Masters thesis of engineering investigation of material and mechanical properties of laser beam welded (lbw) laser powder bed fusion (lpbf) and wrought titanium alloy samples

Investigation of Material and Mechanical Properties of Laser Beam Welded LBW Laser Powder Bed Fusion LPBF and Wrought Titanium Alloy Samples A thesis submitted in fulfilment of the requ

Investigation of Material and Mechanical Properties of Laser Beam Welded (LBW) Laser Powder Bed Fusion (LPBF) and

Wrought Titanium Alloy Samples

A thesis submitted in fulfilment of the requirement for the degree of Master of Engineering

Ali Tamaddon

B Sc (Mechanical Engineering), Azad University, Tehran

P E (Institute of Engineers Australia), Canberra, ACT

School of Engineering College of Science, Technology, Engineering and Mathematics

RMIT University November 2021

Declaration

I certify that except where due acknowledgement has been made, this research is that of the author alone; the content of this research submission is the result of work which has been carried out since the official commencement date of the approved research program; any editorial work, paid or unpaid, carried out by a third party is acknowledged; and, ethics procedures and guidelines have been followed

In addition, I certify that this submission contains no material previously submitted for award

of any qualification at any other university or institution, unless approved for a joint-award with another institution, and acknowledge that no part of this work will, in the future, be used

in a submission in my name, for any other qualification in any university or other tertiary institution without the prior approval of the University, and where applicable, any partner institution responsible for the joint-award of this degree

I acknowledge that copyright of any published works contained within this thesis resides with the copyright holder(s) of those works

I give permission for the digital version of my research submission to be made available on the web, via the University’s digital research repository, unless permission has been granted

by the University to restrict access for a period of time

I acknowledge the support I have received for my research through the provision of an Australian Government Research Training Program Scholarship

Ali Tamaddon, 19/11/2021

Acknowledgements:

This is to acknowledge the start of a new chapter in my life and a career in research

I would like to express my sincere gratitude to Professor Sabu John and Distinguished Professor Milan Brandt for their unequivocal support, guidance, and mentorship, which will never be forgotten Their invaluable knowledge and wisdom have been a beacon of reckoning throughout these years I would like to thank key RMIT staff that assisted me throughout this research: Mr Alan Jones of the Advanced Manufacturing Precinct (AMP) and Dr Matthew Field and Dr Edwin Mayes from the RMIT Microscopy and Microanalysis Facility (RMMF) and

Dr Wei Qian Song from the RMIT Bundoora East Material Testing Laboratories

I would also like to thank my dearest son Mr Daniel Tamaddon for his patience and understanding during this period and Dr Mohammad Mehdizadeh who encouraged me to start this journey A list of many good friends who supported me throughout this process would be too long for this limited space and includes but is not limited to Dr Joe Elambasseril and Dr Nabi Chowdhury to name a few

Table of Contents

Declaration ii

Acknowledgements: iii

Table of Contents iv

List of Tables viii

List of Figures ix

Abstract 1

1 Introduction 3

1.1 Background 3

1.2 Motivation 5

1.3 Project Objectives 6

1.3.1 Aim 6

1.3.2 Research Questions 7

1.3.3 Research Scope 7

1.4 Thesis Outline 8

1.5 References 9

2 Literature Review 10

2.1 Chapter overview 10

2.2 Overview of additive manufacturing (AM) 10

2.3 Metal additive manufacturing 13

2.4 Laser powder bed fusion (LPBF) 15

2.4.1 Powder feedstock 17

2.4.2 Laser fusion 18

2.4.3 Build shielding gas 20

2.5 Laser beam welding (LBW) 23

2.5.1 Introduction to lasers 23

2.5.2 Disk lasers 24

2.5.3 Laser welding 24

2.6 Welding considerations 28

2.7 Microscopy 29

2.8 Research Questions 30

2.9 References 31

3 Materials and Methodology 35

3.1 Introduction 35

3.2.2 Material: 36

3.2.3 Fabrication of Laser Powder Bed Fusion (LPBF) parts 38

3.2.4 Laser Beam Welding 41

3.2.5 Cross section and metallographic preparation 44

3.2.6 Microscopy (area, shape) 47

3.2.7 Weld porosity studies 47

3.2.8 EDS (chemical composition) 49

3.2.9 Microscopy (EBSD) 50

3.2.10 Micro hardness 52

3.3 Second Phase: Main Samples 53

3.3.1 Design of experiments 53

3.3.2 Fabrication of parts 55

3.3.3 Pre-testing measurement and testing: 59

3.3.4 Tensile test set up: 59

3.3.5 Fatigue test set up: 62

3.3.6 Post-test examinations: 64

3.4 References 65

4 Results and Discussion 66

4.1 Introduction 66

4.2 Pilot samples 66

4.2.1 Macroscopy 66

4.2.1.1 Weld appearance and comparison to AWI standards 66

4.2.2 Optical Microscopy 68

4.2.2.1 Weld area measurement, shape, and penetration profile 68

4.2.3 Scanning Electron Microscopy 73

4.2.3.1 EDS and chemical composition results 73

4.2.3.2 Microstructure and EBSD results 75

4.2.4 Porosity analysis 79

4.2.5 Hardness profile 85

4.3 Main samples 89

4.3.1 Design of experiments 89

4.3.2 Tensile tests 90

4.3.2.1 Tensile test setup 90

4.3.2.2 Single piece wrought and LPBF tensile tests 90

4.3.2.3 Porosity analysis – Single piece wrought and LPBF 91

4.3.2.4 Welded wrought to LPBF assembly tensile tests 100

4.3.2.4.1 Pre-test porosity analysis – Wrought to LPBF 100

4.3.2.4.2 Tensile tests – Wrought to LPBF 103

4.3.2.4.3 Post-test CT scan – Wrought to LPBF 106

4.3.2.5 Welded LPBF to LPBF assembly tensile tests 109

4.3.2.6 Tensile tests – LPBF to LPBF 110

4.3.2.7 Post-test CT scans – LPBF to LPBF 111

4.3.3 Fatigue tests 116

4.3.3.1 Design of experiments 116

4.3.3.2 Single piece wrought fatigue tests 117

4.3.3.3 Pre-test porosity analysis – single piece wrought 118

4.3.3.4 Fatigue test – single piece wrought 118

4.3.3.5 Post-test SEM – single piece wrought 118

4.3.3.6 Single piece LPBF fatigue tests 121

4.3.3.7 Pre-test porosity analysis – single piece LPBF 121

4.3.3.8 Fatigue test – single piece LPBF 122

4.3.3.9 Post-test CT scan – single piece LPBF 122

4.3.3.10 Post-test SEM – single piece LPBF 123

4.3.3.11 Welded wrought to LPBF assembly fatigue tests 125

4.3.3.12 Pre-test porosity analysis – wrought to LPBF 125

4.3.3.13 Fatigue test – wrought to LPBF 131

4.3.3.14 Post-test CT scan – wrought to LPBF 136

4.3.3.15 Post-test SEM – wrought to LPBF 138

4.3.3.16 Welded LPBF to LPBF assembly fatigue tests 150

4.3.3.17 Pre-test porosity analysis – LPBF to LPBF 150

4.3.3.18 Fatigue test – LPBF to LPBF 151

4.3.3.19 Post-test CT scan – LPBF to LPBF 157

4.3.3.20 Post-test SEM – LPBF to LPBF 158

4.4 References 168

5 Conclusions 169

5.1 Chapter overview 169

5.2 Welding parameters 169

5.3 Tensile strength and fatigue performance 170

5.4 Microstructural parameters 171

5.5 Future work 172

5.5.3 Build parameters effect on weldability 172 5.5.4 Impact of geometry on weld performance 172 5.5.5 Enhance statistical confidence of results 172

List of Tables

Table 2-1 (Page 15) Cost comparison of titanium vs steel and aluminium

Table 3-1 (Page 36) Primary Design of Experiment for pilot sample weld parameters

Table 3-2 (Page 37) Nominal chemical composition of Titanium alloy grade 5 used for this research, as per ASTM F2924-14

Table 3-3 (Page 39) LPBF base material build parameters

Table 3-4 (Page 53) Proposed design of experiment for tensile tests as per ASTM E8

Table 3-5 (Page 54) Proposed design of experiment for fatigue tests as per ASTM E466 Table 4-1 (Page 68) Surface colour in titanium welds; as per American Welding Society (AWS) G2.4M:2014 – guide for the fusion welding of titanium and titanium alloys

Table 4-2 (Page 72) Summary of weld parameters, resulting energy density and welded joint fusion zone cross section area on pilot samples as measured using digital microscopy Table 4-3 (Page 75) Ti6Al4V raw material properties

Table 4-4 (Page 77) Phase identification in fusion zone of pilot sample number 9

Table 4-5 (Page 98) ASTM E8 tensile test results on wrought single piece samples

Table 4-6 (Page 105) ASTM E8 tensile test results for LPBF to wrought welded samples Table 4-7 (Page 110) ASTM E8 tensile test results for LPBF-to-LPBF welded samples, note the close proximity of the UTS to breaking strength and relative short overall elongation prior to breakage

Table 4-8 (Page 117) ASTM E466 fatigue test design of experiment (DOE)

Table 4-9 (Page 118) ASTM E466 fatigue test results for single piece wrought samples Table 4-10 (Page 122) ASTM E466 fatigue test results for single piece LPBF samples

Table 4-11 (Page 131) ASTM E466 fatigue test results for LPBF to wrought welded assemblies Table 4-12 (Page 152) ASTM E466 fatigue test results for LPBF to LPBF welded assemblies

List of Figures

Figure 2-1 (Page 11) Single step and multi-step AM process principles (Standard 2012) Figure 2-2 (Page 12) Production of AM parts from independent service providers (in millions

of dollars) (Wohlers Report 2021)

Figure 2-3 (Page 13) Experience of AM technologies per industry 2019 (%) Courtesy Ernest Young (EY_Global 2019b)

Figure 2-4 (Page 14) Overview of single-step AM processing principles for metallic materials courtesy ISO ASTM 52900 (Standard 2012)

Figure 2-5 (Page 16) Build rate comparison of small (blue), medium (red) and large (green) machines Courtesy (Khorasani et al 2020)

Figure 2-6 (Page 20) Summary of stress (S) versus cycles to failure (N) (S-N) data for PBF (laser), PBF (E-beam), and wire (DED) at R = 0.1 Metallic Materials Properties Development and Standardization (MMPDS) data for cast, wrought machined data are shown for comparison (Lewandowski & Seifi 2016)

Figure 2-7 (Page 22) Schematic representation of possible process by-products courtesy (Ladewig et al 2016)

Figure 2-8 (Page 22) Schematic of defect induced by recoating during LPBF process

Figure 2-9 (Page 23) Four major components of a laser (AWS 2010)

Figure 2-10 (Page 24) Principles for thin disk laser (AWS 2010)

Figure 2-11 (Page 25) Trump TruLaser Cell 7020 weld setup Laser beam comes from above and nozzle feeds the inert shielding gas to the work area

Figure 2-12 (Page 27) Tensile test results related to the porosity ratio in welding area, according to the average power variation (Akman et al 2009)

Figure 2-13 (Page 27) Microhardness distribution of workpieces for different average power (Akman et al 2009)

Figure 3-1 (Page 37) SEM image of the used Titanium grade 5 (Ti-6Al-4V) powder

Figure 3-2 (Page 37) Titanium grade 5 alloy wrought sheets

Figure 3-3 (Page 39) SLM Solutions ™ SLM 250, metal additive manufacturing machine Figure 3-4 (Page 40) 20mm x 20mm x 4mm LPBF fabricated Ti64 plates for preliminary welding experiments

Figure 3-5 (Page 41) Trumpf TruLaser Cell 7020 welding machine and interface

Figure 3-6 (Page 42) Custom jig to hold the square samples in compression for welding and diagram of weld direction used

Figure 3-7 (Page 42) Welding set up inside the Trumpf Trulaser ™ Cell 7020 robotic laser welding machine chamber

Figure 3-8 (Page 43) Trial runs on provisional samples, to verify weld parameters Note the direction of welds marked on far-right sample in image Start (S) to finish (F)

Figure 3-9 (Page 44) Work in progress of the pilot sample fabrication

Figure 3-10 (Page 45) Struers Labotom-3 manual table-top lab size cutting machine

Figure 3-11 (Page 46) Struers CitoPress phenolic resin mounting machine

Figure 3-12 (Page 46) Struers RotoPol-21 and RotoForce-4 polishing machine

Figure 3-13 (Page 47) Keyence VHX-5000 Digital optical microscope

Figure 3-14 (Page 48) Phoenix v| Tome|x s CT scanner at RMIT Bundoora East campus Figure 3-15 (Page 49) Image of VG Studio 3.0 interface

Figure 3-16 (Page 50) FEI Quanta 200

Figure 3-17 (Page 52) FEI Nova NanoSEM 200

Figure 3-18 (Page 52) Microhardness measurements performed on cross section

Figure 3-19 (Page 55) Work in progress of main samples fabrication in the SLM® 250HL

machine

Figure 3-20 (left) (Page 56) Additively manufactured tensile testing half specimens as per ASTM E8 prior to welding, note elongated tabs the junction for welding, these are incorporated to accommodate any anomalies occurring at the start and finish of each weld These tabs are machined off and result in a fully compliant geometry as per ASTM E8 (right) wrought titanium alloy half specimens alongside baseline samples, these are used to establish the experiment reference data necessary for this research

Figure 3-21 (Page 56) Tensile testing sample geometry (courtesy ASTM E8 standard)

Figure 3-22 (Page 57) (left) Additively manufactured fatigue testing half specimens as per ASTM E466 prior to welding Note the provision for welding tabs, similar to tensile testing samples (right) wrought titanium alloy half specimens alongside baseline fatigue full samples Figure 3-23 (Page 57) Fatigue testing sample geometry (courtesy ASTM E466 standard) Figure 3-24 (Page 58) Custom jig to accommodate ASTM E8 and E466 autogenous welding in

Figure 3-25 (Page 59) Machined Fatigue and Tensile samples post machining and ready for tests

Figure 3-26 (Page 61) MTS Landmark 100kN Servohydraulic Test System, material testing machine Material testing laboratory, RMIT University, Bundoora East, Victoria

Figure 3-27 (Page 62) Close up of the tensile sample loaded onto the rectangular grips and the MTS extensometer attached and armed (note pin is removed and arm is aligned, visible via the opening in which the pin has been removed from)

Figure 3-28 (Page 63) Polycarbonate ballistic shield to protect against flying debris

Figure 4-1 (Page 67) Top side of welded pilot samples

Figure 4-2 (Page 67) Bottom side of welded pilot samples

Figure 4-3 (Page 69 – 71) Combined cross section area of fusion and heat affected zones Figure 4-4 (Page 72) Scatterplot of combined fusion zone (FZ) and heat affected zone (HAZ) area versus welding energy density

Figure 4-5 (Page 74) (Top) Cross section of welded joint as visible via SEM microscope (Bottom) Energy dispersive spectroscopy (EDS) results across welded joint, sample number 8 Figure 4-6 (Page 75) Energy dispersive spectroscopy (EDS) results of base metal (BM) at the additive manufactured side of sample number 8

Figure 4-7 (Page 76) Electron backscatter diffraction images of fusion zone on sample 9 Top left shows the overall area under study, with the zoomed area highlighted in the top right image Needle shaped titanium alloy grains captured middle left using the parameters above middle right Phase identification bottom left and highlighted of Titanium hexagonal close packed (HCP) crystals bottom right

Figure 4-8 (Page 77) Zoomed out phase identification of pilot sample number 9 (Right) Blue area is Ti hexagonal close packed (HCP) and yellow area is Titanium body centre cubic (BCC) Figure 4-9 (Page 80) Porosity profile of pilot sample number 6, welded at 2000 Watts and 1200mm/min Note that 1.28% of overall region of interest (ROI) consists of pores 241 defects were detected at a mean diameter of 0.14mm, with a mean sphericity of 0.5171 Figure 4-10 (Page 81) Porosity profile of pilot sample number 9, welded at 2200 Watts and 1200mm/min Note that only 0.12% of overall region of interest (ROI) consists of pores 99 defects were detected at a mean diameter of 0.12mm, with a mean sphericity of 0.4901 Figure 4-11 (Page 82) There is direct relationship between porosity and mean pore size Figure 4-12 (Page 83) There is direct relationship between laser welding power levels and the amount of porosity in the final resulting welded joint This is shown with a probability value (P-value) of 0.686

Figure 4-13 (Page 83) There is an inverse relationship between the welding laser beam travel speed and the resulting porosity, with a P-value of 0.796

Figure 4-14 (Page 84) (left) Direction relationship between welding energy density and resulting porosity is not evident until (right) outlier datapoint (100 J/mm) are taken out, resulting in a clear demonstration of a relationship between the welding energy density (J/mm) and the resulting porosity (% volume) of the welded joints

Figure 4-15 (Page 84) (left) inverse relationship between laser beam welding power levels with a P-value of 0.448 and Figure 4-16 (right) show the inverse relationship between porosity sphericity to laser beam welding travel speeds with a P-value of 0.690

Figure 4-16 (Page 85) Leco LM700AT micro-indentation hardness testing machine

Figure 4-17 (Page 86) Microhardness profile across three different assemblies built using 1800W power level The LPBF side is labelled as SLM

Figure 4-18 (Page 86) Microhardness profile across three different assemblies built using 2000W power level The LPBF side is labelled as SLM

Figure 4-19 (Page 87) Microhardness profile across three different assemblies built using 2200W power level The LPBF side is labelled as SLM

Figure 4-20 (Page 87) The greatest variation in hardness happens on the HAZ area of LPBF The LPBF side is labelled as SLM

Figure 4-21 (Page 91) CT scan image of porosities Note only pores with a probability greater

or equal to 1 have been annotated

Figure 4-22 (Page 92) CT scan data graphical representation of porosities along the length of the tensile test sample, note that the origin of the z position is around the middle of the sample Circle diameters proportional to diameter of porosities

Figure 4-23 (Page 93) Tensile test on single piece LPBF Ti6Al4V, note the fracture geometry Figure 4-24 (Page 94) Stress-strain diagram of single piece LPBF tensile sample, as per ASTM E8

Figure 4-25 (Page 95) CT scan of single piece LPBF tensile test sample, not showing any significant porosities around the fractured surface, besides the single pore at 180 microns diameter (defect 7)

Figure 4-26 (Page 96) Fractured surface of LPBF single piece tensile sample Note the flat surface of failure perpendicular to the axis of stress loading at the middle of the part Top image at 38x magnification Bottom left at 50x facing the upper and bottom right at 50x, facing the lower area of the face Note dark points as porosities

Figure 4-27 (Page 97) Three identical ASTM E8 wrought Ti6Al4V alloy samples were tested for tensile strength The top 2 resulted in a cup and cone fracture surface and the last sample resulted in a single slanted 45 degrees fracture

Figure 4-28 (Page 98) Stress-Strain diagram of wrought single piece tests as per ASTM E8 Figure 4-29 (Page 99) Fractured surface of wrought single piece tensile sample Note the 45 degrees angle between the surface of failure and axis of stress loading Top image at 50x magnification, bottom left at 200x and bottom right at 400x, highlighting one of the points of failure

Figure 4-30 (Page 100) Sample number 1, CT scan and porosity distribution histogram Figure 4-31 (Page 101) Sample number 2, CT scan and porosity distribution histogram Figure 4-32 (Page 102) Sample number 3, CT scan and porosity distribution diagram

Figure 4-33 (Page 103) Minitab combined analysis of porosity data for test samples 1, 2 and

Figure 4-39 (Page 107) Surface of fractured damage Left image shows entire face of damaged area in 50x magnification, right image shows a 100x magnification of cup and cone fracture detail

Figure 4-40 (Page 108) Top left (200x) and top right (1600x) show details of the cup and cone failure (in ductile mode) at higher magnification

Figure 4-41 (Page 109) Three LPBF to LPBF welded test samples

Figure 4-42 (Page 110) Stress-Strain diagram of LPBF-to-LPBF welded samples as per ASTM E8

Figure 4-43 (Page 111) CT scan images of sample numbers 1 and 2 of the welded LPBF-LPBF tensile tests Both images have annotations showing the location of the top 5 largest porosities by diameter

Figure 4-44 (Page 112) Histograms showing the distribution of porosity diameter sizes in the samples number 1 (top) and 2 (bottom) Both show relatively similar pore sizes, with the mean sizes ranging between 260 and 227 microns

Figure 4-45 (Page 113) Distribution of porosities relative to the z-axis, with pore sizes represented by diagram size As evident, the majority of the pores are at the vicinity of the fusion zone (circa 400 to 600 voxels on the z-axis) Top is sample 1 and bottom is sample 2 Z datum point is at base metal and 600 microns is around FZ

Figure 4-46 (Page 114) Large pores can be seen at the perimeter and on the surface of the fractured area of tensile test sample number 1 of the LPBF-to-LPBF welded assembly

Figure 4-47 (Page 115) Large pores can be seen at the perimeter and on the surface of the fractured area of tensile test sample number 2 of the LPBF-to-LPBF welded assembly

Figure 4-48 (Page 116) Fatigue test loading profile

Figure 4-49 (Page 119) Single piece wrought samples subjected to 500MPa and R0.1 axial fatigue loading Top image is the top view, and the bottom image is the side view As evident from these images the failure mode is ductile, with a fracture plane at 45 degrees to the plane

of loading

Figure 4-50 (Page 120) Fatigue test results for single piece wrought samples as per ASTM E466 (Sample ID W1), which failed at 58,046 cycles, shows a complete ductile failure initiated

at a material defect as evident in the highlighted images above

Figure 4-51 (Page 121) Histogram of sample ID LL1 porosity analysis

Figure 4-52 (Page 122) The largest 5 pores in sample ID LL1 were distributed along the entire length of the sample

Figure 4-53 (Page 123) Distribution of pores in sample ID LL1

Figure 4-54 (Page 124) LPBF single piece fatigue sample fractured surface Note the multiple pores on the surface of the failure plane and edges of the fracture (bottom right)

Figure 4-55 (Page 126) CT scan images of samples (from top to bottom and left to right) LW10, LW11 and LW12 (alongside an image to show the orientation of the axis, note Z-axis is zero at the bottom and increases upwards)

Figure 4-56 (Page 127) Pore diameter to sphercity distribution graph (from top to bottom) LW10 and LW11

Figure 4-56 (continued) (Page 128) Pore diameter to sphercity distribution graph LW12

Figure 4-57 (Page 128) Pore diameter, volume and Z-axis distribution of samples (from top

Figure 4-60 (Page 131) Welded LPBF to wrought fatigue behaviour is best represented by using an S-N curve, with the number of cycles to failure in logarithmic scale in the abscissa and the stress levels on the ordinate

Figure 4-61 (Page 132) Failed welded LPBF to wrought samples subject to 200MPa (sample

ID from top to bottom LW3, LW2 and LW1)

Figure 4-62 (Page 133) Failed welded LPBF to wrought samples subject to 200MPa (top 2, sample ID from top to bottom LW11 and LW12) and 600MPa (bottom sample LW7)

Figure 4-63 (Page 134) Failed welded LPBF to wrought samples subject to 400MPa (sample

ID from top to bottom LW4, LW5 and LW6)

Figure 4-64 (Page 135) Failed welded LPBF to wrought samples subject to 500MPa (sample

ID from top to bottom LW8, LW9 and LW10)

Figure 4-65 (Page 136) LW1 (left) and LW2 (right), both failed at the LPBF side of the welded assembly, but the largest pores were identified to be at the FZ as per CT scans

Figure 4-66 (Page 136) LW3 (left) and LW4 (right), both failed at the LPBF side of the welded assembly, but the largest pores were identified to be at the FZ as per CT scans

Figure 4-67 (Page 137) LW5 (left) and LW6 (right), both failed at the fusion zone (FZ) of the welded assembly and the largest pores were also identified to be at the FZ as per CT scans Figure 4-68 (Page 137) LW7 (left) and LW8 (right), both failed at the fusion zone (FZ) of the welded assembly and the largest pores were also identified to be at the FZ as per CT scans Figure 4-69 (Page 137) LW9 (left) failed at the LPBF side of the welded assembly, but the majority of the largest pores were at the FZ LW10 (right), both failed at the LPBF side of the welded assembly, but the largest pores were identified to be at the FZ as per CT scans Figure 4-70 (Page 138) LW11 (left) and LW12 (right) both failed at the LPBF side of the welded assembly and as evident in the images, in both assemblies the porosities are concentrated at the FZ and HAZ

Figure 4-71 (Page 139) Fracture surface on LPBF base-metal side of the welded assembly LW1, which failed under 200MPa loading, at 144,795 cycles and R=0.1 shows a brittle failure mode, with multiple points of crack initiations at the perimeter of the failure plane

Figure 4-72 (Page 140) Fracture surface at the fusion zone of the welded assembly LW2, which failed under 200MPa axial stress, after 55,513 loading cycles and R=0.1 shows was porosities which acted as multiple points of crack initiations

Figure 4-73 (Page 141) Fracture surface at the fusion zone of the welded assembly LW3, which failed at 200MPa axial stress, after 149,920 loading cycles and R=0.1 Note the porosities and enlarged image of one of the porosities showing crack initiation points at the surface of the pore

Figure 4-74 (Page 142) Fracture surface on LPBF base-metal side of the welded assembly LW4, which failed under 400MPa loading, at 20,347 cycles and R=0.1 Note the pore at the centre of the fracture surface and multiple fracture initiation points at the perimeter of the failure zone

Figure 4-75 (Page 143) Fracture surface on LPBF base-metal side of the welded assembly LW5, under 400MPa loading, at 16,949 cycles and R=0.1 Two major porosities can be clearly seen on these images

Figure 4-76 (Page 144) Fracture surface on LPBF base-metal side of the welded assembly LW6, under 400MPa loading, at 23,352 cycles and R=0.1

Figure 4-77 (Page 145) Fracture surface at fusion zone of welded assembly LW7, under the highest load applied during these tests of 600MPa The part failed after only 60 cycles Large porosities can be observed on the failure surface

Figure 4-78 (Page 146) The contribution of surface roughness acting as crack initiation source can be seen by observing the perimeter of the fracture surface of LW9 which failed at 500MPa after 7,164 loading cycles

Figure 4-79 (Page 147) Large pores and their inherent surface cracks that have led to the failure of sample LW10 at 500MPa of stress after 9,496 loading cycles

Figure 4-80 (Page 148) Surface roughness and inconsistencies have played a significant role

in crack initiation on sample LW11 at 300MPa axial stress after 39,245 loading cycles

Figure 4-81 (Page 149) Surface conditions can also be observed to have had a significant effect on the failure mode of LW12 at 300MPa axial stress after 28,266 loading cycles Crack initiation porosities highlighted in (bottom right) image

Figure 4-82 (Page 150) Sample ID LL7, note the concentration of porosities at z-axis position around FZ and the HAZ

Figure 4-83 (Page 151) Sample ID LL8, shows similar pattern for detectable porosities, with the concentration being at the FZ and HAZ

Figure 4-84 (Page 152) Welded LPBF to LPBF fatigue behaviour is best represented by using

an S-N curve, with the number of cycles to failure in logarithmic scale in the abscissa and the

Figure 4-85 (Page 153) Top and side view of LPBF to LPBF welded assemblies subjected to 200MPa axial loading cycles to failure From top to bottom, sample IDs: LL1, LL2 and LL9 Figure 4-86 (Page 154) Top and side view of LPBF to LPBF welded assemblies subjected to 300MPa axial loading cycles to failure Top LL8 and bottom LL7

Figure 4-87 (Page 155) Top and side view of LPBF to LPBF welded assemblies subjected to 400MPa axial loading cycles to failure Top LL4 and bottom LL3

Figure 4-88 (Page 156) Top and side view of LPBF to LPBF welded samples, subject to 500MPa (top, LL5) and 100MPa (bottom, LL6) The sample subject to 100MPa didn’t fail after 1,000,001 cycles, which for the purpose of this study is considered the limit for infinite life Figure 4-89 (Page 157) (left to right and top to bottom): LL1, LL2, LL3, LL4, LL5, LL7 Note that sample ID LL6 did not fail

Figure 4-90 (Page 158) (left) sample ID LL8 and (right) sample ID LL9

Figure 4-91 (Page 159) Sample LL1, subjected to 200MPa of axial load, survived 117,055 cycles The perpendicular cut surface suggests the brittle nature of the fracture and the pores

at the perimeter, are sources of crack initiation

Figure 4-92 (Page 160) Sample LL2, subjected to 200MPa of axial loading, underwent 161,869 cycles prior to fracture5 as typical in LPBF to LPBF welded assemblies in this research, the failure mode is demonstrated to be brittle, with multiple points of fracture initiation at the perimeter of the welded assembly

Figure 4-93 (Page 161) Sample LL3, subjected to 400MPa of axial stress loading, underwent 19,149 cycles prior to failure Bottom right picture shows the starting point of a microcrack that contributed to the brittle fracture of the part

Figure 4-94 (Page 162) Large pores pertaining to incompletely melted particles can be seen

as dark patches in the images of sample LL4, which was subjected to 400MPa axial loading and maintained its integrity up to 16,544 cycles

Figure 4-95 (Page 163) Sample LL5, which was subjected to the highest axial loading of this research at 500MPa, withstood 10,746 cycles before showing multiple failures, including a secondary fracture on the side of the body, as seen in the bottom two images

Figure 4-96 (Page 164) LL6, close up image of the fusion zone (FZ) and heat affected zones (HAZ) Note that albeit surface anomalies, the welded assembly was able to maintain its integrity after 1,000,001 cycles of 100MPa axial loading

Figure 4-97 (Page 165) Sample LL7, which failed at the fusion zone at 300MPa axial loading after 4,094 cycles, shows a large cluster of porosities at the centre and perimeter of the fusion zone

Figure 4-98 (Page 166) Sample LL8 which also failed at the fusion zone at 300MPa axial loading after only 2,264 cycles, has multiple large pores on the surface and perimeter of the fractured area, contributing to the failure of the part

Figure 4-99 (Page 167) Sample LL9, which failed at the fusion zone after 3,965 cycles at 200MPa axial force, also demonstrates large pores scattered all over the face of the fractured surface and perimeter

This research investigates a means of negating such shortcomings and reaping the benefits of both LPBF additive manufacturing and traditional manufacturing methods by joining such parts together using laser welding in the autogenous mode (no fillers required)

Based on the literature review conducted in this field and the gaps identified, a two-stage methodology was utilised, as per following

At the first stage of the research, 20x20x4mm pilot samples were made using wrought Ti6Al4V titanium alloys and LPBF fabricated Ti6Al4V titanium alloys with the same geometry and dimensions, then welded together using disc laser beam at various power levels and speed, which were selected using existing literature The welded joints were examined for porosity, geometry of weld zones, microhardness and microstructure Energy density and rate of cooling resulting from the various welding parameters were the driving factors of the welded joint results The results of examinations were analysed, and laser beam power level of 2200 Watts and 1200 mm/ min were the most promising parameters

At the second stage of the research more samples were fabricated, in a similar fashion to the first stage, but had geometries and dimensions as per ASTM E8 and E466 standards All samples in the second stage were welded using the most promising weld parameters found

during the first stage and subjected to tensile and fatigue tests Each sample was checked for porosity before and after each test and the results produced a profile of tensile and fatigue characteristics of such assemblies

It is demonstrated that the welding of LPBF to wrought samples, and LPBF to LPBF is a viable solution to overcoming the limitations in build volume and production speed in LPBF technologies The governing factors in the success of such welds are laser energy density (cooling rates) and LPBF base metal initial build quality (porosities)

1 Introduction

1.1 Background

Additive manufacturing has come a long way from the early days when it was regarded as a rapid prototyping technology, used to create polymer components for visualisation and demonstration purposes, to this day where it has become a viable production method for end-use parts of various materials, including complex cellulous fibres, carbon fibre embedded plastics and even metallic parts There is a great amount of expectation of this technology, due to the unique characteristics it provides, such as ability to create complex geometries, freedom from the necessity of jigs and fixture and the time required to transform 3D CAD data into actual parts All which are done seamlessly with minimal need for post processing and minimal wastage of raw material This makes the technology extremely appealing in low-volume, high-complexity scenarios Many OEM (original equipment manufacturer) organisations have started showing interest in such technologies and the market is growing

at a rapid pace and is not showing any signs of deceleration (Brandt 2016)

Additive manufacturing has especially become extremely appealing in fabricating dimensional metallic end-use parts, providing the possibility to fabricate novel shapes and geometrical features never deemed possible prior to the birth of this technology With the proliferation of this technology the affordability factor is also becoming more favourable; to such an extent that owning and operating a metal printing machine is no longer a capability confined to only large OEMs but also available to small and medium enterprises (SMEs) This change of dynamics has led to 3D printing becoming a changing force used to optimise the supply chain and manufacturing today, filling in the gaps of small quantity, high complexity and high value parts With the increase of the quality of the end product and development of 3D-printing software applications for design and analysis, we are witnessing a large shift in the application of additive manufacturing from prototyping to end-use products, with an overall increase of 65% in applying such technologies in industries in 2019 compared to 2016; and there are strong signs and economic model predictions that are showing these figures to accelerate significantly by the end of 2022, with US companies alone shifting from a miniscule share of 4% end use parts fabricated using additive manufacturing to a major share of 54% by end of 2022 (EY_Global 2019a)

three-The entire product development lifecycle requires to be re-engineered to harness the full potential of this new technology This includes but is not limited to all aspects of the supply chain, from design, raw material to fabrication, testing, quality assurance and control and recycling (Huang et al 2015)

Additively manufactured metal components have many benefits, such as complex geometry not possible achieving using traditional subtractive methods, freedom from manufacturing process restrictions, low lead times, minimum wastage of material, near-zero set up costs to name a few New concepts and parameters become important in this new technology, such

as the capital investment in machinery, customised computer aided design (CAD) tools; machine feedstock; skillsets of operators, verification methods for final product and quality assurance and control to safeguard client confidence in the product The marketplace is now much more aware of the potentials as well as the restrictions of this new technology and the race has started to overcome those hurdles and gain a competitive advantage in this new-found marketplace

Metal additive manufacturing has evolved over the years and has seen the rise and fall of various technologies and fabrication methods, such as stereolithography and sheet lamination The most common and highly popular method is the laser powder bed fused (LPBF), formerly known as selective laser melted (SLM) technology, due to its advantages such

as moderate surface roughness, dimensional accuracy, high layer resolution and critical design fabrication to name a few LPBF is used to fabricate a large range of metal alloys such Titanium (Ti), Nickle (Ni), Aluminium (Al) and Ferrous (Fe) based alloys (Revilla et al 2020) Titanium alloys have a special place in the Aerospace industry, due to their high strength to weight ratio; high mechanical strength in the range from cryogenic temperatures to moderately elevated temperatures as well as high resistance to corrosion in the most harshest environments (Putyrskii, Yakovlev & Nochovnaya 2018) Hence there has been a motivation to utilise titanium alloys in fabricating additively manufactured end-use parts, using various technologies LPBF has proliferated in the past decade to be the driving force in industrial scale metal additive manufacturing (Curtis 2013)

1.2 Motivation

Engineering is the practical application of science that also has to deal with any limitations

As for additive manufacturing of metals is concerned, the main draw backs that have hindered the adoption of this technology at full scale, print size and speed of fabrication, make to the top of the list Current build chamber sizes are still restrictive, in comparison to subtractive manufacturing techniques Even with the rapid progress during the duration of this research were their build volumes have increased dramatically from a mere 248x248x250mm (on SLM Solutions SLM250) and their build methods have evolved from a single laser fabrication at a rate of 20 cm3/ hour to a whopping build chamber volume of 600x600x600mm; using 12 concurrent lasers of 1000 Watts each delivering built parts at a rate of 1000cm3/ hour, i.e 1 litre/ hour (on SLM Solutions NXG XII 600) Even such progress is not comparable to the rapid production rates of mass subtractive and traditional manufacturing, thus, to overcome such shortfalls, it is proposed in this research that additively manufactured parts be combined with conventionally manufactured components; to reap the benefits of both methods and overcome the limitations in size and speed of fabrication Traditionally fabricated parts will

be especially economical in areas where additively manufactured components do not demonstrate competitive advantage, such as parts that lack geometrical complexity, are bulky and easily available in off the shelf formats, shapes, and cross sections such as in sheets, standard beams and rods of various sizes An example of such a synergic combination would

be an assembly in which long slender beams would be joined to complex geometrical topologies For instance, in the fabrication of an aircraft wing, the long slender beams would

be the wing spars and the complex geometries would be the wing root assembly and/ or topologically optimised wing cross sections The slender beams would benefit from traditionally manufactured process such as extrusion and rolling and the complex geometry would benefit from additive manufacturing Joining two differently created titanium alloys of same material and understanding the overall performance of the resulting assembly is the main topic addressed in this research Titanium alloys in wrought format have well documented mechanical, weldability and fatigue characteristics; but the same material when subject to production processes involving making the LPBF powder stock and eventual fabrication of parts were layers upon layers of laser are melted and subjected to rapid and extreme thermal cycles are yet to be fully investigated and documented Laser beam welding

of material that have undergone such extreme thermal cycles and combining them with

traditionally fabricated wrought parts will result in unique mechanical and fatigue performance characteristics, that depend on the macroscopic characteristics, microstructure

of the LPBF and wrought parts as well as the parameters used in welding these parts together (Qian et al 2016)

1.3 Project Objectives

This research investigates the material, mechanical properties, and fatigue performance of titanium grade 5 (Ti-6Al-4V) laser powder bed fused (LPBF) additively manufactured parts welded to wrought material of same chemical composition This is achieved using Laser Beam Welding (LBW), in autogenous mode, which means the two metals are fused using laser energy and without any kind of filler material involved The surface condition of each of the welded faces requires to meet certain criteria Since laser beam welding can be conducted using various parameters, such as speeds, power settings, and beam geometry, the focus of this research is to find the co-relationship between these weld parameters with the overall mechanical and fatigue performance of the welded assembly There is no prior literature on the study the mechanical and fatigue performance of Ti-6Al-4V LPBF and wrought alloy assemblies of same material using laser beam welds in autogenous mode A detailed study of macroscopic, microstructure, porosity, hardness, chemical composition, tensile strength and fatigue performance are analysed in detail and related to the welding parameters that are within the control of the technician, during the assembly process

1.3.1 Aim

The aim of this research is to understand how laser beam welding parameters such as laser power level and scanning speed can affect the overall performance of titanium alloy assemblies made using laser powder bed fused (LPBF) and wrought parts welded together using a disc laser beam weld method in autogenous mode Failure modes are related to the variations in the structure due to presence of porosities, non-uniformities, and other material features within the individual parts

1.3.2 Research Questions

The research questions addressed during this research are as follows:

1- What are the most predominant factors of laser beam welding on the overall

mechanical and fatigue performance of a welded SLM assembly?

2- What are the interaction effects of such parameters on the end result?

3- What microscopic parameters (phase, grain size and orientation) effect the overall performance of the welded SLM assembly?

1.3.3 Research Scope

For practical purposes, the scope of the research is confined to the investigation of assemblies fabricated using laser beam welding using disc lasers (1030nm) in autogenous mode with variations in the following parameters only:

1) Three discreet laser power levels of 1800 Watts, 2000 Watts and 2200 Watts and 2) Three discreet scanning speeds of 800 mm/ min, 1000 mm/ min and 1200 mm/ min The parameters used for the fabrication of the individual LPBF parts are deemed to be constant (at laser power of 100 Watts, focal offset at 0mm, spot size at 80 microns, hatch space at 120 microns, layer thickness at 30 microns and beam scanning speed at 375 mm/ min) and irrelevant to the scope of this research Other laser beam welding parameters such

as focal offset distance (at 0mm) and the flow rate of the inert shielding gas (Argon at 20 litres per minute) are also kept constant throughout the research

The assemblies are scrutinised for the following:

1) Macroscopic appearance of the welded joint

2) Microscopic geometry of the fused area and heat affected zone of joint

3) Microstructure of base metal, heat affected zone and fusion zone

4) Chemical composition and any potential changes throughout the cross section of the welded assembly

5) Hardness profile across the base metal, heat affected zone and fusion zone

6) Porosity at the fusion zone and heat affected zone

7) Tensile strength

8) Fatigue performance

All such parameters were then correlated to the laser beam welding parameters and

conclusions are made

1.4 Thesis Outline

The research is documented in five chapters The outline of which are as follows:

Chapter one – Introduction: Describes the problem statement of the research The industrial background, motivation and research tasks are presented

Chapter two – Literature review: Reviews the state of the art and literature available in the field of laser beam welding (LBW), laser powder bed fusion technology (LPBF), welding of titanium, electron back scattering (EBSD), energy dispersive spectroscopy (EDS), tensile and fatigue testing

Chapter three – Materials and methodology: Details the materials and goes into detail of the methodology used to address the research questions, using a scientific and proven approach Chapter four – Results and discussion: Experiment results are studied and discussed in detail against established theory and prior works and literature in the field

Chapter five - Conclusion: Study results are used to draw a conclusion and verdict in response

to research questions The stage is set for any further investigation into this research topic in the future

1.5 References

Brandt, M 2016, Laser Additive Manufacturing: Materials, Design, Technologies, and Applications, Woodhead Publishing series in electronic and optical materials, Elsevier Science & Technology, Cambridge

Curtis, DI 2013, Microstructural and mechanical property characterization of Ti-6Al-4V produced by SLM and EBM

EY_Global 2019a, How 3D printing brings transformational change,

<https://www.ey.com/en_gl/consulting/how-3d-printing-brings-transformational-change>

Huang, Y, Leu, MC, Mazumder, J & Donmez, A 2015, 'Additive manufacturing: current state, future potential, gaps and needs, and recommendations', Journal of Manufacturing Science and

Engineering, vol 137, no 1

Putyrskii, SV, Yakovlev, AL & Nochovnaya, NA 2018, 'Benefits and Applications of High-Strength Titanium Alloys', Russian engineering research, vol 38, no 12, pp 945-948

Qian, M, Xu, W, Brandt, M & Tang, HP 2016, 'Additive manufacturing and postprocessing of Ti-6Al-4V for superior mechanical properties', MRS bulletin, vol 41, no 10, pp 775-784

Revilla, RI, Van Calster, M, Raes, M, Arroud, G, Andreatta, F, Pyl, L, Guillaume, P & De Graeve, I 2020, 'Microstructure and corrosion behavior of 316L stainless steel prepared using different additive manufacturing methods: A comparative study bringing insights into the impact of microstructure on their passivity', Corrosion science, vol 176, p 108914

2.2 Overview of additive manufacturing (AM)

As per ASTM F2792-12a, which made an attempt to provide standard terminology for additive manufacturing technologies, before being withdrawn in 2015, in favour of ISO ASTM 52900: Additive Manufacturing (AM) is the process of joining material to make objects from 3D model data, usually layer upon layer, as opposed to subtractive manufacturing methodologies (ASTM_International 2013) Subtractive manufacturing methodologies include cutting, drilling, machining, milling and other methods, in which the net shape is formed by removing material from a bulk solid In Additive Manufacturing the focus is creating a net shape by adding material which come in various raw formats such as powder, gel, liquid or filaments

to name a few The difference between rapid prototyping and additive manufacturing is the difference in the objectives which are served Where rapid prototyping products are used for

a design often iterative for form, fit or function testing or combination thereof (ASTM_International 2013) Whereas additive manufacturing (AM) focuses on making end use parts and not prototypes

Rapid prototyping was a consequence of the speedy adoption and advancement of computer aided design (CAD) in the 1980’s which brought about the need for prototyping designs in real physical models and parts Rapid prototyping is one of the earlier forms of additive manufacturing CAD data is converted to a stereolithography (STL) file, in which three-dimensional data created in CAD is approximated by triangles and sliced containing the information of each layer that is going to be printed The thickness of each layer as well as the resolution of the build depends on many factors, including raw material, machinery used and heat and material flow limitations

Categorisation of additive manufacturing technologies in the third decade of the twenty first century has started to look like Figure 2-1 (Standard 2012) With the main single step additive manufacturing processes being categorised as metallic, polymer, ceramic and composite This research has focused on metallic single-step additive manufactured titanium Ti6Al4V alloys joined via autogenous disc laser beam welding to other AM Ti64 parts or wrought parts

Figure 2-1 Single step and multi-step AM process principles (Standard 2012)

Additive manufacturing is now expanded to more industries than ever before and now serves not just engineers in aerospace and manufacturing, but also medical practitioners in the health sector, architects in the building industry and scientists and students in academia and research Rapid prototyping is also a very useful tool for market researchers to gauge public and target market opinion on a new product

Additive manufacturing has grown significantly since the early days of its inception in the form

of Selective Laser Sintering in the 1980s, until it became a rapid prototyping tool in the 1990s and then as a rapid direct manufacturing tool in the 2000s (Rayna & Striukova 2016)

Additive manufacturing has always rapid growth and even has been relatively immune to massive market forces such as the global pandemic in the past year, with a market size increase of 7.5% to USD 12.8 billion in 2020 Yet, this was significantly much less compared to the average rate of 27.4% over the last 10 years, as seen in Figure 2-2

Figure 2-2 Production of AM parts from independent service providers (in millions of US dollars) (Wohler’s Report 2021)

A thorough study of 404 firms providing 3D printing services to the European market has revealed the major impact of additive manufacturing on supply chains across various industries Impacts on reducing setup costs, decreasing costs per unit for small production runs and transferring the risk of production to 3D printing service providers Inventory and

warehousing Supply chains also enjoy greater flexibility with the ability to absorb variations

in production volume (Rogers, Baricz & Pawar 2016)

Experience in using additive manufacturing technologies is increasing substantially with 78%

of the aerospace industry having used this technology in their operations and other industries following suit, as seen in Figure 2-3

Figure 2-3 Experience of AM technologies per industry 2019 (%) Courtesy Ernest Young (EY_Global 2019b)

2.3 Metal additive manufacturing

Metals are one of the primary class of materials used for additive manufacturing In 2010, the results of a US Navy workshop were published with the title “direct digital manufacturing (DDM) of metallic components: affordable, durable and structurally efficient aircraft” which laid out a vision of on demand part availability for aircraft The technology at the time was still facing major challenges in design, qualification and certification, maintenance and repair and the science and technology of DDM (Frazier 2010) ISO ASTM 52900 categorises metallic additive manufacturing process as per Figure 2-4

Figure 2-4 Overview of single-step AM processing principles for metallic materials courtesy ISO ASTM 52900 (Standard 2012)

One of the main reasons that titanium alloys are not as widely used in aerospace engineering

is the high cost of such material Additive manufacturing provides many benefits, one such benefit is the minimal material wastage and hence improving the cost effectiveness of such alloys for high demanding applications in highly corrosive environments Table 2-1 provides a cost comparison of titanium to other common engineering material such as steel and aluminium The high cost of titanium is due to the high extraction and processing costs associated with this mineral (Dutta & Froes 2017)

Hence titanium and namely Ti-6Al-4V which due to its high strength to weight ratio and anticorrosive characteristics is one of the most common titanium alloys becomes an attractive option as a metal additive manufacturing feedstock

Table 2-1 Cost comparison of titanium vs steel and aluminium (Dutta & Froes 2017)

Cost of titanium - Comparison

2.4 Laser powder bed fusion (LPBF)

Laser powder bed fusion (LPBF) technology employs, as the naming suggests, a bed of powder that is selectively melted layer upon layer by high power laser, then is lowered and covered

by another layer of powder which in turn is melted again, as per the slices of the part design pattern The melted layers fuse together at their interface due to the energy of the scanning laser A variety of material can be utilised in this technology Other forms of additive manufacturing such sheet lamination do not enjoy the same level of freedom and material savings In sheet lamination thin layers are cut and fused together forming the final shape Excessive material is produced as the result which would need to be reprocessed again and turned into sheets Directed energy deposition methods, use a filament wire or metal powder that is fed through a nozzle and is melted layer upon layer using high energy laser or electron beams The resolution of such processes is yet to be competitive with the precision of LPBF technology Laser powder bed fusion (LPBF) technologies have enjoyed special attention from industry and have gained popularity This research focuses on this technology and attempts

to address these objectives (Nichols 2021)

LPBF has the lowest build rate compared to other metal additive manufacturing processes such as directed energy deposition (DED), electron beam powder bed fusion, binder jetting and sheet lamination Yet, it has superior surface quality and the highest dimensional accuracy The most significant parameters in LPBF technology are laser power, scan speed, layer thickness, hatch space, build size, and powder handling system (Khorasani et al 2020)

Figure 2-5 Build rate comparison of small (blue), medium (red) and large (green) machines Courtesy (Khorasani et al 2020)

Part quality and build rates are the two main areas of recent advancements in LPBF technology by the manufacturers of such machinery This is characterised by an increase in the number of concurrent lasers, maximum wattage and build chamber size (Figure 2-5) Ti6Al4V parts fabricated using LPBF and electron beam melting (EBM) technologies have been demonstrated to have relatively similar fatigue properties, albeit differences in microstructure, both methods being sensitive to the main three types of defects, surface defects, unmelted zones and small internal defects Machining improves the mechanical properties of the parts by removing the surface defects (Chastand et al 2018)

But for the purpose of this research, additional processes post-fabrication is not considered, since that would contradict the benefits of direct digital additive manufacturing

Improvements in each of the process parameters involved in the powder bed fusion process results in an overall improvement in part quality, decrease lead time and increase the safety and user experience of the machines

2.4.1 Powder feedstock

Additive layer manufacturing (ALM) processes such as laser powder bed fusion (LPBF), aim to produce fully dense parts, hence metallic powder processing gains great importance Recent studies have showed a direct relationship between the particle size distribution of a metallic powder material to the density, surface quality and mechanical properties of the part fabricated using LPBF technology Although process parameters can be adjusted accordingly

to each particle size to enable a densely built part using various particles sizes, but as-built surface roughness will always be greater with larger particles sizes (Spierings, Herres & Levy 2011)

Laser powder bed fusion (LPBF) part characteristics are well documented to be governed by the laser power, scanning speed, focal length, layer thickness and powder material to name

a few The effects of particle size distribution on the mechanical properties of the end result parts has been investigated in detail (Mori et al 2018) Particle size has little effect on elongation but is inversely proportional to the tensile strength of the produced part, at constant energy densities One contributing factor to this is the increased oxygen content during the sieving process of powders and processing of LPBF parts, from the oxidised film on surface of the particles Another factor is the grain size decrease due to repeated heat cycles Therefore it is essential that stockfeed particle sizes be controlled prior to the LPBF process (Mori et al 2018)

The chemistry of the surface of Ti6Al4V components fabricated using additive manufacturing and no secondary process and also Ti6Al4V fabricated using AM, but post-processed using polishing methods, showed some variations in the thickness of the oxide layer thickness, and

a higher aluminium content on the surface than as-built AM Ti6Al4V (Vaithilingam, Prina, et

al 2016), due to this finding, it was decided to measure the changes in chemistry across the various areas in the assembly using electron dispersion spectroscopy (EDS) in this project

Various types of titanium alloys have been investigated for additive manufacturing purposes, alloys such as Ti–24Nb–4Zr–8Sn (Zhang, LC et al 2018), Ti-Si, Ti-8.5Cu, Ti-5Cu, Ti6.5Cu, Ti-7Cu, Ti10Cu (Zhang, D et al 2019) Ti-Ta-Zr, Ti-Nb-Zr and Ti-Fe-Ta (Trevisan et al 2018), Ti-Cr, Ti-

Nb, Ti-Mo, Ti-Ta (Nagase et al 2019) but Ti6Al4V and pure titanium are the most common material used for additive manufacturing and are considered the most common types due to their excellent strength to weight ratio and corrosion resistance characteristics

Significant research has been performed to study the application of Ti6Al4V as a superior material of choice of metal additive manufacturing (Qian et al 2016)

Titanium Ti-6Al-4V equilibrium microstructure contains alpha (α) and beta (β) phases which respectively have hexagonal close packed (HCP), and body centred cubic (BCC) lattice structures The α is stable at room temperature and provides the strength of the alloy, whereas the β phase metastable and contributes to the ductility of the alloy Ti-6Al-4V under rapid cooling conditions, produces a brittle HCP martensitic phase (α’) which is not favourable

to the overall mechanical properties of the part (Galindo-Fernandez et al 2018)

2.4.2 Laser fusion

Multiple experiments have been conducted to establish the relationship between the main parameters in laser powder bed additive manufactured parts such as laser power levels, layer thickness, scanning velocity, hatch spacing, energy density and focal offset distance (FOD) and the mechanical properties of the resulting parts In this section we will look into the literature available for studying these effects under constant feedstock and build shielding gas conditions One of such studies concluded that LPBF fabricated Ti6Al4V can achieve yield strengths over 1300MPa, but tensile elongations to be noticeably below the minimum threshold of 10% suggested for critical structural applications The mechanical properties of LPBF Ti6Al4V are defined by the phases, morphology, length scales, size and orientation of its crystal grains (Xu et al 2015) Laser power levels used and scanning speed used in fabricating LPBF Ti6Al4V samples was noted during the study and their effects documented and correlated to the microstructure and overall behaviour of the samples For a layer thickness

of 60µm and constant energy density, a focal offset distance (FOD) of zero to 2mm was proven

improved ductility o the samples Reducing the energy density, which is governed by equation 2-1, results in preserving alpha-prime martensite phase Thus to achieve an ultrafine lamellar (alpha + beta) structure, it is necessary to maintain energy densities at the desired level (Xu

in performance was assumed to result from process-induced defects; with machining, and heat treatment each having positive impact on the fatigue behaviour of resultant parts (Lewandowski & Seifi 2016)

Figure 2-6 Summary of stress (S) versus cycles to failure (N) (S-N) data for PBF (laser), PBF (E-beam), and wire (DED) at R = 0.1 Metallic Materials Properties Development and Standardization (MMPDS) data for cast, wrought machined data are shown for comparison (Lewandowski & Seifi 2016)

2.4.3 Build shielding gas

Shielding gas flow in the laser powder bed fusion (LPBF) additive manufacturing process plays

a crucial factor in the stability of the laser melting and the consequent layer by layer build-up

of material This is not just to avoid oxidization during the melting process, but also influences the state of the powder bed during the melting process, removing spatter and fumes away from the build area, just in time for the molten area to solidify Shielding gas are usually recirculated in the build chamber The shielding gas used in this research flowed directly on the build area from above (SLM Solutions, SLM 250) The two main by-products of the laser

result of the focused high energy laser beam taking some of the metal elements in the molten pool of the alloy to their evaporation point and ejecting them out of the molten pool (Ladewig

et al 2016) This was one of the reasons that an electron dispersive spectroscopy (EDS) analysis was performed across the welded samples in this research project to verify if such a mechanism has come into play and created any major changes in the chemical composition

in the individual heat affected zones (HAZ) or fusion zone (FZ) The plasma plumes are generated by the ionisation of the gas between the energy (laser) source and the interaction zone (Beck, Berger & Hugel 1995) The second by-product of the laser powder bed fusion process is spatter and ejected powder which upon analysis of spatter formation during laser melting it has been found that the size of such formations are significantly larger in size than the used powder, are spherical and have chemical compositions identical to the original raw powder, all which suggests that spatters are a result of instabilities in the molten pool rather than being the result of vaporisation (Simonelli et al 2015) Laser powder bed fusion process by-products also have secondary effects on the process, namely beam attenuation and reposition of ejected powder and spatter Beam attenuation can be minimised by changing the focal point position of the laser beam and minimise the chances of a “splashy process” (Grünberger & Domröse 2015) Redeposition of spatter material could potentially lead to major variations in layer thickness and decrease in the accuracy of build This is usually in the form of increased layer thickness, caused by large spherical particles from the spatter and redeposits, which can then be torn off the base material by the re-coater and leave behind a hold in the part, as demonstrated schematically in Figure 2-8 (Gong et al 2013)

This research has utilised an argon shielding gas at maximum flow rate of the welding machine, which was 15 litres per minute at 1.5 bars of pressure (SLM_Solutions 2016)

Figure 2-7 Schematic representation of possible process by-products courtesy (Ladewig et al 2016)

Figure 2-8 Schematic of defect induced by recoating during LPBF process (Gong et al 2013)