waves were used for the investigation, with the 0º longitudinal waves used for observing the coupling check.

The scans were produced as a combination of images processed by the in-built software. A multiple display was set to show a combination of three images, the waveform (A-scan), the sectorial / top view linear scan (C-scan) and the partial top view (B-scan).

The phased array inspection procedure was carried out as follows:-

1. The probe set up was calibrated before inspection using the phased array calibration block, reference 1.5 side drilled hole set at 80% screen height for Aluminium inspections.

2. A mark was identified at 8mm distance from the root of the weld to the edge of the plate. 3. The probe was placed 8mm away from the weld of side A, longitudinally traversing as shown in Figure 3.18. 4. A scan alongside the marked distance was carried out and the scan was saved. 5. A repeat of the same scanning procedure on side B of the plate was performed.

Figure 3:18: UT Phased Array probe longitudinal traverse directions relative to the weld

Figure 3:19: UT Phased Array probe set-up

3.11 SUMMARY

Friction stir welding input process parameters used to produce the nine AA6061-T6 of similar material weld samples were discussed. The rotational speeds were classified as low, medium and high of 700 rpm, 900 rpm and 1100 rpm respectively. A weld

matrix was developed using the feed rates of 60mm/min, 80mm/min and 100mm/min classified as low, medium and high respectively.

The samples were sectioned to produce nine samples clearly marked for microhardness and microstructure analysis. Twenty seven tensile specimens were produced and tests conducted representative of all the nine weld sample’s plunge zone, stabilized zone and the retraction zone. Non-destructive tests were carried out on all the nine samples for defect analysis. The experimental procedures for all the various analysis were discussed in this chapter.

The next chapter reports the results and discussion of the nine welds produced using the weld matrix.

4. CHAPTER FOUR - RESULTS AND DISCUSSION

4.1 INTRODUCTION

This chapter presents the results and discusses the effects of process parameters on the mechanical properties and microstructure of the friction stir processed butt weld joint of similar AA6061-T6 aluminium alloys. The weld defects results and analysis using visual inspection and non-destructive techniques are also discussed in this chapter.

4.2 PHYSICAL APPEARANCE OF THE WELDS

The most straight forward and easiest inspection technique is visual inspection, which is an effortless method of inspecting for surface features including excess flash, galling, shoulder voids, and even weld misalignment.

Figure 4.1 shows the physical appearances of the friction stir processed weld joints at different weld speeds and tool rotational speeds. Visual inspections on all the samples produced showed no visible defects such as cracks, wormholes, or any other near surface deformities. A key hole was observed on all the weld samples at the point of the tool retraction as common phenomenon on FSW.

All FSW samples produced were observed to be having weld flashes on the tool retreating side with flashes that are curled one over the previous one. The excessively large curled flashes concentrated towards the retraction point are more noticeable on welds produced using the tool rotational speed of 900 rpm at a feed rate of 60 mm/min. Likewise, the largest curled flushes concentrated on the retreating side around the plunge zone are seemingly noticeable for the weld sample produced using the tool rotational speed of 1100 rpm at feed rate of 80 mm/min. The large flash curls can be attributed to the resultant outflow of the highly plasticised material from underneath of the shoulder necessitated by the excessive frictional heat generation and large forces. Reduction in flash size is evident when the same rotational speed of 1100mm/min is used to produce a weld at increased feed rate of 100 mm/min, which is attributed to optimal heat energy to necessitate effective material flow of the plasticized material.

Table 4.1: Physical appearances of the FSW samples

Rotational Speed Feed rate matrix and weld physical appearance 60mm/min 80mm/min 100mm/min (rev/min)

700

900

1100

In essence, large flash curls appear to be caused by the generation of high temperature combined with severe local shearing action due to the tool movement. The curling wavelike flashes are a result of the movement of plasticised material 76

displaced by the pin with the tool shoulder not confining the material. The tool rotational speed of 700 rpm and 100 mm/min appear to have processed a weld with the least flash sizes.

It is evident from the results that flashes are necessitated by material flow behaviour which is predominantly influenced by the FSW process parameters such as the tool dimensions and the rotational speed. It is noticeable, rotational speed appear to be the predominant influencing process variable since it is associated with the translational velocity. Murr et al., [129], reported that increase in strain rate can influence the recrystallization process and these strain rates are associated with very high rotational speeds which influence the generation of excessively large flashes. Also, low heat inputs are predominantly associated with higher welding speeds which causes the welded joint to be cooled at a faster rate. The curled flashes could thus be attributed to incorrect combinations of the rotational speed and the feed rate which in instances yields low axial forces. Such low axial forces, contributes to formation of non-symmetrical semi-circular features and in instances large flashes at the top surface of the weld due to poor plasticization and consolidation of the material under the influence of the tool shoulder [9]. The non-symmetrical semi-circular features were not evident with the visual inspection on all the welds. This implies that the welds were homogenous and of sound quality. Optical microscopy and other analytical techniques will be required to verify this.

The tool pin used in this study for the friction stir welding was of a fixed pin length. The major disadvantage is the singularity ability to weld materials of the same thickness. It has been observed and reported that all weld samples had a keyhole in the workpiece left when the pin is retracted. The elimination of the disadvantage associated with the keyhole is eliminated by using a retractable tool pin.

4.3 MACROSTRUCTURAL CHARACTERIZATION

Figure 4.1 show the typical macrograph of the friction stir welded sample while the rest of the macrographs in respect of the various process parameters implemented in this study are shown in Appendix C1. The macrographs shows no internal defect presence in any of the weld samples. The morphological feature obtained using the optical microscope revealed all the four distinct FSW microstructural zones namely: the 77

Thermo-mechanically Affected Zone (TMAZ), the Heat Affected Zone (HAZ) and the Stir Zone (SZ) or the Nugget Zone (NZ), and the Parent material.

The shape of the weld nugget was not easily revealed but a closer observation on the 700rpm_80mm/min weld sample seems to resemble a basin shaped nugget. It is important to note that a sharp demarcation between the WN and TMAZ is more visible on the advanced side and more diffuse on the retreating side. This observation was also confirmed by Huang et al. [141], who reported that the strain rate and temperature gradients are much steeper in advancing side than that in the retreating side. However, the morphology can be noticed to be more diffuse for the 700rpm_60mm/min and 1100rpm_80mm/min samples, both on the advanced and retreating side. This could be attributed to less heat energy towards the base metal.

Figure 4:1: Typical Macrograph images of FSW samples

4.4 MICROSTRUCTURAL EXAMINATION

An optical microscopy and SEM examinations were used to study the FSW microstructure zones for all the welds produced. Figure 4.2 (a) shows the parent metal microstructure with a honeycomb-like shape as revealed in the SEM micrograph presented in Figure 4.2(b). Presence of the second phase precipitates can be seen in the microstructure analysed using SEM. These precipitates are mostly elements of Mg2Si [93]. Figure 4.3 presents the microstructural evolution for the weld sample produced by 700rpm and 80mm/min. The rest of the weld samples micrographs grouped in their respective rotational and feed speeds in all the three weld zones are presented in Appendix D. The weld zones’ diffuse contrast can be observed indicating the different material thermos-flow and stresses not analysed in this study [103]. 78

(a) (b)

β” Precipitates

Figure 4:2 – AA6061-T6 Base Metal Microstructure; (a) OM micrograph (b) SEM

SEM was used to analyse the welded zone microstructure as well as presence of the intermetallic and the resultant micrographs are presented in Figure 4.4 to 4.11. All the micrographs have white and grey spots which are evident within the grains. All these spots represent intermetallic particle constituents. Weck’s reagent can be seen to have attacked some of the intermetallic particles and the grains during the etching processing. These attacks should not be confused with the weld defects such as voids which will be analysed in the next section using non-destructive testing methods. As such, similar to the standard FSW, three different weld microstructural zones, i.e., heat-affected zone (HAZ), thermo-mechanically affected zone (TMAZ), the weld nugget (WN) also known as the stirred zone (SZ) were exhibited. This study earlier reported that, the BM is characterized by elongated grain structures which are much larger when compared to all the other weld zones. Figure 4.3(a) to 4.11(a) show the HAZ, which was common to all welding processes subjected to thermal cycle with no deformation, which however, has almost no detectable change from the parent metal. Although the macrographs for TMAZ (Figure 4.3(b) to 4.11(b)) presents the TMAZ grain structure observed from the optical microscopy showed a mixed rotated elongated and equiaxed grains which reflects grains of partially recrystallized morphology. The WN (Figure 4.3(c) to 4.11(c)), has much smaller fine equiaxed grains due to dynamic recrystallization. In addition, a sharp transition can be seen between the TMAZ and the SZ on the AS while a much more diffuse interface at the RS is observed from the images shown in Appendix D. This contrast is in relation to the different material flow conditions on the two sides. 79

Figure 4:3 – Optical Microstructure micrographs of the welds produced at 700rpm and 80 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget

Figure 4:4 – SEM Microstructure of the welds produced at 700rpm and 80 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget

Figure 4:5 – SEM Microstructure of the welds produced at 700rpm and 60 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget

Figure 4:6 – SEM Microstructure of the welds produced at 700rpm and 100 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget

Figure 4:7 – SEM Microstructure of the welds produced at 900rpm and 60 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget

Figure 4:8 – SEM Microstructure of the welds produced at 900rpm and 100 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget

Figure 4:9 – SEM Microstructure of the welds produced at 1100rpm and 60 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget

Figure 4:10 – SEM Microstructure of the welds produced at 1100rpm and 80 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget

Figure 4:11 – SEM Microstructure of the welds produced at 1100rpm and 100 mm/min (a) HAZ; (b) TMAZ (c) Weld Nugget

Table 4.2 shows the average grain sizes in the different 4 regions. The grains in NZ are much smaller than those in other regions. The average grain size in the four zones are as follows in the order of HAZ > BM > TMAZ > NZ, similar to results by Fadaeifard et al., [103], and contrary to results obtained by Chen et al., [177], who reported results in the order of BM > HAZ > TMAZ > NZ, but also observed that the grain sizes were also similar. The almost similar average grain sizes between HAZ and BM agrees with most of the results in Table 4.2.

Table 4.2: Average grain size in various region

Rotational Speed Welding NZ TMAZ HAZ BM Sample (rpm) Speed (rpm) (µm) (µm) (µm) (µm)

BW1 700 60 4.4 85.4 129 125 BW2 700 80 3.7 83.1 127 125 BW3 700 100 3.43 79.2 126 125

BW4 900 60 7.4 92.7 122 125 BW5 900 80 6.8 95.1 121 125 BW6 900 100 5.43 80.2 118 125

BW7 1100 60 14 95.3 133 125 BW8 1100 80 10.1 95.1 128 125 BW9 1100 100 8.7 80.2 125 125

As presented in the above Table 4.2, in the HAZ, similar trend to the NZ and TMAZ, the average grain sizes decreased with increase of welding speed though it was minimal. For instance, at rotational speed of 700 rpm, the average grain size in the HAZ decreased from 129 µm to 127 µm and finally 126 µm with increase in welding speed from 60 mm/min, 80 mm/min and 100 mm/min respectively. This further confirms the findings by Fadaeifard et al., [103], which reports that the grain size decrease with increase in weld speeds. It has also been reported that the increase in deformation rate caused by the rotating tool plays a major role in reducing grain size. However, it is important to note that both frictional heat and deformation rate reduce with increase in traverse speed [103].

In most of the weld samples, the HAZ grain size is slightly larger than in the BM. This grain growth is highly influenced by heat input in the weldment. Unlike the rotational speed, the weld traverse speed shows a reverse effect on the heat generated during FSW [177], as such the heat energy reduction is relative to increase in the feed rate. Previous studies have shown that the stirred zones has lower temperature and a short duration is available for grain growth after recrystallization to cause smaller grain sizes in the WNZ of welded joints.

Akinlabi et al., [31], reported that there is a derived condition that changes the grain size necessitated by a competition between frictional heat and deformation rate at each welding condition such that the grain size is a net effect of deformation rate and frictional effect.

In this study, it can be deduced that the lower welding speeds were effective in increasing the deformation rate and frictional heat resulting in reduced grain size. However, the reduction of grain sizes with increase in welding speeds suggest that frictional heat is dominant than the deformation rate. The increase in weld speeds also necessitated the reduction in peak temperatures and faster cooling rates which assisted in effecting the reduced grain sizes. It is therefore imperative to conclude that the welding parameters such as rotational and traverse speeds play a crucial role in reduced grain sizes.

In Figure 4.2 (b) above, small rod/needle-like shaped β” precipitates were observed in the parent metal microstructure. The dissolution of these precipitates in the weld samples during FSW can be attributed to the feed rate and the tool rotational speed which are some of the key parameters, as they respectively influence the heating time and the peak temperature. As such dynamic recrystallization (DRX) occurs due to intense plastic flow at high temperatures, more so, in the weld nugget to give fine grains as shown in Table 4.2 and demonstrated in the SEM results from Figure 4.3(c) to 4.11(c). The precipitates which are known to have intense interactions with dislocations and grain boundaries may start to have effect on dislocation density (driving force) and the recrystallization kinetics. This leads to a microstructural evolution that results in dissolution of precipitates. The smaller grain size in the weld nugget could be attributed to grain recrystallization because of the pinning effect of

these particles. The β’’ precipitates are obviously dissolved in the nugget which gave rise to some supersaturated solution in the stirred zone. These heterogeneously distributed precipitates observed in weld nugget gives rise precipitates hardening. The next section on micro-hardness will discuss the hardness distribution across the weld traverse which will be seen to be high in the weld nugget when compared to the TMAZ and HAZ. Sauvage et al., [125] contributed that two mechanisms exists that lead to the formation of heterogeneously distributed precipitates in the weld nugget. The first consideration is the heterogeneous precipitation due to the great increase of the temperature causing the β’’ precipitates dissolve. This they explained is experienced when the FSW pin is moved away, which drops the temperature and increases the driving force for precipitation. The other scenario is during interaction between the coarsening precipitates and dislocations. In hardened aluminium alloys, the dramatic increase of the temperature in the weld nugget has effect on both the solubility of Mg and Si and the atomic mobility. “The solubility of Mg and Si increases in the Al matrix, and this leads to a high driving force for the decomposition of β’’ needle” [125]. This effect plays a crucial role in grain sizes and grain boundaries as well as the associated mechanical properties.

4.5 TENSILE PROPERTIES

The standard deviation range was established by comparing the mean in the ultimate tensile strength (UTS) and yield strength (YS) using a standard deviation equation as shown below.

𝜎̅푈푇푆 = 3𝜎̅푠 (8)

It was observed from the results that a standard deviation range of values was within the 1 – 3 MPa limits. This was deemed to be quite acceptable.

The influence of the tool rotational and weld speeds with different FSW conditions on load-extension behaviour was studied and compared against each weld matrix. Figure 4.12 shows the load-extension behaviour of different combinations of the tool rotational and weld speeds. Among the nine curves, it is clear that the weld sample produced using 1100rpm and 100mm/min has a maximum load of 5.03kN with the corresponding extension of 2.35mm however failing as the second least at the 91

corresponding extension of 3.25mm, due to lesser elongation. The weld sample processed using the combination of 900 rpm and 100mm/min gave the highest extension value of 4.225mm, and had the second highest ultimate tensile strength value of 191MPa (Table 4.3), thus demonstrating highest elongation. The parent metal’s ductility, yield strength and ultimate tensile strength were 15%, ~271 and ~300 MPa respectively. Chen et al., reported the AA6061-T6 base metal elongation, yield strength and ultimate tensile strength of 12.1%, 242 MPa and 272 MPa respectively. Table 4.3 demonstrates that all the weld samples exhibited reduction in the ultimate tensile strength and yield strength by between 34.8± 7% compared to the parent metal. The parent metal of AA6061-T6 is a hardened alloy which relies on precipitates. The thermal gradients and mechanical workings of the pin generated during FSW processing causes dissolution of the precipitates. This decomposition of precipitates and softened material gives rise to the reduction of the mechanical properties such as UTS and yield strength when compared with the base. The reduction of tensile strength inadvertently affects reduced ductility.

Figure 4:12 Load-Extension behaviour for the weld samples processed at different tool rotational and weld speeds

As tabulated in Table 4.3 and illustrated in Figure 4.13 showing the stress - strain curve, the FSW sample processed with the tool rotational and weld speeds of 1100rpm and 100mm/min respectively, recorded the maximum tensile strength, yield strength and weld efficiency of 192.51 MPa, 149.51 and 64.2% respectively. The variation of the ultimate tensile strength, yield strength and efficiency was within close range of between 181.73-192.51 MPa, 116.79-149.51 MPa, 60.5 %- 64.2%, respectively, for all the welded samples. It is also distinctively clear that for all the different tool rotation speeds, the ultimate tensile strength, yield strength and weld efficiency improved with increase in the weld speeds. It can be observed that all weld achieved their maximum UTS (192.51MPa, 191.17MPa, 189.94MPa) at the highest combinations of rotational and weld speeds. This could be attributed to the precipitates size and distribution improvement relative to optimum heat generation, which results in increase in mechanical properties relative to the increase in weld speeds. It can be noticed that a 93

decrease in ultimate tensile strength is also uniform with the decrease in the yield strength and is relative to the decrease of the rotational and weld speeds.

Stress - Strain Curve 250

200

150 [MPa)

Stress 100

0 0.00 0.35 0.70 1.05 1.40 1.75 2.10 2.45 2.80 3.15 3.50 3.85 4.20 4.55 4.90 5.25 5.60 5.95 6.30 6.65 7.00 7.35 7.70 8.05 8.40 8.75 9.10 9.38 9.39 9.40 9.50 Strain %

Figure 4:13: Stress-strain curves obtained from tensile test results of the weld sample processed using tool rotational and weld speeds of 1100 rpm and 100mm/min.

Table 4.3: Mechanical Properties and Weld efficiency obtained by tensile tests Rotational Weld ᶯ compared to base UTS Yield Strength Speed Speed metal Sample (rpm) mm/min (MPa) (MPa) %

BW1 700 60 183.63 114.26 61.2 BW2 700 80 186.57 147.88 62.2 BW3 700 100 189.94 127.24 63.3 BW4 900 60 181.73 116.79 60.5 BW5 900 80 189.45 145.37 63.2 BW6 900 100 191.17 122.46 63.7 BW7 1100 60 184.49 113.76 61.5 BW8 1100 80 188.65 134.08 62.9 BW9 1100 100 192.51 149.51 64.2

Oyindamola [93], concurred that the ultimate strength decreases relative to the decrease in the yield strength. He argued that this could be attributed to the history of age-hardened alloys. The age hardened alloys do not always rely on the grain size but rather on the precipitate sizes and distribution for improved mechanical properties. He also observed that the re-precipitation were ~1.5 times larger than the precipitates in the parent material.

He further argued that the size of the precipitates has a significant influence on how the precipitates influence the mechanical properties of materials. He noticed that the Mg2Si precipitates were divided into several forms of the β” Mg2Si, β’ Mg2Si, and the β Mg2Si. The densely sparse small rod shaped β” Mg2Si (double beta prime) are the major contributors to mechanical properties and are highly characterised in the base metal. The dissolution of the precipitate due to thermal effects and mechanical workings yielding large strains during FS welding resulted in the decrease in the ultimate tensile strength, the yield strength and the ductility when compared to the base metal.

Whilst fracto-graphs were not obtained, by inspection of the fractured specimen, it seemed the fracture location was observed in the weld joint’s weakest region resembling ductile mode of failure. This can be interpreted from the W-shaped reduced hardness values distribution seen at lowest in the TMAZ/HAZ interface, where material softening occurred due to dissolution of precipitates. Hihara et al. [178], confirmed that the interface of the TMAZ/HAZ was the relatively weak region were the tensile results showed it as the fracture location.

Ilangovan et al. [179], observed that there was a 26% reduction from the base metal in the tensile strength of friction stir welded similar AA6061 Al joint which achieved 222 MPa. He confirmed similar mode of failure as this study by arguing that the fine populated dimples observed at higher magnification at the point of fracture resembled ductile mode of the failure.

4.6 MICROHARDNESS PROFILING

Figure 4.14 shows the results of the microhardness test carried out in the stabilized region of the weld. The microhardness profiling was conducted across the weld 95

traverse at a depth of 1.5 mm from the weld top surface. The unaffected parent metal average Vickers microhardness values is ~99 HV±7. The weld zone showed degradation of hardness on all the welded samples when compared to the BM.

Microhardness Results

120 Advancing Side Retreating Side

100 0.5

40 Vickers microhardness Hv microhardness Vickers 20

0 -13.2 -11 -8.8 -6.6 -4.4 -2.2 0 2.2 4.4 6.6 8.8 11 -13.2 Distance from the weld centre (mm)

1100x60 1100x80 1100x100 700x60 700x80 700x100 900x60 900x80 900x100

Figure 4:14: Effect of processing speeds on microhardness across the weld traverse.

The plot shows a hardness decrease towards the weld retreating side for lower rotational speeds. The most affected hardness values are approximately 6.6 to 11mm from the weld centre on both sides of the welds. The rotational speed of 1100rpm_60mm/min recorded the least hardness value of ~ 49 Hv in the HAZ/TMAZ interface of the retreating side. This 48% degradation in hardness is mostly characterized in the TMAZ interface where a combination of high stress and large strains results in the deformation of grain structure were recrystallization does not take place causing coarse grain structure. The graph also shows degradation in hardness values at the interface of the regions attributed to softening of the material. This can be also be attributed to the absence of solid solution in the weakest zone due to the effect of dynamic recrystallization and the effect of dynamic recovery processes [132].

Employing the rotational speed of 700 rpm showed distinct hardness variations across the weld nugget traverse, more-so, for the lower feed rates of 60 and 80mm/min. This is attributed to higher than normal fast cooling rates and varying thermal cycles resulting in poor material flow. Increasing the feed rate to 100mm/min improved the hardness values uniformity, however, with a decrease noticeable in the weld nugget towards the retreating side. The decrease is possibly due to fast cooling rates.

Further increasing the rotational speed to 900 rpm, results in more uniform hardness values in the weld nugget attributed to optimum heat energy and stable thermal cycle effecting good material flow and stabilized cooling rates. However, the hardness value seem to be lower in all the regions and this could be attributed mainly to the coarsening and dissolution of strengthening precipitates induced by the thermal cycle at the lower feed rates of 60 and 80 mm/min. Increasing the feed rates to 100mm/min increases the hardness values which is a result of more plasticized material due to sufficient heat energy and permissible cooling rates.

However, increasing the rotational speed to 1100 rpm at the lower feed rates manages to improve the hardness in the weld nugget region with a huge decrease of hardness in the HAZ/TMAZ interface due to softening of the material which could be attributed to increased heat input. Further maintaining the rotational speed at 1100 rpm and increasing the feed rate to 100mm/min, influenced the hardness values to increase steeply towards the retreating side.

The hardness distribution can be noticed ranging from 67Hv in the HAZ/TMAZ advanced interface to 80Hv in the retreating HAZ side. This unsymmetrical distribution can be to excessive plasticized strains leading to an inhomogeneous weld joint. This behaviour is a result of the substantial fluctuation related to the heterogeneous constitution of the nugget. Chen et al., used a higher rotational speed of 1000 rpm to process an AA6061 FSW butt joint and analysed the dynamic recrystallization effect on hardness. They reported a hardness curve which was asymmetrical with respect to the weld centre line and attributed it to the plastic flow field in the two sides of the weld centre as not being uniform. They noted that the asymmetrical microhardness distribution was also as a result of larger distorted grains and distortion energy which caused the strain-hardness to increase.

4.7 WELD DEFECTS

This section discusses the weld defects analysis determined by conducting non- destructive testing using methods such as the liquid dye penetrant, the ultrasonic testing methods applying the phased array technique as well as the ToFD of the shear wave form principle and the digital x-ray radiography testing. The UT of the pulse-echo mode is used and results of the various couplants were obtained to discuss the attenuation effects. The attenuation effect investigation results are were used to identify the optimal couplants with least signal-noise ratio and highest frequency gain.

4.7.1 Liquid Dye Penetrant Weld Defect Analysis

Table 4.4 shows the nine welds at the root with the dye penetrant trace observations. The results were obtained by taking photos of the samples during weld defect analysis, by the application of liquid dye penetrant testing technique. The results show no significant evidence of the red penetrant trace along the weld root indicating no root defects are found on all the weld samples. The absence of the root defect can be attributed to optimum selection of the rotating tool and the effective plunge depth which were successful in producing quality welds with no root defects at the bottom of the joint. It can be observed that a plunge depth of 3.65mm is effective to produce a quality weld on a 4.09mm thick aluminium sheet using an optimum selected tool. The tool used had a concave shoulder diameter of 25 mm and a cylindrical pin of length 3.8mm and tapered diameter of 6-7 mm.

However, all weld showed a roundish shaped red penetrant trace on the weld ends. These traces indicate a keyhole formed when the tool is retracted on completion of the friction stir welding process.

Patil et al. [160], studied weld defects of dissimilar FSW between AA6061 and A7705 and observed that increase of defects occurrence increased with increase in traverse speed and concluded that that the base metal 6061 which was positioned on advancing side gave good mixing of materials. In this study, it can be noted that by 98

application of liquid dye penetrant, no root defects were observed in all the combinations of speeds for the similar FSW welded AA6061 joint. This is attributed to careful selection of optimum processing parameters aided by the AA6061 base material properties which resulted in conducive material flow conditions.

Table 4.4: Liquid dye penetrant tested weld joint photos

Rotational Feed rate matrix and weld physical Comments Speed appearance 60mm/min 80mm/min 100mm/min (rev/min) No evidence of root defects was observed on 700 all weld produced using the 700 rpm rotation speed and the respective feed rate speeds. Keyhole defects are observed.

No evidence of root defects was observed on

all weld produced using the 900 rpm rotation 900 speed with the 60 and 80 mm/min feed rates. An insignificant tunneled root defect can be seen for the weld produced with the feed rate of 100 mm/min. Keyhole defects are observed.

No evidence of root defects was observed on

all weld produced using the 1100 rpm rotation 1100 speed and the respective feed rate speeds. Keyhole defects are observed.

4.7.2 Ultrasonic Phased Array Weld Defect Analysis

The best non-destructive technique for inspecting defects in the entire volume of the friction stir welds is ultrasonic phased array testing employing a water-coupled wedge normally achieved with a single pass. Traverse defects can also be analysed using phased arrays by permitting lateral scans. In this study the probability of detection was maximised by optimising the probe beam inspection angle. The optimisation of the angle was also done to increase the number of zones that could be covered by the phased arrays so as to provide accurate flaw sizes and location.

It was expected that phased array inspections of friction stir welds would detect all volumetric-type defects such as incomplete penetration, lack of fusion and cracking. As such, all nine samples were examined for the presence of the above mentioned defects using the phased array technique. Figures 4.15 to 4.22 show scans for the weld samples with simultaneous combinations of image results for the corresponding A-scan waveform, a sector scan and a planar C-scan of the weld profile. It is evident from all these results that no volumetric defects were observed on all the samples.

100

(a) (b)

(c)

Figure 4:15 - UT Phased Array scan images for weld produced by 700 rpm and 60 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile

(a) (b)

(c)

Figure 4:16 - UT Phased Array scan images for weld produced by 700 rpm and 80 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile

101

(a) (b)

(c)

Figure 4:17 - UT Phased Array scan images for weld produced by 700 rpm and 100 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C- scan of the weld profile

(a) (b)

(c)

Figure 4:18 - UT Phased Array scan images for weld produced by 900 rpm and 60 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile

102

(a) (b)

(c)

Figure 4:19 - UT Phased Array scan images for weld produced by 900 rpm and 80 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile

(a) (b)

(c)

Figure 4:20 - UT Phased Array scan images for weld produced by 900 rpm and 100 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C- scan of the weld profile

103

(a) (b)

(c)

Figure 4:21 - UT Phased Array scan images for weld produced by 1100 rpm and 60 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C- scan of the weld profile

(a) (b)

(c)

Figure 4:22: UT Phased Array scan images for weld produced by 1100 rpm and 80 mm/min in the (a) A-scan waveform, (b) sectorial scan and (c) planar C-scan of the weld profile

104

4.7.3 Ultrasonic ToFD Weld Defect Analysis

TOFD UT testing was performed on all nine samples to defect sub-surface flaws. Improvements in ultrasonic instruments have eliminated the past challenges of ToFD during initial detection scans that failed to detect surface breaking planar flaws. It has been reported that flaws that “are open to the scanned surface can often be obscured by the presence of the lateral wave response which makes identification and sizing difficult” [13]. To achieve flaw detection, the below listed measures were taken:-

 Use of high energy, short pulse transducers to reduce the lateral wave effect to 2-3mm.  Signal processing was enhanced to ensure lateral wave straightening was removed.  The beam was configured to target the weld specific region.

Due to these control measures the near surface indication detection were significantly improved, however if very near surface critical sizing is called for it may be necessary to ally TOFD with a NDT surface inspection technique.

Figures 4.23 - 4.31 presents the TOFD scan images in time domain over the longitudinal length of all the nine welds. Figure 4.23 shows the scan for the weld sample processed using rotational and weld speeds of 700rpm and 60 mm/min which indicated attenuation in the shear wave zone indicating lower fusion at the longitudinal mid-point of the weld. The weld sample processed using rotational speeds of 900 rpm for both 80 and 100mm/min weld speeds, Fig 4.27-4.28 respectively, showed the scans with attenuation in the compression and shear wave zone indicating lower fusion also in the centre on the longitudinal weld. The rest of the weld sample showed no fusion problems indicating good material flow was achieved.

These results with lower material fusion correspond well with the lower hardness values obtained and reported earlier in this study for these specific mentioned samples. This possibly could be attributed to poor material mixing conditions. In instances of wrongly selected process parameters, at weld mid-point, thermal variations at peak heat energy results in excessively plasticized material at fast cooling

105

rates which results in coarsened grains. During TOFD scanning, attenuation mostly by wave scattering and absorption of waves is mostly experienced respectively at grain boundaries and in coarse grains. This effect of attenuation is thus used to interpret poor material mixing and lower fusion.

Attenuation in shear wave zone indicating lower fusion

Figure 4:23 - TOFD scan images obtained for the weld produced by 700rpm and 60 mm/min

106

Figure 4:24 - TOFD scan images obtained for the weld produced by 700rpm and 80 mm/min

Figure 4:25 - TOFD scan images obtained for the weld produced by 700rpm and 100 mm/min

107

Figure 4:26 - TOFD scan images obtained for the weld produced by 900rpm and 60 mm/min

Attenuation in compression and shear wave zone indicating lower fusion

Figure 4:27 -TOFD scan images obtained for the weld produced by 900rpm and 80 mm/min

108

Attenuation in compression and shear wave zone indicating lower fusion

Figure 4:28 - TOFD scan images obtained for the weld produced by 900rpm and 100 mm/min

Figure 4:29 - TOFD scan images obtained for the weld produced by 1100rpm and 60 mm/min

109

Figure 4:30 - TOFD scan images obtained for the weld produced by 1100rpm and 80 mm/min

Figure 4:31: TOFD scan images obtained for the weld produced by 1100rpm and 100 mm/min

4.7.4 Digital X-ray Radiographic Weld Defects Analysis

The results of the radiographic examinations conducted on all the welds produced are presented in Table 4.5. The table contains the X-ray radiograph scan images and the respective comments on each examined sample. All the friction stir welds were 110

observed to be defect free from voids and porosity, which could be attributed to good material flow ability due to optimal heat energy. The presumed defect identified during the liquid penetrant testing was not noticeable. The presumed defect could have been either an uneven surface dent or misalignment on the parent metal not attributable to any friction stir processed weld flaws such as lack of bonding type defect.

Table 4.5: X-ray radiograph scan images of the weld samples

Rotational Feed rate matrix and weld physical Comments Speed appearance 60mm/min 80mm/min 100mm/min (rev/min) No evidence of root defects was observed 700 on all weld produced using the 700 rpm rotation speed and the respective feed rate speeds. Keyhole defects are observed.

111

No evidence of root defects was observed

on all weld produced using the 900 rpm 900 rotation speed with the 60 and 80 mm/min feed rates. An insignificant root defect can be seen for the weld produced with the feed rate of 100 mm/min. Keyhole defects are observed.

No evidence of root defects was observed

on all weld produced using the 1100 rpm 1100 rotation speed and the respective feed rate speeds. Keyhole defects are observed.

112

4.7.5 Ultrasonic Testing pulse-echo Couplant Attenuation Effect

Ultrasonic attenuation technique measurements have been used to analyze if conditions for forming kissing bonds are present or absent [10]. Kissing bonds predominantly occurs due to low penetration of the FSW rotating tool during processing of the weld. This condition thus prevents proper stirring within the root region. As such, the weld zone will have much finer grain size than the base material exuberated by material plasticizing of this weld region. Smaller grains are expected to be characterized by less ultrasonic attenuation also referred to as less acoustic “noise”. The signal processing principle is used to quantify the attenuation to determine if proper mixing and FSW has occurred.

In this study, different couplant were studied to investigate the attenuation effect. This approach was not used to actually detect the kissing bonds, however it does reliably detect the conditions under which kissing bonds occur [11]. This study was thus limited to the study of attenuation effects by comparing four different types of couplants employed.

Figures 4.32 – 4.38 show the different results of the frequency waveforms in time domain measured from the back wall echo obtained from scanning with the use of varying four couplants namely grease, water, oil and ultrasonic gel. The ultrasonic pulse-echo mode was applied during the investigation. The attenuation effect on the acoustic of the longitudinal waveform shows the influence of the couplant properties noticeable by the waveform peak frequencies for the re-coupling variations. These attenuation effects can be attributed to the coupling conditions such as the acoustic impedance values and the viscous properties of the couplants.

The waveform results are shown after setting the frequency for the single crystal probe at a higher frequency of 44.7dB. Figure 4.32 shows oil had the least gain compared to all the couplants when applied and a probe with a membrane used for scanning. This indicates that oil has the lowest attenuation coefficient at higher frequencies. Application of ultrasonic gel as shown in Figure 4.35, recorded the highest frequency gain value of -12.7dB from the set high frequency when the ultrasonic gel couplant was used with a probe attached with a membrane. The high frequency gain can be

113

attributed to the ultrasound gel having benefited from its good wetting and air removal properties. Air pockets are not conducive for wave transmission.

Ultrasonic gel, oil and particularly water with the frequency waveform demonstrated in Figure 4.38, provides the best frequency gain when applied and a probe without a membrane used for scanning as demonstrated by the frequency gain of -14.4dB. Conversely, grease provides poorer frequency gain than the ultrasound gel. This indicates that ultrasound gel has a lower attenuation coefficient at the higher frequencies than grease as both have comparable acoustic impedance. The ultrasound gel could as well benefited from good wetting and air removal properties. Water produced the highest frequency gain of -14.4dB. This can be attributed to water’s higher viscous properties which allows acoustic waves to travel without scattering and absorption.

The signal-to-noise ratio in frequency measurements can be adversely affected by poor selection of the couplant which results in distortion of the waveform being measured. It can be observed that noise was generated on the back-echo when a probe attached with a membrane was used with couplants such as, oil and ultrasonic gel. High magnitudes of noise were generated when probe with or without a membrane was used in the oil couplant application. This shows that the frequency spectrum can be characterized by various parameters such as the coupling conditions.

114

Figure 4:32 - UT frequency (y-axis) waveform in time (x-axis) domain scan done using Oil and Membrane attached to probe using and reference of 44.7 dB

Figure 4:33 - UT frequency (y-axis) waveform in time (x-axis) domain scan done using Oil and no membrane attached at a reference of 44.7 and showed a frequency gain of – 11.4 dB

115

Figure 4:34 - UT frequency (y-axis) waveform in time (x-axis) domain scan was done using Ultrasonic Gel with a membrane to the probe at a reference of 44.7 and showed frequency gain of – 5.6 dB

116

Figure 4:36 - UT frequency (y-axis) in time (x-axis) domain scan was done using Grease and a membrane attached to the probe using a reference of 44.7, results displayed a frequency gain of –1.5dB.

117

4.8 SUMMARY

The influence and effects of tool rotational and weld speeds on ultimate tensile strength, grain size and microhardness were discussed. It was observed that weld heat input is a derivative of rotational and weld speeds and influenced the integrity and quality of the welds produced.

Macrostructure and microstructural analysis were carried out. The weld’s stirred zone showed fine equiaxed grains that underwent full dynamic recrystallization as effect of highest temperature and highest deformations. The TMAZ had elongated grains considered to be along the flow line which had a grain structure that appeared to have been dynamically recovered due insufficient thermal and deformation conditions which were not enough to cause grain structure recrystallization. There was a slight change of grain morphology in the HAZ due thermal gradient. The thermal gradient and heat energy effects caused by the processes parameters were demonstrated to have influenced the grain size and grain boundaries in the various weld regions.

118

Tensile tests were conducted and average ultimate tensile strength tabulated. The base metal had ultimate tensile and yield strength values higher than that of the weld region. It was observed that increasing the weld speeds improved the increase of the tensile and yields strength towards the values of the base metal. The results showed that the increase of the tensile strength was uniform to the increase of the yield strength. The increase in mechanical properties was attributed to the attainment of heterogeneously distributed precipitates, dynamic recrystallization and the absence of weld defects. The tensile tests conducted showed ductile failure mode in the weld’s weakest zone, predominantly the TMAZ/HAZ interface.

Non-destructive testing imaging was successfully used and no defects were observed on all the welds. Selection of suitable process speeds aided in optimum generation of heat energy sufficient to plasticize the material thus contributing to improved material flow that achieved uniform microhardness across the weld traverse. Higher rotational speeds and higher feed rates resulted in increased hardness values, with the highest values recorded in the weld nugget and the least values recorded in the SZ/TMAZ interface. The least values in the interface were attributed to the material softening and decomposition of precipitates, while the high values were attributed to dynamic recrystallization leading to fine grains, and more-so, strain-hardening due to the pinning effect of the tool particularly in the weld nugget. This is seen defined by a W- shaped hardness distribution data across the welds traverse.

Four different couplants were investigated for attenuation effects during weld defects analysis using the ultrasonic pulse-echo mode. These attenuation effects were observed to be attributed to the coupling conditions such as the acoustic impedance values and the viscous properties of the couplants.

119

5. CHAPTER FIVE - CONCLUSION AND FUTURE WORK

5.1 SUMMARY OF RESEARCH WORK

The aim of this study was to assess, evaluate and characterize the microstructural integrity within the weld region of the friction stir welded AA6061-T6 butt joint and to evaluate the attenuation effect of sound waves during contact ultrasonic testing of the weld defects examination. Additional non-destructive tests such X-ray radiography, UT phased array and TOFD were conducted during the course of the research to validate the weld defects analysis results obtained using contact pulse-echo ultrasonic testing.

The role of non-destructive testing in the examination of weld defects and material flow conditions was discussed. The knowledge gap identified was that there is no previous study on UT pulse-echo attenuation effects during weld defects analysis specifically in the friction stir welding field.

5.2 CONCLUSION

In this research it was demonstrated that 4.09 mm AA6061-T6 can be welded using FSW process with a concave shoulder tool having a cylindrical pin length of 3.8mm constructed from tool steel material W302.

The base metal showed higher mechanical properties when compared to all the weld regions which showed degradation in tensile strength, ductility and hardness values. Higher rotational speeds and higher feed rates result in increased hardness values, with the highest values recorded in the weld nugget and the least values recorded in the SZ/TMAZ interfaces. This was defined by a W-shaped hardness distribution data across the welds traverse. Selection of suitable process speeds led to optimum generation of heat energy sufficient to plasticize the material, thus, contributing to improved material flow to achieve uniform microhardness across the weld traverse.

All welds showed no presence of weld defects. This was attributed to optimum process parameters that gave complete homogeneous FSW which improved mechanical 120

workings and material mixing of the plasticized AA6061-T6 aluminium alloy. The samples with reduced hardness values specifically 900rpm (for both 80 and 100mm/min) revealed fusion problems. This was attributed to thermal variations at fast cooling rates. The degradation to the hardness was due to decomposition of the precipitates necessitated by the thermal cycles. These reduced values were local to the TMAZ/HAZ interface. Ductile failure mode was local to this weakest region (TMAZ/HAZ). This region corresponds with the lowest hardness values that can be explained by softened material with larger grain sizes not as fine when compared to the weld nugget.

The couplant attenuation effect was minimal when the Ultrasonic gel, oil and water were applied with the UT pulse echo longitudinal wave. This was attributed to the coupling conditions such as the acoustic impedance values and the viscous properties of the couplants. The high frequency gain was attributed to the ultrasound gel having benefited from its good wetting and air removal properties. Conversely, grease provided poorer frequency gain than all the couplants. This indicates that ultrasonic gel had lower attenuation coefficient at the higher frequencies than grease as both have comparable acoustic impedance.

In addition, grease has poor wetting conditions, which creates air pockets that are not conducive for wave transmission. Correct non-destructive testing procedure carried out in accordance to specifications is crucial in weld defect analysis. All non- destructive testing used were successful in detecting no weld defects. On the other hand, UT TOFD proved to be a good technique in analysing volumetric defects due to material mixing, which in this study, identified lack of weld fusion which in samples which collated with the lowest hardness values.

5.3 FUTURE WORK

This research made great strides towards achieving the objectives of the research, however, some areas require further and expanded study:-

121

 The process window needs to be widened by varying the tool tilt angle and tool design in order to correlate detailed microstructure evolution and the presence of weld defects.  The UT TOFD was able to detect lack of weld fusion. A comprehensive study in material flow needs to be investigated using the TOFD technique and evaluated against other established material flow analysis techniques such as the marker method and finite element method.  Residual stress and fatigue life needs to be evaluated to determine the influence of process parameters. This is important in understanding the endurance cycle of FSW butt welded AA6061 aluminium alloy of a wide thickness spectrum.  A CFD simulation should be considered to study the thermo-flow phenomena at the plunge, stabilised and retraction points.  An understanding on the influence of thermal gradients needs to be derived using finite determination of welding temperature in time and distance domain to determine the influence of process parameters on UTS and hardness (HV) on AA6061-T6.  This study need to be extended to the research for fatigue optimization and fracture mechanics of AA6061-T6 to deepen the ductile fracture mode interpreted in this study.  During the study, revealing of AA6061-T6 microstructure using SEM investigation was a challenge due to unconfirmed development of a passive layer on the samples once certain types of etchants were applied. A correct etchant procedure need to be investigated and the results would benefit future students.

122

6. REFERENCES

[1] Wikipedia, Friction Stir Welding, https://en.wikipedia.org/wiki/Friction_stir_welding. [2] U. Patent, The Welding Institute, 6 December 1991. [Online]. Available: https://www.google.com/patents/US5460317#backward-citations. [3] W. Arbegast and P. Hartley, "Friction stir welding technology development at Lockheed Martin Michoud Space system - An Overview," in ASM International, Georgia, US, 1998. [4] E. T. Akinlabi, R. M. Mahamood, S. A. Akinlabi and E. Ogunmuyiwa, "Processing Parameters Influence on Wear Resistance Behaviour," Advances in Materials Science and Engineering, vol. 2014, no. Article ID 724590, pp. 1- 2, 2014. [5] B. Singh, in A hanbook on Friction Stir Welding, 978-3-659-10762-7,, UK, LAP Lambert Academic Publishing, UK, June 2012, pp. 44-54. [6] B. Kumar, V. Umapathi and J. Kumar, "A comparative study of mechanical properties and microstructures of FSW joints with Conventional welded joints," International Journal of Engineering Trends and Technology, vol. 35, no. 6, 2016. [7] M. Patil, "MTech Dissertation on Study of effect of process parameters in friction stir welding of 5083 Al alloy on power input and micro hardness contours in weld cross-sections," https://www.scribd.com/document/233479059/Friction-Stir-Welding-Mahesh- Thesis-10ME61R09, Indian Institute of Technology, Kharagpur, 2012. [8] N. Khan, A. Siddiquee and S. Shihab, "Investigations on tunneling and kissing bond defects in FSW joints for dissimilar aluminum alloys," Alloys and Compounds, vol. 648, pp. 360-367, November 2015. [9] P. Kah, R. Rajan, K. Martikainen and R. Suoranta, "Investigation of weld defects in friction-stir welding and fusion welding of aluminium alloys,"

123

International Journal of Mechanical and Materials Science, vol. 4, no. ISSN: 1823-0334, December 2016. [10] H. Chena, K. Yanb, L. Tao, S. Chena, C. Jiangb and Y. Y. Zhaob, "The investigation of typical welding defects for 5456 aluminum alloy friction stir welds," Elsivier Material Science and Engineering, vol. 433, no. 1-2, pp. 64- 69, October 2006. [11] M. Kathrotiya and K. Bhavsar, "Various Parameter Effects on Friction Stir Welding," International Research Journal of Engineering and Technology, vol. 04, no. 02, Feb 2017. [12] C. Devanathan and S. Babu, "Effect of Plunge Depth on Friction Stir Welding of Al 6063," in 2nd International Conference on Advanced, INCAMA, 2013. [13] S. Bayazida and H. Farhangia, "Effect of Pin Profile on Defects of Friction Stir Welded 7075 Aluminum Alloy," Elsevier Material Science, vol. 11, pp. 12-16, 2015. [14] C. Brook and C. Hanstead, "Impact of Non Destructive Testing," in Proceeding on the 28th Annual British Conference of NDT, Bristol, UK, September 1989. [15] P. Kah, B. Mvola, J. Martikainen and R. Suoranta, "Real Time non destructive testing of welding," Trans Tech Publications Advanced Material Research, vol. 933, no. 01, pp. 109-116, 2013. [16] C. Bird, "Ultrasonic phased array inspection technology for the evaluation of friction stir welds," The British Institute of Non-Destructive Testing and Condition Monitoring, vol. 46, no. 1, pp. 31-36, 1 January 2004. [17] C. Meola, A. Squillace, F. Memola, C. Minutolo and R. Morace, "Analysis of stainless steel welded joints: a comparison between destructive and non- destructive techniques," Elsevier Material Processing Technology, Vols. 155- 156, pp. 1893-1899, 2004. [18] B. Li, Y. Shen and W. Hu, "The study on defects in aluminum 2219-T6 thick butt friction stir welds with the application of multiple non-destructive testing methods," Elsevier Materials and Design, vol. 32, no. 4, p. 2073–2084, April 2011.

124

[19] S. Rajakumar and V. Balasubramanian, "Establishing relationships between mechanical properties of aluminium alloys and optimised friction stir welding process parameters," Elsevier Materials & Design, vol. 40, pp. 17-35, September 2012. [20] C. Genevoisa, A. Deschampsa, A. Denquinb and B. Doisneau-cottigniesa, "Quantitative investigation of precipitation and mechanical behaviour for AA2024 friction stir welds," Elsevier Acta Materialia, vol. 53, no. 8, p. 2447– 2458, May 2005. [21] "Joining of Al 6061 alloy to AISI 1018 steel by combined effects of fusion and solid state welding," Elsevier Science Inc, Machine Tools and Manufacture, vol. 44, no. 11, p. 1205–1214, September 2004. [22] S. Babu, S. Pavithran, M. Nithin and B. Parameshwaran, "Effect of Tool Shoulder Diameter During Friction Stir Processing of AZ31B Alloy Sheets of various Thicknesses," Elsevier Science Inc Procedia Engineering, vol. 97, pp. 800-809, 2014. [23] V. Godiganur and S. Biradar, "Comparison of Friction Stirs Welding Technique with Conventional welding methods," International Journal of Research in Engineering and Technology, vol. 03, no. 04, May 2014. [24] A. Squillace, A. D. Fenzo, G. Giorleo and F. Bellucci, "International Journal of Research in Engineering and Technology," Elsevier Science Inc Journal of Materials Processing Technology, vol. 152, no. 1, pp. 97-105, 1 October 2004. [25] J. Guo, "Solid State Welding Processes in Manufacturing," in Handbook of Manufacturing Engineering and Technology, Springer Research & development, September 2014, pp. 569-592. [26] M. A. Bodude and I. Momohjimoh, "Studies on Effects of Welding Parameters on the Mechanical Properties of Welded Low-Carbon Steel," Science research Publication Journal of Minerals and Materials Characterization and Engineering, vol. 3, no. 1, pp. 142-153, 31 March 2015. [27] T. S. Kumara, V. Balasubramanianb and M. Sanavullahc, "Influences of pulsed current tungsten inert gas welding parameters on the tensile properties of AA 6061 aluminium alloy," Elsevier Science Inc Materials & Design, vol. 28, no. 7, p. 2080–2092, 2007.

125

[28] J. Lancaster, "The physics of welding," International Conference of Advanced Materials Physics in Technology, vol. 15, no. 2, 1984. [29] "Fusion Welding Processes Vol 3, 7th ed," in Welding Handbook, Miami, American Welding Society, 1980, p. 170–238. [30] B. Shivakumar and S. Vishwanath, "Comparison of Friction Stir Welding technique with conventional welding techniques". [31] E. Akinlabi, "Doctoral Thesis on Characterisation of dissimilar Friction Stir Welds Between 5754 Aluminium Alloy and C11000 Copper," Nelson Mandela Metropolitan University, Port Elizabeth, South Africa, December 2010. [32] E. A. Association, "Joining dissimilar materials," The Aluminium Automotive Manual, pp. 1-31, 2015. [33] S. Kalpakjian and S. Schmid, "Solid State Welding Processes," in “Manufacturing Engineering and Technology, 7th Edition, Singapore, Pearson Publications ISBN-13 : 9780133128741, 2015, p. Chapter 3. [34] X. Liu, C. Wu and G. K. Padhy, "Characterization of plastic deformation and material flow in ultrasonic vibration enhanced friction stir welding," Elsevier Science Inc Scripta Materialia, vol. 102, pp. 95-98, June 2015. [35] R. Rai, A. De, H. K. D. H. Bhadeshia and T. DebRoy, "Review: friction stir welding tools," Institute of Materials, Minerals and Mining Science and Technology of Welding and Joining, vol. 14, no. 4, February 2011. [36] P. Asadi, K. Choobi and A. Elhami, Welding of Magnesium Alloys Chapter 6, Tehran, Iran: ISBN 978-953-51, 2012. [37] S. Mandal, J. Rice and A. Elmustafa, "Experimental and numerical investigation of the plunge stage in friction stir welding," Elsevier Science Inc , vol. 203, no. 1-3, pp. 411-419, 18 July 2008. [38] L. Fourment and S. Guerdoux, "3D numerical simulation of the three stages of Friction". [39] P. Sahu and S. Pal, "Multi-response optimization of process parameters in friction stir welded AM20 magnesium alloy by Taguchi grey relational analysis," Elsevier Science Inc Journal of Magnesium and Alloys, vol. 3, no. 1, p. 36–46, March 2015.

126

[40] B. London, M. Mahoney, B. Bingel, M. Calabrese and D. Waldron, "Proceedings of the Third International Symposium on Friction Stir Welding," Kobe, Japan, September 2001. [41] R. Mishra and Z. Ma, "Friction stir welding and processing," Elsevier Science Inc Material Science & Engineering, vol. 50, pp. 1-78, August 2005. [42] X. Hea, F. Gub and A. Ballb, "A review of numerical analysis of friction stir welding," Elsevier Science Inc Progress in Material Science, vol. 65, pp. 1-66, August 2014. [43] R. Mishra, P. De and N. Kumar, Springer International Science and Engineering Friction stir welding and processing, p. 347pp, 2014. [44] P. Rambabu, N. E. Prasad, V. Kutumbarao and R. Wanhill, "Chapter 2," in Aluminium Alloys for Aerospace Applications, Singapore, Springer Science, 2017, pp. 29-52. [45] P. Mandez and T. Eager, "Welding Processes for Aeronautics," ASM International Advanced Material & Processes, pp. 39-43, May 2001. [46] D. Lohwasser and . Chen, "Aerospace," in Friction Stir Welding: From Basics to Applications, Cambridge, UK, Woodhead Publishing, 2009, pp. 230-231. [47] B. Halverson and J. Hinrichs, "Friction Stir Welding (FSW) of Littoral Combat Ship Deckhouse Structure," American Institute of Shipping, vol. Friction Stir Link Inc. [48] X. Sun, "Parameter Effects," in Failure Mechanisms of Advanced Welding Processes, Elsevier Science Inc Technology & Engineering, 2010, pp. 153- 154. [49] R. Mishra and H. Sidhar, in Friction Stir Welding of 2XXX Aluminum Alloys including Al-Li Alloys, ISBN: 128092920, Butterworth-Heinemann, Technology & Engineering, Oct 2010, p. 110. [50] M. J. Zehetbauer and Y. Zhu, "Friction Stir Processing: Features & Properties," in Bulk Nanostructured Materials, Erlangen, Germany, Wiley Publishers, July 2009, pp. 260-288.

127

[51] M. Rao, B. Kumar and M. Hussain, "Friction Stir Welding: Process and its Variables - A Review," in 9th International Conference of Recent Innovations in Science, Engineering & Management, ISBN:978-93-86171-22-1, 2017. [52] FSW-Technical-Handbook, Welding Automation, LAXÅ, Sweden: ESAB AB, www.esab.com.. [53] R. Leala, C. L. A. Loureiroa, D. Rodriguesa and P. Vilaçac, "Material flow in heterogeneous friction stir welding of thin aluminium sheets: Effect of shoulder geometry," Elsevier Science Inc Material Science & Engineering, vol. 498, no. 1-2, p. 384–391, 20 December 2008. [54] S. Malarvizhi and V. Balasubramanian, "Influences of tool shoulder diameter to plate thickness ratio (D/T) on stir zone formation and tensile properties of friction stir welded dissimilar joints of AA6061 aluminum–AZ31B magnesium alloys," Elsevier Science Inc Materials & Design, vol. 40, p. 453–460, September 2012. [55] D. Burford, B. Tweedy and C. Widener, "Influence of shoulder configuration and geometric features on FSW track properties," in Proceedings of the 6th International Symposium, Montreal, Canada, October 2016. [56] I. Galvão, J. Oliveira, A. Loureiro and D. Rodrigues, "Formation and distribution of brittle structures in friction stir welding of aluminium and copper: Influence of shoulder geometry," Elsevier Science Inc Intermetallics, vol. 22, pp. 122-128, March 2012. [57] M. D. Giorgi, A. Scialpi, F. W. PanellaL and A. C. D. Filippis, "Effect of shoulder geometry on residual stress and fatigue properties of AA6082 fsw joints," Springer Science Journal of Mechanical Science and Technology, vol. 23, no. 1, pp. 26-35, January 2009. [58] I. Galvãoa, R. Leal, D. Rodriguesa and A. Loureiroa, "Influence of tool shoulder geometry on properties of friction stir welds in thin copper sheets," Elsevier Science Inc Journal of Materials Processing Technology, vol. 213, no. 2, pp. 129-135, February 2013. [59] I. Fridlyander, "Aluminium alloys with lithium and magnesium," Metal Science Heat treatment, vol. 45, pp. 344-347, 2003.

128

[60] A. Scialpia, L. D. Filippisa and P. Cavaliereb, "Influence of shoulder geometry on microstructure and mechanical properties of friction stir welded 6082 aluminium alloy," Elsevier Science Inc Materials & Design, vol. 28, no. 4, pp. 1124-1129, 2007. [61] M. Said, D. Hamid, A. Ismail, M. Rojan, M. Zainuddin, B. Baharudin and A. Ali, "The Effect of Pin Size on Friction Stir Welded AA5083 Plate Lap Joint.," in Int'l Conference on Production, Automobiles and Mechanical Engineering, Kuala Lumpur, Malaysia , 2015. [62] E. Akinlabi and K. Surekha, "Surface engineering by friction stir processing and friction surfacing.," in Surface Engineering Techniques and Applications: Research Advancements, 2014, p. 44. [63] S. Jia, Q. Shi, L. Zhang, A. Zou, S. Gao and L. Zan, "Numerical simulation of material flow behavior of friction stir welding influenced by rotational tool geometry," Elsevier Science Inc Journal of the Computational Materials Science, vol. 63, pp. 218-226, October 2012. [64] R. Nandan, T. Debboy and H. Badheshia, "Recent advancement in friction stir welding process, Weldment structure and properties," Progress in Material Science, vol. 53, pp. 980-1023, 2008. [65] W. Thomas, A. Braithwaite and R. John, "Skew Stir Technology," in 3rd International FSW Symposium, Kobe, Japan, September 2001. [66] E. Akinlabi, A. Andrews and S. Akinlabi, "Effects of processing parameters on corrosion properties of dissimilar friction stir welds of aluminium and copper," Elsevier Science Press Transactions of Nonferrous Metals Society of China, vol. 24, p. 1323−1330 , 2014 . [67] I. Galvão, J. Oliveira, A. Loureiro and D. Rodrigues, "Formation and distribution of brittle structures in friction stir welding of aluminium and copper: Influence of shoulder geometry," Elsevier Science Inc Journal of Intermetallics, vol. 22, pp. 122-128, 2012. [68] N. Muhayat, M. Z. Yuliadi and A. Zubaydi, "Effect of Tool Tilt Angle and Tool Plunge Depth on Mechanical Properties of Friction Stir Welded AA 5083 Joints," Transtech Publication Advances in Applied Mechanics and Materials, vol. 493, pp. 709-714, January 2014.

129

[69] D. Lohwasser and Z. Chen, Friction stir welding, Cambridge, UK: Woodhead Publishing Ltd, 2009. [70] P. Kush and V. Bhadheka, "Effects of Tilt Angle on the Properties of Dissimilar Friction Stir Welding Copper to Aluminum," Taylor Francis Online Journal of Materials and Manufacturing Processes, vol. 31, no. 3, pp. Pages 255-263, December 2014. [71] S. S. Kumar, S. Ashok and S. Narayanan, "Investigation of Friction Stir Butt Welded Aluminium Alloy Flat," Elsevier Science Inc International Conference on DESIGN AND MANUFACTURING, vol. 64, p. 915 – 925, 2013. [72] A. Pradeep, S. Muthukumaran, P. Dhanush and T. Nakkeran, "A Study on Friction Stir Welding of Low Alloy steel by using a tool with a conical pin," International Journal of Mechanical and Materials Engineering, vol. 8, no. 2, p. 2013, 132-137. [73] G. Krishna, P. Reddy and M. Hussain, "Effect of Tool Tilt Angle on Aluminum 2014 Friction Stir Welding," Global Journal of Researches in Engineering, vol. 14, no. 7, 2014. [74] T. Shinoda, "Effect of tool angle on metal flow phenomenon in friction stir weld," in 3rd International FSW Symposium, Kobe, Japan, 2001. [75] S. Celik and R. Cakir, "Effect of Friction Stir Welding Parameters on the Mechanical and Microstructure Properties of the Al-Cu Butt Joint," MDPI Journal of Metals, vol. 6, no. 6, p. 133, 31 May 2016. [76] S. Ugender, A. Kumar and A. Reddy, "Influence of Friction Stir Welding Parameters on Mechanical Properties of 6061- T6 Aluminum Alloy," Research & Reviews: Journal of Engineering and Technology, Vols. ISSN: 2319-9873, 2014. [77] M. Tehyo, P. Muangjunburee, A. Binraheem, S. Chuchom and N. Utamarat, "Influence of friction stir welding parameters on metallurgical and mechanical properties of dissimilar joint between semi-solid metal 356-T6 and aluminum alloys 6061-T651," Songklanakarin Journal of Science and Technology, vol. 34, no. 4, pp. 415-421, August 2012.

130

[78] D. Neto and P. Neto, "Numerical modeling of friction stir welding process: a literature review," Springer, The International Journal of Advanced Manufacturing Technology, vol. 65, no. 1, pp. 115-126, March 2013. [79] K. Vepakomma, "Three Dimensional Thermal Modeling of Friction Stir Processing," Master's Dissertation, The Florida State University, Florida, USA, 2006. [80] M. Melendez, W. Tang, C. Schmidt, J. C. McClure, A. C. Nunes and L. E. Murr, "Tool Forces developed during Friction Stir welding," NASA Marshall Space Flight Center, Huntsville, Alabama, 2005. [81] P. Shahi, M. Barmouz, T. Modares and P. Asadi, "Force and torque in friction stir welding," in Advances in Friction Stir Welding and Processing, Woodhead Publishers, 2014, pp. 459-498. [82] D. Trimble, J. Monaghan and G. O'Donnell, "Force generation during friction stir welding of AA2024-T3," CIRP Annals - Manufacturing Technology, vol. 61, no. 1, p. 9–12, 2012. [83] X. Hea and A. B. F. Gub, "A review of numerical analysis of friction stir welding," Elsevier Science In Journal of Progress in Materials Science, vol. 65, pp. 1-66, August 2014. [84] P. Threadgill, A. Leonard, H. Shercliff and P. Withers, "Friction stir welding of aluminium alloys," International Materials Reviews, vol. 54, no. 2, pp. 49-93., March 2009. [85] K. Colligan, "Material Flow Behavior during Friction Stir," American Welding Society and the Welding Research Council - The Welding Journal, New Orleans, Los Angelos, July 1999. [86] T. Dickerson, H. Shercliff and H. Schmidt, "A Weld Marker Technique for Flow Visualization in FSW," in 4th International Symposium on Friction Stir Welding, Park City, Utah, USA, 14-16 May 2003. [87] B. S. Roy, T. Medhi and S. C. Saha, "Material Flow Modeling in Friction Stir Welding of AA6061-T6 Alloy and Study of the Effect of Process Parameters," International Journal of Environmental, Chemical, Ecological, Geological and Geophysical Engineering, vol. 9, no. 6, pp. 658-666, 2015.

131

[88] A. Reynolds and T. Seidel, "Visualisation of the material flow in AA2195 T8 FSW using Marker Technique," Springer Journals of Metallurgical and Materials Transactions, vol. 32, no. 11, p. 2879–2884, Nov 2001. [89] R. K. Uyyuru and S. Kailas, "Numerical analysis of friction stir welding process," Springer Publications Journal of Materials Engineering and Performance, vol. 15, no. 5, p. 505–518, October 2006. [90] N. Dialami, M. Chiumenti and M. Cervera, "Material Flow Visualization in Friction Stir Welding vs Particle Tracing," International Center for Numerical Methods in Engineering, Barcelona, Spain., 2013. [91] J. Chen, H. Zhang and Z. Zhang, "3D modeling of material flow in friction stir welding under different process parameters," Elsevier Inc Journal of Material Processing Technology, vol. 183, no. 1, pp. 62-70, 5 March 2007. [92] D. Devaiah, K. Kishore and P.Laxminarayana, "Effect of Welding Speed on Mechanical Properties of Dissimilar Friction Stir Welded AA5083-H321 and AA6061-T6 Aluminum Alloys," International Journal of Advanced Engineering Research and Science, vol. 4, no. 3, Mar- 2017. [93] K. Oyindamola, "Master's Dissertation on Effects of Multi-Pass Friction Stir Processing on Aluminium," University of Johannesburg, Johannesburg, SA, 2016. [94] ASM International, [Online]. Available: http://products.asminternational.org/hbk/content/...... /0082748.jpg. [Accessed 23 March 2017]. [95] Z. Chen, S. Li and L. Hihara, "Microstructure, mechanical properties and corrosion of friction stir welded 6061 Aluminum Alloy," Hawaii Corrosion Laboratory, Department of Mechanical Engineering, University of Hawaii , Manoa, Honolulu, HI 96822, USA. [96] C. Shi, J. Lai and X. Chen, "Microstructural Evolution and Dynamic Softening Mechanisms of Al-Zn-Mg-Cu Alloy during Hot Compressive Deformation," MDPI Journal of Materials, vol. 7, no. 1, pp. 244-264, 2014. [97] R. Mishra and N. Kumar, Friction stir welding and processing, Switzerland: Springer International Publishing, 2014, p. 347.

132

[98] H. Kokawa, Y. Sato, M. Enmoto and S. Jogan, "Metals," Materials Transaction, vol. 30, no. 2429, 1999. [99] P. Prangnell and C. Heason, "Grain structure formation during friction stir welding observed by the stop action technique," Elsevier Science Inc Acta Materialia , vol. 53, pp. 3179-3192, 2005. [100] S. K. Tiwari, D. Shukla and R. Chandra, "Friction Stir Welding of Aluminum Alloys: A Review," World Science Academy International Journal of Mechanical, Aerospace, Industrial, Mechatronic and Manufacturing Engineering , vol. 7, no. 12, 2013. [101] "AA8090 Al-Li Alloy FSW parameters to minimize defects and increase fatigue life," Springer Publications International Journal of Material Forming, vol. 3, no. 1, p. 1003–1006, April 2010. [102] O. Salih, H. Ou, W. Sun and D. McCartney, "A review of friction stir welding of aluminium matrix composites," Elsevier Science Inc Journal of Materials & Design, vol. 86, no. 5, pp. 61-71, 5 December 2015. [103] F. Fadaeifard, K. Matori, S. Aziz, L. Zolkarnain and M. B. A. Rahim, "Effect of the Welding Speed on the Macrostructure Microstructure and Mechanical Properties of AA6061-T6 Friction Stir Butt Welds," MDPI Journal of Metals, vol. 7, no. 48, pp. 2-16, 2017. [104] C. Rhodes, M. Mahoney, W. Bingel, R. Spurling and C. Bampton, Scripta Material, vol. 36, no. 69, 1997. [105] G. Liu, L. Murr, C. Niou and J. McClure, Scripta Material, vol. 26, no. 355, 1997. [106] K. Jata and S. Semitian, Scripta Material, vol. 43, no. 743, 2000. [107] Z. Ma, R. Mishra and M. Mahoney, Acta Material, vol. 50, no. 4419, 2002. [108] Y. Sato, M. Urata and H. Kokawa, MDPI Material Transactions, vol. 33, no. 625, 2002. [109] B. Heinz and B. Strotzki, Metallurgical Material Transactions, vol. 6, no. 489, 2002. [110] L. Murr, G. Liu and J. McClaure, Journal of Material Letrters, vol. 16, no. 1081, 1997.

133

[111] G. Frankel and Z. Xia, Journal of Corrosion, vol. 55, no. 139, 1999. [112] Y. Sato, S. Park and H. Kokawa, Metallurgy Journal of Material Transactions, vol. 32, no. 3023, 2001. [113] S. Kazi and L. Murr, "Friction stir welding Processing," PA, USA, TMS Warrendale, 2001, p. 139. [114] H. Salem, A. Reynolds and J. Lyons, Scripta Material, vol. 46, no. 337, 2002. [115] R. Braun and L.Litynska-Dobrzynska, Journal of Material Science Forum, no. 1531, pp. 396-402, 2002. [116] A. Norman, P. Prangell and I. Brough, Journal of Material Science Forum, no. 1713, pp. 331-337, 2002. [117] A. Norman, P. Prangell and K. Hassan, Material Science Forum, no. 1549, pp. 396-402, 2002. [118] J. Sun, T. Nelson, R. Mishra and M. Mahoney, Acta Material, vol. 51, no. 713, 2002. [119] R. Mishra and I. Charit, Material Science Engineering, vol. 359, no. 290, 2003. [120] I. Charit, R. Mishra and M. Mahoney, Scripta Material, vol. 47, no. 631, 2002. [121] I. Charit, Z. Ma and R. Mishra, in Hot Deformation of Aluminium Alloys III, TMS, 2003, pp. 331-342. [122] P. Pao, E. Lee, C. Feng, H. Jones and D. Moon, "Friction Stir Welding Processing II," Warrendale, PA, USA, TMS, 2003, p. 113. [123] C. Rhodes, M. Mahoney, W. Bingel, R. Spurling and C. Bampton, "Effects of friction stir welding on microstructure of 7075 aluminum," Elsevier Science Inc Scripta Materialia, vol. 36, no. 1, pp. 69-75, 1 January 1997. [124] H. McQueen, M. Cabibbo, E. Evangelista, S. Spigorell and M. Paoski, "Microstructure and Mechanical Properties of AA6056 friction stir welded," Elsevier Science Inc Journal of Material Science & Engineering, Vols. 460- 461, pp. 86-94, July 2007. [125] X. Sauvage, A. Dédé, A. C. Muñoz and B. Huneau, "Precipitate stability and recrystallisation in the weld nuggets of friction stir welded Al-Mg-Si and Al-Mg- Sc alloys.," Elsvier Science Inc Material Science Engineering, vol. 491, pp. 364-371, 2008.

134

[126] C. A. W. Olea, L. Roldo, T. R. Strohaecker and J. F. d. Santos, "Friction Stir Welding of Precipitate Hardenable Aluminium Alloys: A Review," Springer Journal for Research & Development, vol. 50, no. 11, p. 78–87, November 2006. [127] Y. Chen, J. Feng and H. Liu, "Precipitate evolution in friction stir welding of 2219-T6 aluminum alloys," Elsevier Science Inc Journal of Material Characterization, vol. 60, no. 6, pp. 476-481, June 2009. [128] H. Liua, J. Houa and H. Guoa, "Effect of welding speed on microstructure and mechanical properties of self-reacting friction stir welded 6061-T6 aluminum alloy," Elsevier In Journal of Materials & Design, vol. Volume 50, p. 872–878, September 2013. [129] L. E. Murr, G. Liu and J. C. McClure, "A TEM study of precipitation and related microstructures in friction-stir-welded 6061 aluminium," Spinger Journal of Material Science, vol. 33, no. 5, p. 1243–1251, March 1998. [130] G. Liu, J. McClure and F. Vega, "Microstructural aspects of the friction-stir welding of 6061-T6 aluminum," Elsevier Science Inc Journal of Scripta Materialia, vol. 37, no. 3, pp. 355-361, August 1997. [131] W. Lee, Y. Yeon and S. Jung, "Evaluation of the microstructure and mechanical properties of friction stir welded 6005 aluminum alloy," Taylor & Francis Journal of Materials Science and Engineering, vol. 19, no. 11, pp. 1513-1518, 2003. [132] A. Denquin, "Metallurgy and Friction Stir Welding," Onera Department of Metallic Structures and Materials, Onera, Châtillon, France, 2009. [133] V. Balasubramanian, "Relationship between base metal properties and friction stir welding process parameters," Elsevier Science In - Materials Science and Engineering: A Review, vol. 480, no. 1-2, p. 397–403, 15 May 2008. [134] Y.-H. Zhao, S. Lin, Z. He and L. Wu, "Microhardness prediction in friction stir welding of 2014 aluminium alloy," Taylor & Francis Journal of Science and Technology of Welding and Joining , vol. 11, no. 2, pp. 178-182, 2006. [135] L. D. Filippis, L. Serio, F. Facchini, G. Mummolo and A. Ludovico, "Prediction of the Vickers Microhardness and Ultimate Tensile Strength of AA5754 H111

135

Friction Stir Welding Butt Joints Using Artificial Neural Network," MDPI Journal of Materials, vol. 9, no. 11, 2016. [136] P. Cavalierea, A. Squillace and F. Panellaa, "Effect of welding parameters on mechanical and microstructural properties of AA6082 joints produced by friction stir welding," Elsevier Science Inc Journal of Materials Processing Technology, vol. 200, no. 1-3, pp. 364-372, May 2008. [137] J. Leon and V. Jayakumar, "Investigation of Mechanical Properties of Aluminium 6061 Alloy Friction Stir Welding," International Journal of Students’ Research in Technology & Management, Vols. 2 ISSN 2321-2543, no. 04, pp. 140-144, 2014. [138] D. Rodriguesa, A. Loureiroa, C. Leitaoa, R. Leala, B. C. b and P. Vilaçad, "Influence of friction stir welding parameters on the microstructural and mechanical properties of AA 6016-T4 thin welds," Elsevier Science Inc Journal of Material Science, vol. 30, no. 6, p. 1913–1921, June 2009. [139] P. Podrazaj, B. Jerman and D. Kloncar, "Welding Defects at Friction stir welding," Metalurgija ISSN 0543-5846, vol. 54, no. 2, pp. 387-389, 2015. [140] H. Chen, K. Yan, T. Lin, S. Chen, C. Jiang and Y. Zhao, "The investigation of typical welding defects for 5456 aluminum alloy friction stir welds," Elsevier Science Inc Materials Science and Engineering: A, vol. 433, no. 1-2, pp. 64- 69, October 2006. [141] Y. Huang, L. Wan, T. Huang and &. L.Zongliang, "The weld formation of self- support friction stir welds for aluminum hollow extrusion," Springer International Journal on Adv Manuf Technol, 2016. [142] P. Kah, R. Martikainen and R. Suoranta, "Investigation of weld defects in friction-stir welding and fusion welding of aluminium alloys," Springer International Journal of Mechanical and Materials Engineering, December 2015. [143] S. Lockeyer and A. Leonard, "Flaws in Friction Stir Welds," in 4th International Symposium on Friction Stir Welding, Park City, Utah, USA, 2003. [144] P. Theobald, B. Zeqiri and J. Avison, "Couplants and their influence on AE sensor sensitivity," in 28th European Conference on AE Testing, Middlesex, UK, 2008. 136

[145] M. Larson, "Ten Criteria for Accurate Ultrasonic Couplant Selection," Springer Publication, Bellingham, Washington, 2009. [146] Y. Kim and S. Song, "A study on the couplant effects in contact ultrasonic testing," in 10th Annual Conference Proceedings NDT, Brisbane, Australia, September 2011. [147] N. Netshidavhini and R. Mabuza, "Effects of Various Couplants on Carbon Steel and Aluminium Materials Using Ultrasonic Testing," in 18th World Conference on Nondestructive Testing, Durban, South Africa, 20 April 2012. [148] J. Drury, Ultrasonic Flaw detection for technicians, OIS Power & Nuclear Division, 1992. [149] L. Kinsler, A. Coppens, A. Frey and J. Sanders, "Fundamentals of Acoustics," New York, USA, John Wiley & Sons, 1980, pp. 124-134. [150] T. Young. [Online]. Available: https://en.wikipedia.org/wiki/Interference_(wave_propagation). [Accessed 2017 March 15]. [151] Olympus Scientific Solution, [Online]. Available: https://www.olympus- ims.com/en/ndt-tutorials/intro/breif-history/. [Accessed 15 March 2017]. [152] J. Dziewierz, R. L. O'Leary, A. Gachagan, A. Velichko and P. D. Wilcox, "An annular array with fiber composite microstructure for far field NDT imaging applications," in Ultrasonics Symposium (IUS), 2009 IEEE International, Rome, Italy, 20-23 Sept 2009. [153] N. Gaurav, K. Manu and M. Sivashanmugam., "Digital Radiography Test for the GTAW and FSW Weldments on AA 7075-T651 Aluminium Alloy," Research india Pulblications International Journal of Advanced Mechanical Engineering, vol. 4, no. 7, pp. 747-754, 2014. [154] P. Britos, "Master's Thesis for Probability of Detection in Friction Stir Welding using Non-Destructive Testing," Wichita State University, Wichita, USA, March 2010. [155] M. Tabatabaeipoura, J. Hettlera, S. Delruea and K. V. D. Abeelea, "Nondestructive Ultrasonic Inspection of Friction Stir Welds," in International Congress on Ultrasonics , Metz, 2015.

137

[156] C. Bird, O. Dupuis and A. Lamarre, "New developments of ultrasound phased array for the evaluation of friction stir welds," in Friction Stir welding and Processing II TMS annual meeting, Quebec Canada, 2-6 March 2003. [157] L. Cederqvista and T. Öbergc, "Reliability study of friction stir welded copper canisters containing Sweden's nuclear waste," Elevier Science Inc Journal of the Reliability Engineering & System Safety, vol. 93, no. 10, pp. 1491-1499, October 2008. [158] T. Saravanana, B. Lahirial, V. Pillaib, J. Philip, B. Rao and T. Jayakumar, "Non-Destructive Evaluation of Friction Stir Welded Joints by X-ray Radiography and Infrared Thermography," in 1st International Conference on Structural Integrity, 2014. [159] P. Edwards and M. Ramulub, "Material flow during friction stir welding of Ti- 6Al-4V," Elsevier Science Inc Journal of Materials Processing Technology, vol. 218, p. 107–115, April 2015. [160] C. Patil, H. Patil and H. Patil, "Investigation of Weld Defects in Similar and Dissimilar Friction Stir Welded Joints of Aluminium Alloys of AA7075 and AA6061 by X-ray Radiography," American Journal of Materials Engineering and Technology, vol. 4, no. 1, pp. 11-15, 2016. [161] T. W. I. G. Sites. [Online]. Available: http://www.twi-global.com/technical- knowledge/job-knowledge/non-destructive-examination-nde-part-1-liquid- penetrant-and-magnetic-particle-inspection-122/. [Accessed 2017 May 15]. [162] "The Approval of Friction Stir Welding in Aluminium," American Bureau of Shipping, Houston, USA, October 2011. [163] D. Kinchen and L. Martin, "NDE of Friction Stir Welds in Aerospace Applications," in NDE Inspection Trends, New Orleans, USA, NASA, pp. 1-7. [164] A. International, "Standard: ASTM B308/B308M Specification for Aluminium- Alloy 6061-T6 Standard structural profiles," October 2010. [165] "Hot Mounting," Struers, [Online]. Available: http://www.struers.com/Content/Documents/Brochures/English/Hot%20Mounti ng_1015_English.pdf. [Accessed 09 June 2017].

138

[166] E. Weidmann and A. Guesnier, "Metallographic preparation of aluminium and aluminium alloys," Struers, Copenhagen, Denmark, April 2008. [167] "Standard test methods for determining average grain size, ASTM E112-12," ASTM International, West Conshohocken, PA, 2012. [168] "Standard Test Method for Microindentation Hardness of Materials ASTM 384- 16," ASTM International, West Conshohocken, PA, 2016. [169] "Standard Test Methods for Tension Testing Wrought and Cast," ASTM International, West Conshohocken, PA, 2014. [170] "Standard Test Methods for Tension Testing of Metallic Materials," ASTM International, West Conshohocken, PA, 2013. [171] "Standard Practice for Liquid Penetrant Testing, ASTM E1417 / E1417M - 16," ASTM International, West Conshohocken, PA, 2016. [172] "Standard Reference Radiographs for Inspection of Aluminum and Magnesium Castings, ASTM E155 - 15," ASTM International, West Conshohocken, PA, 2015. [173] "Standard Practice for Contact Ultrasonic Testing of Weldments, ASTM E164 - 13," ASTM International, West Conshohocken, PA, 2013. [174] M. Berke, "Nondestructive Material Testing with Ultrasonics," KrautKramer Non-Destructive Testing, vol. 5, no. 09, September 2000. [175] "Standard Practice for Use of the Ultrasonic Time of Flight, ASTM E2373 - 04," ASTM International, West Conshohocken, PA, 2004. [176] "Standard Guide for Evaluating Performance Characteristics of Phased-Array Ultrasonic Testing Instruments and Systems, ASTM E2491 - 13," ASTM International, , West Conshohocken, PA, 2013. [177] A. Sedmak, R. Kumar, S. Chattopadhyaya, S. Hloch, S. Tadic´, A. Djurdjevic´, I. Cˇ ekovic´ and E. Doncˇeva, "Heat input effect of friction stir welding on aluminum alloy AA 6061-T6 welded joint," Thermal Sciences, vol. 20, p. 637– 641, 2016. [178] Z. Chen, S. Li and S. Hihara, "Microstructure, mechanical properties and corrosion of friction stir welded 6061 Aluminum Alloy," Hawaii Corrosion

139

Laboratory Department of Mechanical Engineering University of Hawaii, Manoa, Honolulu, HI 96822, USA. [179] M. Ilangovan, S. Boopathy and V. Balasubrmanian, "Microstructure and tensile properties of friction stir welded dissimilar AA6061-AA5086 aluminium alloy joints," Trans. Nonferrous Met. Soc. China, vol. 25, pp. 1080-1090, 3 November 2014 . [180] C. M. V. B. S. Rajakumar, "Optimization of the friction-stir-welding process and the tool parameters to attain a maximum tensile strength of AA7075-T6 aluminium alloy," Elsevier Engineering & Manufacturing, vol. 224, pp. 1175- 1191, 2010. [181] H. Lombord, D. Hattingh, A. Steuwer and M. James, "Optimising FSW process parameters to minimise defects and maximise fatigue life in 5083- H321 aluminium alloy," Elsevier Engineering Fracture Mechanics, vol. 75, pp. 341-354, 2008. [182] A. K. Lakshminarayanan, V. Balasubramanian and K. Elangovan, "Effect of welding processes on tensile properties of AA6061 aluminium alloy joints," The International Journal of Advanced Manufacturing Technology, vol. 40, no. 3, p. 286–296, January 2009. [183] T. DebRoy and S. A. David, "Physical processes in fusion welding," American Physics Society, Reviews of Modern physics, vol. 67, no. 1, pp. 67- 85, 1 January 1995. [184] V. Bindal, "Water based couplants for general use for ultrasonic NDT testing," Journal of Scientific and Industrial research, vol. 59, pp. 935-939, November 2000. [185] Y. Kim, S. Song, S. Lee, J. Lee, S. Hong and H. Eom, "A study of Couplant Effects on Contact Ultrsonic Testing," Journal of Korean Society for Non- Destructive Testing, vol. 22, no. 6, pp. 621-626, 2002. [186] M. Anderson, D. Kass, M. Moles and T. Nelligan, Phased Array Testing Basic Pinciples in Industrial Applications, Waltham, USA: Olympus NDT, Nov 2010.

140

7. APPENDICES

141

Appendix A

A1. Aluminium Alloy Elemental Composition

Figure A1: AA6061-T6 Aluminum alloy chemical composition

142

APPENDIX B

B1. Metallographic preparation of aluminium used for AA6061 weld sample

Table B1-1: Grinding and Polishing procedure

Table B1-2: Micro Etching

143

APPENDIX C

C1. Macrographs of the welds and corresponding process parameters

Figure C1.1: BW1 Sample_ 700rpm_60mm/min

Figure C1.2: BW2 Sample_ 700rpm_80mm/min

Figure C1.3: BW4 Sample_ 900rpm_60mm/min

144

Figure C1.4: BW5 Sample_ 900rpm_80mm/min

Figure C1.5: BW7 Sample_ 1100rpm_80mm/min

Figure C1.6: BW83 Sample_ 1100rpm_80mm/min

145

APPENDIX D

D1. Optical Micrographs of the welds retreating side

Figure D1:1: Typical Microstructures of the (a) HAZ; (b) TMAZ; for 700rpm_60mm/min

Figure D1:3: Typical Microstructures of the (a) HAZ; (b) TMAZ; for 700rpm_100mm/min

146

Figure D1:4: Typical Microstructures of the (a) HAZ; (b) TMAZ; for 900rpm_60mm/min

Figure D1:5: Typical Microstructures of the (a) HAZ; (b) TMAZ; for 900rpm_80mm/min

Figure D1:6: Typical Microstructures of the (a) HAZ; (b) TMAZ; for 1100rpm_60mm/min

147

Figure D1:8: Typical Microstructures of the (a) HAZ; (b) TMAZ; for 1100rpm_80mm/min

Figure D1:9: Typical Microstructures of the (a) HAZ; (b) TMAZ; for 1100rpm_100mm/min

148

APPENDIX E

E1. Fractured Mode of Tensile Test Samples for the weld’s stabilized region

Figure E1:1: Fractured Tensile Samples for weld sample process parameters 700rpm_60mm/min

Figure E1:2: Fractured Tensile Samples for weld sample process parameters 700rpm_80mm/min

Figure E1:3: Fractured Tensile Samples for weld sample process parameters 700rpm_100mm/min

Figure E1:4: Fractured Tensile Samples for weld sample process parameters 900rpm_60mm/min

Figure E1:5: Fractured Tensile Samples for weld sample process parameters 900rpm_80mm/min

Figure E1:6: Fractured Tensile Samples for weld sample process parameters 900rpm_100mm/min

149

Figure E1:7: Fractured Tensile Samples for weld sample process parameters 1100rpm_60mm/min

Figure E1:8: Fractured Tensile Samples for weld sample process parameters 1100rpm_80mm/min

Figure E1:9: Fractured Tensile Samples for weld sample process parameters 1100rpm_80mm/min

150

E2. Force_Extension data of pulled Tensile Tests

Figure E2.1: Tensile behaviour of welds

E3. Stress-Strain Relationship of welds

151

Figure E3.1: BW1 Sample_ 700rpm_60mm/min Stress-strain behaviour

Stress-Strain Curve 200

150

100

50 Stress Stress [MPa] 0

-50 4.12E-05 7.03E-05 0.00097717 0.001470689 0.001144558 0.000394849 0.001161848 0.000784636 0.000537091 0.000673829 0.000996818 0.000239251 0.000744558 0.000364202 0.001170492 0.001290729 0.001357525 0.001155562 0.000280901 0.001540632 0.001281299 Strain % -0.000167824 -0.001067633 -0.000713996 -0.001080992

Figure E3.2: BW2 Sample_ 700rpm_80mm/min Stress-strain behaviour

Stress-Strain Curve 200 180 160 140 120 100 80

Stress Stress [MPa] 60 40 20 0 0 4.59905 6.89881 9.53616 1.379684 1.839683 2.299525 2.759446 3.219366 3.679366 4.139287 5.059049 5.519049 5.978891 6.438969 7.358653 7.818573 8.278573 8.738415 9.198415 9.528378 9.648272 0.4598421 0.9198414 Strain %

Figure E3.3: BW3 Sample_ 700rpm_100mm/min Stress-strain behaviour

152

Stress_Strain Curve 200

150

100

50 Stress Stress [MPa}

0 0.00 0.44 0.88 1.32 1.76 2.21 2.65 3.09 3.53 3.97 4.41 4.85 5.29 5.74 6.18 6.62 7.06 7.50 7.94 8.38 8.82 9.26 9.71 10.11 10.11 10.16 -50 Strain %

Figure E3.4: BW4 Sample_ 900rpm_60mm/min Stress-strain behaviour

Stress_Strain Curve 200 180 160 140 120 100 80

Stress Stress [MPa] 60 40 20 0 0.00 0.43 0.87 1.30 1.74 2.17 2.60 3.04 3.47 3.91 4.34 4.78 5.21 5.64 6.08 6.51 6.95 7.38 7.81 8.25 8.68 9.12 9.55 9.99 10.38 10.38 10.38 Strain %

Figure E3.5: BW5 Sample_ 900rpm_80mm/min Stress-strain behaviour

153

Stress_Strain Curve 250

200

150

100 Stress Stress [MPa] 50

0 0.00 0.51 1.01 1.52 2.03 2.53 3.04 3.55 4.05 4.56 5.07 5.57 6.08 6.58 7.09 7.60 8.10 8.61 9.12 9.62 10.13 10.64 11.14 11.65 12.16 12.28 12.29 Strain %

Figure E3.6: BW5 Sample_ 900rpm_100mm/min Stress-strain behaviour

Stress_Strain Curve 200 180 160 140 120 100 80

Stress Stress [MPa] 60 40 20 0 0.00 0.48 0.97 1.45 1.93 2.41 2.90 3.38 3.86 4.34 4.83 5.31 5.79 6.27 6.76 7.24 7.72 8.20 8.69 9.17 9.65 10.13 10.33 10.34 Strain %

Figure E3.7: BW7 Sample_ 1100rpm_60mm/min Stress-strain behaviour

154

Stress_Strain Curve 200 180 160 140 120 100 80

Stress Stress [MPa] 60 40 20 0 0.00 0.48 0.95 1.43 1.91 2.39 2.86 3.34 3.82 4.30 4.77 5.25 5.73 6.20 6.68 7.16 7.64 8.11 8.59 9.07 9.54 10.02 10.50 10.64 10.64 Strain %

Figure E3.8: BW8 Sample_ 1100rpm_80mm/min Stress-strain behaviour

Stress_Strain Curve 250

200

150

100 Stress Stress [MPa)

0 0.00 0.42 0.84 1.26 1.68 2.10 2.52 2.94 3.36 3.78 4.20 4.62 5.04 5.46 5.88 6.30 6.72 7.14 7.56 7.98 8.40 8.82 9.24 9.38 9.40 9.50 Strain %

Figure E3.8: BW9 Sample_ 1100rpm_100mm/min Stress-strain behaviour

155

APPENDIX F

E1. Publications

Figure E3.8: Presentation paper extract for AMPM International Conference 7-8 March 2018 Port Elizabeth, South Africa

156