Post Processing For Nylon 12 Laser Sintered

Post Processing for Nylon 12 Laser Sintered
Components
Ahmad Kamil
079039969
School of Mechanical and Systems Engineering

Newcastle University
Newcastle upon Tyne, United Kingdom
Thesis submitted for partial fulfilment for

the degree of Doctor of Philosophy
November 2016
Post Processing for Nylon 12 Laser Sintered Components
ACKNOWLEDGEMENTS
I would like to express my appreciation and gratitude to my supervisor, Prof. Kenny

Dalgarno, for his guidance and support during the research work.
I gratefully acknowledge the Government of the Republic of Indonesia, through
the Ministry of Research, Technology and High Education for financial support
during my PhD programme.
I would also like to thank my colleagues in Rapid Manufacturing Group at
Newcastle University, for their friendship, helpful discussions, suggestions and
advice. Thanks also to the laboratory operators and technicians for their assistances.
Special acknowledgements go to my lovely wife, Rina Wahyuni, and my
daughters Farah Nabilah and Raisa Kamila, for their understanding and patience
during my study, as well as to Naila Ulfa Safitri (latest).
I deeply express my thanks to my beloved mother, who has always prayed for
me. Without her prayers, I would not have been able to complete this research.
Finally, thanks to all my relatives for their overwhelming support for me and
my family when we have faced difficult situations.
Newcastle upon Tyne, November 2016
Acknowledgements Page i
DEDICATED TO
MY WONDERFUL FAMILY
Acknowledgements Page ii
ABSTRACT
This research investigates the effect of post-processing on the mechanical

characteristics and behaviour of laser sintered components produced by selective laser
sintering (SLS). It aims to understand the material’s behaviour and to develop post-
processing methods that can be used to improve and maintain consistency in the
mechanical properties of the layer manufactured material.
Duraform Polyamide (Nylon 12) and a Sinterstation VanguardTM SLS machine
were used to produce test specimens. The behaviour of the layer material
characteristics was established using different fabrication orientations and tensile,
compression, shear and flexure tests as benchmarking investigations. The results show
that there are significant variations in mechanical properties, as well as divergences
from previous results.
In addition, section thickness in closed and open hollow structures was
investigated in order to establish its effect on mechanical properties. The larger a
sintered area, the greater the tensile properties gained when there is an increase of
section thickness and when solid specimens are used. Moreover, when fill and outline
scanning strategies were implemented in producing the specimens, the improvements
were obtained in the tensile properties of nylon 12 laser-sintered material with no
impact on geometry.
To further improve the mechanical properties, a new post processing method
that included heat treatment in air and vacuum environments was investigated.
Experiments were conducted in air from room temperature to 140oC with a treatment
time of 120 minutes and vacuum heat treatment was conducted from room
temperature to 180oC with 16 hours and 100 hours treatment time. The material
properties in both conditions were then analysed in terms of tensile properties, thermal
characterisations, microstructure and geometrical changes.
Heat treatment in air showed no significant improvement in mechanical
properties. However, Nylon 12 SLS material heat-treated in a vacuum showed
considerable improvement in crystallinity and peak melting point. Heat treatment for
a longer period to approach the melting point, especially on material with the different
section thicknesses and solid specimens and particularly in a vacuum, has a greater
impact on mechanical properties, but this may not be sufficient to justify the cost and
time involved.
Abstract Page iii

TABLE OF CONTENTS
Page
Acknowledgement ......................................................................................... i
Abstract ........................................................................................................ iii
Table of Contents ......................................................................................... iv
List of Figures .............................................................................................. x
List of Tables ................................................................................................ xix
Abbreviations and Notations ......................................................................... xxi
CHAPTER 1. INTRODUCTION 1
1.1 Background ......................................................................................... 1
1.2 Research Gap and Problems ............................................................... 2
1.3 Research Aim and Objectives .............................................................. 4
1.4 Research Scope and Limitation ........................................................... 4
1.5 Structure of the Thesis ........................................................................ 5
CHAPTER 2. LITERATURE REVIEW 7

2.1 Introduction ......................................................................................... 7
2.2 Additive manufacturing (AM) and applications ................................. 7
2.2.1 Basic principles ...................................................................... 7
2.2.2 Development of AM processes .............................................. 9
2.2.2.1 Liquid based process ................................................ 9
2.2.2.2 Powder based process .............................................. 10
2.2.2.2.1 Three dimensional printing (3DP) systems ..... 10
2.2.2.2.2 Selective laser sintering (SLS) ............... 11
2.2.2.3 Solid based process .................................................. 12
2.2.2.3.1 Fused deposition modelling (FDM) ....... 12
2.2.2.3.2 Multi jet modelling (MJM) systems ...... 13
2.2.3 Advantages and disadvantages of AM technology ................ 14
2.2.4 The application of AM Systems ............................................ 16
2.2.4.1 Rapid prototyping (RP) ............................................ 16
2.2.4.2 Rapid tooling (RT) ................................................... 16
2.2.4.3 Rapid manufacturing (RM) ...................................... 17
2.2.5 Rapid manufacturing (RM) .................................................... 17
Table of Contents Page iv
2.2.5.1 Definition and basic process .................................... 17

2.2.5.2 Process chain of rapid manufacturing ...................... 17
2.2.5.3 Issues in RM systems ............................................... 19
2.3 Selective laser sintering (SLS) ............................................................ 20
2.3.1 Definition and classification .................................................. 20
2.3.2 Important processing parameters ........................................... 21
a. Part bed temperature (Tb) .................................................. 22
b. Fill laser power (P) ........................................................... 22
c. Laser beam speed (BS) ...................................................... 23
d. Scan spacing (SP) ............................................................. 23
e. Slice/layer thickness (h) .................................................... 24
f. The energy density (ED) .................................................... 24
2.3.3 Melting viscosity, surface tension and particle size ............... 25
a. Melting viscosity (0) ...................................................... 28
b. Surface tension () ........................................................... 29
c. Particle size ...................................................................... 30
2.3.4 Process parameters and material characteristics .................... 30
2.3.5 Advantages and disadvantages of SLS ................................... 34
2.4 Polymer characteristics ....................................................................... 34
2.4.1 Introduction ............................................................................. 34
2.4.2 Polymer classifications ........................................................... 35
2.4.3 Nylon 12 (Duraform Polyamide) ............................................ 37
2.4.3.1 Mechanical and physical properties ......................... 38
2.4.3.2 Melting, crystallisation and density ......................... 40
2.4.3.3 Degree of crystallinity .............................................. 43
2.4.3.4 Effect of crystallinity on mechanical properties ...... 43
2.5 Heat treatment of polymers ................................................................. 45
2.5.1 Basic process ........................................................................... 45
2.5.1.1 Heating temperature/ time .......................................... 46
2.5.1.2 Heating media .......................................................... 47
2.5.1.3 Heating and cooling rates .......................................... 48
2.5.1.4 Cooling medium ........................................................ 48
2.5.1.5 Distortion .................................................................. 48
Table of Contents Page v
2.5.2 Review of heat treatment of polymers and its impact .............. 49

2.5.3 Review of geometry impact and scanning strategy on mechanical
properties of SLS and SLM .................................................... 52
2.6 Summary ............................................................................................. 56
2.7 Engineering Context of Research Work .............................................. 57
CHAPTER 3. EXPERIMENTAL STUDIES AND PROCEDURE 60

3.1 Introduction .......................................................................................... 60
3.2 SLS machine and materials ................................................................. 60
3.3 Specimen design for testing .......................................................................... 62
3.4 Mechanical testing equipment ............................................................ 63
3.5 Experimental tests selection ................................................................ 63
3.5.1 Tensile test .............................................................................. 63
3.5.2 Compression test ..................................................................... 66
3.5.3 Shear test ................................................................................. 67
3.5.4 Four point bend/flexure test .................................................... 70
3.5.5 Three point bend/flexure test .................................................. 73
3.6 Differential Scanning Calorimetry (DSC) .......................................... 75
3.7 Statistical analysis tool ........................................................................ 76
3.8 Scanning Electron Microscope (SEM) ............................................... 77
CHAPTER 4. MECHANICAL PROPERTY CHARACTERISATION 78

4.1 Introduction ......................................................................................... 78
4.2 Tensile test .......................................................................................... 78
4.3 Compression test ................................................................................. 83
4.4 Shear test ............................................................................................. 85
4.5 Three point bend/flexure test .............................................................. 86
4.6 Four point bend/flexure test ................................................................ 88
4.7 Summary, Benchmarking, and Discussion .......................................... 90
4.7.1 Summary of test results ........................................................... 90
4.7.2 Comparison of test results with other researchers .................. 91
4.7.3 Discussion ................................................................................ 94
Table of Contents Page vi

CHAPTER 5. HEAT TREATMENT STUDIES 97

5.1 Introduction ......................................................................................... 97
5.2 Heat treatment selection ...................................................................... 97
5.3 Specimen preparation .......................................................................... 98
5.4 Characterisation .................................................................................. 98
5.5 Experimental work and procedures ..................................................... 98
5.5.1 Heat treatment in air (140oC, 2 hours) ..................................... 98
5.5.1.1 Tensile tests results .................................................. 99
5.5.1.2 Distortion performance ............................................ 103
5.5.1.3 DSC test results ........................................................ 105
5.5.1.4 Optical microscopy .................................................... 107
5.5.1.5 Heat treatment in air (180oC, 100 hours) ................. 109
5.5.2 Heat treatment in vacuum ........................................................ 109
5.5.2.1 Test results at holding time 100 hours ............................... 110
A. Tensile tests results ............................................. 110
B. Distortion performance ........................................ 112
C. DCS test results ................................................... 113
D. Optical microscopy ............................................. 114
5.5.2.2 Test results at holding time 16 hours ................................. 116
A. Tensile tests results ............................................. 116
B. Distortion performance ........................................ 118
C. DCS test results ................................................... 118
5.6 Summary and Discussion ..................................................................... 120
CHAPTER 6. HEAT TREATMENT AND MECHANICAL PROPERTIES OF

CLOSED AND OPEN HOLLOW STRUCTURES 129
6.1 Introduction ......................................................................................... 129
6.2 Specimen preparation and Material ..................................................... 129
6.3 Equipments, experimental work and procedures ................................ 132
6.4 Impact of Section Thickness ............................................................... 134
6.4.1 Tensile Test Results for Non-heat treated Specimens ............. 134
6.4.2 Crystallinity Characterisation Results for Non-heat treated
Specimens .............................................................................. 139
Table of Contents Page vii

6.4.3 Melting Peak Temperature (Tpeak) Results for Non-heat treated

Specimens ............................................................................... 140
6.4.4 Morphology and Microstructure for Non-heat treated Specimens ... 141
6.5 Impact of Heat Treatment on Properties of Different Section Thickness .... 142
6.5.1 Tensile Tests Results for Heat treated Specimens ..................... 143
6.5.2 Crystallinity Characterisation Results for Heat Treated Specimens .. 145
6.5.3 Melting Peak Temperature (Tpeak) Results for Heat treated Specimens146
6.5.4 Morphology and Microstructure for Heat treated Specimens ....... 147
6.6 Impact of Scanning Strategy (fill and Outline Scanning) on Mechanical
Properties ............................................................................................ 149
6.6.1 Tensile tests results ..................................................................... 151
6.6.2 Energy Density ............................................................................ 153
6.7 Summary and Discussion ...................................................................... 156
6.7.1 Effect of Section Thickness of Closed and Open Hollow Structure .... 163
6.7.2 Effect of heat treatment on properties ......................................... 166
6.7.3 Effect of scanning strategy on mechanical properties ................ 170
CHAPTER 7. CONCLUSIONS AND FUTURE WORK 173

7.1 Summary of Findings and Conclusion on Practical implications of the work173
7.1.1 Summary of Findings …………………………………………… 173
7.1.2 Conclusion on practical implication of the work ……………………. 175
7.2 Achievement of objectives .................................................................. 177
7.3 Research contributions ......................................................................... 178
7.4 Future work ......................................................................................... 179
REFERENCES 182
APPENDIX
A: SHEAR AND BENDING TEST RESULTS ............................ 193
B: ANOVA RESULTS .................................................................... 195
C: DSC TEST RESULTS ............................................................... 206
D: DATA RESULTS FOR CH. 6 .................................................. 218
Table of Contents Page viii

E: JIG AND FIXTURE FOR SHEARING TEST ....................... 223
Table of Contents Page ix

LIST OF FIGURES
Figure 2-1 Principle process of AM .......................................................... 8

Figure 2-2 AM process classification ........................................................ 9
Figure 2-3 Schematic view of stereolithography apparatus (SLA) ........... 10
Figure 2-4 Schematic view of 3DP systems .............................................. 11
Figure 2-5 Schematic view of selective laser sintering (SLS) ................... 12
Figure 2-6 Schematic view of fused deposition modelling (FDM) ........... 13
Figure 2-7 Schematic view of multi jet modelling (MJM) ........................ 14
Figure 2-8 Process chain of rapid manufacturing ...................................... 18
Figure 2-9 Process variables which affect sintering in SLS process ......... 21
Figure 2-10 Processing parameters in SLS process ..................................... 22
Figure 2-11 Scan spacing in the SLS process .............................................. 24
Figure 2-12 Frenkel sintering model ........................................................... 26
Figure 2-13 Cross-section of particle-particle bonding in x, y and z axes (Dx > Dy >
Dz) …………………………………………………………………… 26
Figure 2-14 Melt flow response of polyamide with and without nanoclay
reinforcement .................................................................................... 28
Figure 2-15 Powder bed temperature-density relationship of Duraform Polyamide
.......................................................................................................... 31
Figure 2-16 Relationship between the molecular weight and melting viscosity of
polystyrene ......................................................................................... 32
Figure 2-17 Variation in density (a) and tensile strength (b) with energy density 33
Figure 2-18 Amorphous and crystalline structure of a polymer .................... 35
Figure 2-19 Variation of volume with temperature for amorphous and crystalline
polymers ................................................................................... 35
Figure 2-20 Polymer chain structure (a) linear, (b) branched, (c) cross-linked, (d)
networked ................................................................................. 37
Figure 2-21 Structure of nylon ................................................. ................... 38
Figure 2-22 Nylon12 molecule chain structure ........................................... 38
Figure 2-23 Relationship of crystal growth rate and crystallisation temperature Tc for
different fractions of poly(tetramethyl-p-phenylene) siloxane (molar
mass given in gmol-1) ………………………………………… 41
Figure 2-24 Stress-strain behaviour as a function of crystallinity for Nylon 6
List of Figures Page x

polymerised at 453 K ............................................................... 44

Figure 2-25 Heat treatment cycle .................................................................. 45
Figure 2-26 Main factors influencing heat-up and soak time ....................... 47
Figure 2-27 Thermal soak time for a packed load ........................................ 47
Figure 2-28 Illustration of technique and specifies of helix and progressive scanning
strategy (a) the part model, (b) the random five scan layers with
progressive scan strategy, (c) the random five scan layers with helix scan
strategy ............................................................................................... 54
Figure 2-29 Scanning strategy (a) with “S” scan pattern and (b) with Fractal scan
pattern .......................................................................................... 55
Figure 2-30 Flow diagram of heat treatment of nylon 12 laser sintered product …. 59
Figure 3-1 VanguardTM SLS System ......................................................... 61

Figure 3-2 Orientation of fabrication of test specimens ........................... 62
Figure 3-3 Tensile test specimen drawing ............................................... . 64
Figure 3-4 Tensile test specimens: (a) on testing machine, (b) after testing ... 64
Figure 3 5 Slope measurement on stress-strain curve .................................. 66
Figure 3-6 Compression test specimen drawing .......................................... 66
Figure 3-7 Compression test specimens for X, Y, and Z orientations ......... 67
Figure 3-8 Compression specimen in the test equipment ............................ 67
Figure 3 9 Shear test specimen drawing ...................................................... 68
Figure 3 10 Jig fixture for shear test ............................................................ 68
Figure 3-11 Shear specimen on the test equipment ...................................... 69
Figure 3-12 Load/deflection curve ............................................................... 70
Figure 3-13 Four-point bend/flexure test specimen loading set-up ............. 70
Figure 3-14 Four-point bend/flexure test specimen being tested ................. 71
Figure 3-15 Three-point bend/flexure test specimen loading set-up ........... 73
Figure 3-16 Three point bend/flexure test specimen being tested ............... 74
Figure 3-17 DSC Mettler Toledo 823e ......................................................... 75
Figure 4-1 Stress-strain graphs for each of the tensile tests for the different part
orientations .............................................................................. 79
Figure 4-2 Tensile strength of different fabrication orientations (distribution for all
specimens) ............................................................................... 80
List of Figures Page xi

Figure 4-3 Linear fit line on stress-strain relationship of specimen in tensile test for
the Y oriented parts .................................................................. 80
Figure 4-4 Tensile modulus of different fabrication orientations (distribution for all
specimens) ................................................................................ 81
Figure 4-5 Tensile elongation at break of different fabrication orientations
(distribution for all specimens) ............................................... 82
Figure 4-6 Stress-strain graphs for each of the compression tests for the different
orientation of parts ................................................................... 83
Figure 4-7 Linear fit line on stress-strain relationship of specimen in compression test
for the Y-oriented parts ............................................................ 84
Figure 4-8 Compression modulus of different fabrication orientations (distribution for
all specimens) .......................................................................... 84
Figure 4-9 Load-deflection graphs for each of the shear tests for the YZ and ZY
oriented parts ............................................................................ 85
Figure 4-10 Shear modulus of different fabrication orientations (distribution for all
specimens) ............................................................................... 86
Figure 4-11 Stress-strain graphs for each of the three point bending tests for the YX
and ZX oriented parts .............................................................. 86
Figure 4 12 Linear fit line on stress-strain relationship of specimen in three point
bending/flexure test for the ZX-oriented parts ........................ 87
Figure 4 13 Three point flexural modulus of different fabrication orientations
Figure 4-14 Stress-strain graph for each of the four point bending tested for the XY
and YZ oriented parts .............................................................. 88
Figure 4-15 Linear fit line on stress-strain relationship of specimen in four point
bending/flexure test for the XY-oriented parts ....................... 89
Figure 4-16 Four point flexural modulus of different fabrication orientations
Figure 4-17 Summary of result for tensile, compression, shear and flexural modulus
(three- and four-point) of nylon 12 laser sintered material in the X, Y, Z, XY,
YX, ZX, XZ, YZ and ZY orientations ……………………………. 90
Figure 4-18 Comparison of results of tensile modulus test ………………….91
Figure 4-19 Comparison of results of compression modulus tests ……………92
Figure 4-20 Comparison of results of test of flexural modulus (three-point) …….93
List of Figures Page xii
Figure 4-21 Comparison of results of test of flexural modulus (four-point) …… 93

Figure 4-22 Comparison of tensile strength results ……………………………. 94
Figure 5-1 Heating oven and tensile specimen …………………………… 99
Figure 5-2 Stress-strain graphs for each of the tensile tests in X, Y and Z orientation
(a), (c) and (e) without heat treatment; (b), (d) and (f) with heat
treatment………………………………………………………….… 100
Figure 5-3 Tensile strength of different fabrication orientations with and without heat
treatment (distribution for all specimens). The suffix ht indicates heat
treatment ........................................................................................... 101
Figure 5-4 Tensile modulus result with and without heat treatment (distribution for all
specimens).The suffix ht indicates heat treatment ................... 102
Figure 5-5 Tensile elongation at break result with and without heat treatment
(distribution for all specimens). The suffix ht indicates heat treatment.
……………………………………………………………………… 103
Figure 5-6 Solid tensile specimen .......................................................................... 103
Figure 5-7 DSC plots for X orientation specimen sample without heat treatment 105
Figure 5-8 Crystallinity results with and without heat treatment (distribution for all
specimens). Suffix ht indicates heat treatment ......................... 106
Figure 5-9 Melting peak temperature result with and without heat treatment
........................................................................................................ 106
Figure 5-10 Microstructure samples with and without heat treatment in air, with 5X
magnification: (a) sample no.3-X orientation without heat treatment; (b)
sample no.3-X orientation with heat treatment; (c) sample no.2-Z
orientation without heat treatment; (d) sample no.2-Z orientation with
heat treatment; (e). the slicing direction of the section .............. 108
Figure 5-11 Cut-off specimens that burned after heat treated in air atmosphere for
180oC and 100 hours . ............................................................. 109
Figure 5-12 Vacuum oven facilities and tensile specimens ........................... 110
Figure 5-13 Stress-strain graphs for each of the tensile tests for Y oriented parts, (a)
without heat treatment and (b) with heat treatment in vacuum for 100
hours ........................................................................................... 110
Figure 5-14 Tensile strength result with and without heat treatment (distribution for
all specimens). The suffix ht indicates heat treatment . ........... 111
List of Figures Page xiii
Figure 5-15 Tensile modulus result with and without heat treatment (distribution for
all specimens). The suffix ht indicates heat treatment. ...................... 111
(distribution for al specimens). The suffix ht indicates heat treatment.
............................................................................................................ 112
Figure 5-17 Crystallinity result with and without heat treatment (distribution for all
specimens). The suffix ht indicates heat treatment. ........................... 113
(distribution for all specimens). The suffix ht indicates heat
treatment…………………………………………………….……….114
Figure 5-19 Microstructure samples with and without heat treatment in vacuum, with
5X magnification: (a) sample no.2-Y orientation without heat treatment;
(b) sample no.2-Y orientation with heat treatment; (c) sample no.6-Y
orientation without heat treatment; (d) sample no.6-Y orientation with
heat treatment ..................................................................................... 115
without heat treatment and (b) with heat treatment in vacuum for 16
hours ......................................................................................... 116
all specimens). The suffix ht indicates heat treatment ............. 116
all specimens). The suffix ht indicates heat treatment .............. 117
(distribution for all specimens).The suffix ht indicates heat treatment
........................................................................................................... 118
specimens). The suffix ht indicates heat treatment . ................. 119
.......................................................................................................... 119
Figure 5-26 Summaries of test results: average tensile strength with and without heat
treatment. The X, Y and Z indicate fabrication orientation. The suffix ht
indicates heat treatment. Error bars are 95% confidence interval...... 120
List of Figures Page xiv

Figure 5-27 Summaries of test results: average tensile modulus with and without heat
Figure 5-28 Summaries of test results: average elongation at break with and without
heat treatment. The X, Y and Z indicate fabrication orientation. The
suffix ht indicates heat treatment. Error bars are 95% confidence interval
........................................................................................................... 122
Figure 5-29 Summaries of test results: average crystallinity with and without heat
Figure 5-30 Summaries of test results: average melting peak temperature with and
without heat treatment. The X, Y and Z indicate fabrication orientation.
The suffix ht indicates heat treatment. Error bars are 95% confidence
interval ..................................................................................... 124
Figure 6-1 Tensile test specimens drawing with shell and closed hollow structure
with different section thickness (a) 1mm, (b) 2mm and (c) 3mm...... 130
Figure 6-2 Tensile test specimens drawing with shell and open hollow structure with
2mm section thickness……………………………………………….131
Figure 6-3 Specimens position and orientation in the Vanguard SLS machine 132
Figure 6-4 Cross section of hollow specimen …………………………………. 133
Figure 6-5 Stress-strain graphs for each of the tensile tests non-heat treated
specimens for closed hollow with section thickness (a) 1mm, (b) 2mm,
(c) 3mm and (d) solid ……………………………………….. 134
Figure 6-6 Stress-strain graphs for each of the tensile tests non-heat treated
specimens for open hollow with 2mm section thickness …… 134
Figure 6-7 Tensile test specimens (after testing) for closed hollow structure with
1mm, 2mm and 3 mm section thickness ……………………. 135
Figure 6-8 Tensile strength result of different section thickness of closed hollow and
solid non-heat treated specimens. Error bars are at 95% confidence
interval ……………………………………………… …... 135
Figure 6-9 Tensile modulus result of different section thickness of closed hollow
and solid non-heat treated specimens. Error bars are at 95% confidence
interval………………………………………………………………. 136
List of Figures Page xv

Figure 6-10 Elongation at break results of different section thickness of closed

hollow and solid non-heat treated specimens. Error bars are at 95%
confidence interval …………………………………………… 137
Figure 6-11 Summaries of test results: average tensile strength of closed and open
hollow structure specimens. Error bars are at 95% confidence interval
………………………………………………………………………. 137
Figure 6-12 Summaries of test results: average tensile modulus of closed and open
hollow structure specimens. Error bars are at 95% confidence
interval……………………………………………………………… 138
Figure 6-13 Summaries of test results: average elongation at break of closed and
open hollow structure specimens. Error bars are at 95% confidence
interval ................................................................... ....................138
Figure 6-14 Degree of crystallinity of sintered and un-sintered material of non-heat
treated specimens with different section thickness and solid. Error bars
are at 95% confidence interval ...................................................139
Figure 6-15 Melting peak temperature sintered and un-sintered material of non-heat
treated specimens with different wall thickness and solid. Error bars are
at 95% confidence interval ....................................................... 140
Figure 6-16 Microstructure samples without heat treatment in vacuum; sample with
3mm section thickness (a) surface cutting, (b) in the middle, (c) outer
edge (d) inner edge of specimen ……………………………… 141
Figure 6-17 Un-sintered material particle grain (without heat treatment) …… 142
Figure 6-18 Stress-strain plots for each of the tensile tests of heat treated specimens
with section thickness, (a) 1mm, (b) 2mm, (c) 3mm and (d) solid… 143
Figure 6-19 Tensile strength results of different section thickness of closed hollow
and solid heat treated specimens. Error bars are at 95% confidence
interval ………………………………………………………… 144
Figure 6-20 Tensile modulus results of different section thickness of closed hollow
and solid heat treated specimens. Error bars are at 95% confidence
interval …………………………………………………………. 144
hollow and solid heat treated specimens. Error bars are at 95%
confidence interval ....................................................................145
List of Figures Page xvi

Figure 6-22 Degree of crystallinity of sintered and un-sintered material result of heat
are at 95% confidence interval ………………………………. 146
Figure 6-23 Melting peak temperature of sintered and un-sintered material heat
are at 95% confidence interval ……………………………….. 147
Figure 6-24 Microstructure specimens with heat treatment in vacuum, sample with
3mm section thickness (a) surface cutting; (b) outer edge (c) inner edge
(d) in the middle of specimen ………..……………………… 148
Figure 6-25 Scan path of fill and outline scanning strategy for (a) solid and (b)
section thickness specimens …………………………………. 150
Figure 6-26 Stress-strain graphs for each of the tensile tests with fill and outline
scanning specimens for closed hollow with section thickness, (a) 1mm
(b) 2mm, (c) 3mm and (d) solid ............................................... 151
and solid specimens with fill and outline scanning. Error bars are at 95%
confidence interval ................................................................... 151
and solid specimens with fill and outline scanning. Error bars are at 95%
confidence interval ............................................................. … 152
hollow and solid specimens with fill and outline scanning. Error bars are
at 95% confidence interval ………………………………….. 153
Figure 6-30 Scan path length of Fill (F) and Fill and outline (F/O) for different
section thickness and solid specimens …………………………154
Figure 6-31 Pythagorean theory of each layer ……………………………. 155
Figure 6-32 Total energy density of different section thickness and solid specimens
for non-heat treated with fill (non-HT with F scan) and non-heat treated
with fill and outline (non-HT with F/O scan) strategy ………. 155
Figure 6-33 Comparison of average tensile strength of different section thickness and
solid specimens of non-heat treated (non-HT) with fill (F) scan, heat
treated (HT) with fill (F) scan and non-heat treated (non-HT) with fill
and outline (F/O) scan. (*) open hollow and others are closed hollow
structure. Error bars are at 95% confidence interval …………. 156
List of Figures Page xvii
Figure 6-34 Comparison of average tensile modulus of different section thickness and
solid specimens of non-heat treated (non-HT) with fill (F) scan, heat
and outline (F/O) scan. (*) open ……………..……………… 157
Figure 6-35 Comparison of average elongation at break of different section thickness
and solid specimens of non-heat treated (non-HT) with fill (F) scan, heat
and outline (F/O) scan. (*) open hollow and others are closed hollow
structure. Error bars are at 95% confidence interval ………… 159
Figure 6-36 Summaries of test results: average crystallinity sintered material with and
without heat treatment of different section thickness and solid specimens.
Error bars are 95% confidence interval ……………………… 160
Figure 6-37 Summaries of test results: average crystallinity un-sintered material with
and without heat treatment of different section thickness and solid
specimens. Error bars are 95% confidence interval …………. 161
Figure 6-38 Summaries of test results: average melting peak temperature sintered
material with and without heat treatment of different section thickness
and solid specimens. Error bars are 95% confidence interval ...162
Figure 6-39 Summaries of test results: average melting peak temperature un-sintered
material with and without heat treatment of different section thickness of
specimens. Error bars are 95% confidence interval ................... 162
Figure 6-40 End of Vector (EoV) effect (a). Hollow structure and (b). Solid
structure ……………………………………………….................... 164
Figure 6-41 Average energy density and average tensile strength relationship for
specimens with default fill scanning and fill and outline scanning, with
1mm, 2mm, 3mm section thickness and solid ………….......... 171
Figure 6-42 Average total energy density and average tensile modulus relationship
for specimens with default fill scanning and fill and outline scanning,
with 1mm, 2mm, 3mm section thickness and solid …………. 171
Figure 6-43 Average total energy density and average elongation at break
relationship for specimens with default fill scanning and filling and
outline scanning, with 1mm, 2mm, 3mm section thickness and solid 171
List of Figures Page xviii

LIST OF TABLES
Table 2-1 Mechanical Properties of Polymers ........................................................ 39

Table 2-2 Physical Properties of SLS powder of 3D Systems ............................ 40
Table 2-3 Advantages and disadvantages of heat treatment fluid .................. 48
Table 2-4 Summary of Heat Treatment Studies of Nylon and Polymer
Materials ........................................................................................... 51
Table 3-1 Process parameters used for specimen fabrication .......................... 61
Table 3-2 Duraform PA material properties ..................................................... 61
Table 4-1 Tensile strength results (MPa) ......................................................... 79
Table 4-2 Tensile modulus results (MPa) ....................................................... 80
Table 4-3 Tensile elongation at break (%) ...................................................... 81
Table 4-4 Compression modulus results (MPa) ............................................. 84
Table 4-5 Shear Modulus Results (GPa) ......................................................... 85
Table 4-6 Three point flexural modulus results (MPa) .................................... 87
Table 4-7 Four point flexural modulus results (MPa) ....................................... 89
Table 4-8 Comparison of results of tensile and compression modulus of
Duraform PA (SLS) material .......................................................... 91
Table 4-9 Comparison of results of flexural modulus of Duraform PA (SLS)
Material (Three-point test) ............................................................... 92
Table 4-10 Comparison of results of flexural modulus of Duraform PA (SLS)
material (Four-point test) .............................................................. 93
Table 4-11 Comparison of tensile strength results .......................................... 94
Table 5-1 Tensile Strength Results (MPa) – With and Without Heat Treatment ... 100
Table 5-2 Tensile Modulus Results (MPa) - With and Without Heat Treatment ... 101
Table 5-3 Tensile Elongation at Break (%) - With and Without Heat Treatment… 102
Table 5-4 Distortion for Heat Treated Sample (X Orientation) ..................... 104
Table 5-5 Distortion for Heat Treated Sample (Y Orientation) ..................... 104
Table 5-6 Distortion for Heat Treated Sample (Z Orientation) ........................ 104
Table 5-7 Crystallinity Results (%) – With and Without Heat Treatment ........ 105
Table 5-8 Melting Peak Temperature (oC) – With and Without Heat Treatment... 106
Table 5-9 Tensile Strength Results (MPa) – With and Without Heat Treatment .... 111
Table 5-10 Tensile Modulus Results (MPa) – With and Without Heat Treatment . 111
Table 5-11 Tensile Elongation at Break (%) – With and Without Heat Treatment . 112
List of Tables Page xix

Table 5-12 Distortion for Heat Treated Sample (Y Orientation) .................... 113
Table 5-13 Crystallinity Results (%) – With and Without Heat Treatment ....... 113
Table 5-14 Melting Peak Temperature (oC) – With and Without Heat Treatment . 114
Table 5-15 Tensile Strength Results (MPa) – With and Without Heat Treatment .. 116
Table 5-16 Tensile Modulus Results (MPa) – With and Without Heat Treatment . 117
Table 5-17 Tensile Elongation at Break (%) – With and Without Heat Treatment . 117
Table 5-18 Distortion for Heat Treated Sample (Y Orientation) .................. 118
Table 5-19 Crystallinity Results (%) – With and Without Heat Treatment ....... 119
Table 5-20 Melting Peak Temperature (oC) – With and Without Heat Treatment . 119
Table 6-1 Relationship of changes in mechanical property and changes in
degree of crystallinity after heat treated .......................................... 168
Table 7-1 Comparison of Mechanical Properties of Specimens with: default
fill scanning strategy and non-heat treated (A); default fill scanning
strategy and heat treated (B); fill and outline scanning strategy and
non-heat treated (C) ....................................................................... 176
List of Tables Page xx

ABBREVIATIONS AND NOTATIONS
ABBREVIATIONS
3DP : Three dimensional printing
AM : Additive manufacturing
CAD : Computer aided design
DMLS : Direct metal laser sintering
EBM : Electron beam melting
DLF : Direct laser formin
DSC : Differential scanning calorimetry
FDM : Fused deposition modeling
LC : Laser Cusing
MJM : Multi Jet Modelling
OLM : Optical light microscopy
PA : Polyamide
RM : Rapid manufacturing
RP : Rapid prototyping
RT : Rapid tooling
SLA : Stereolithography
SLM : Selective laser melting
SLS : Selective laser sintering
SPLS : Solid phase laser sintering
NOTATIONS
 : the surface tension (mN/m)
 : thermal absortivity ratio
 : the characteristic of beam radius (mm)
 : the apparent density (kg/m3)

 : the solid fraction  
s
o : the melting viscosity (Pa.S)
S : theoretical solid density (kg/m3)
BS : beam speed (mm/sec)
Abbreviation and Notations Page xxi

Cp : specific heat (J/kg K)

d : laser beam diameter (mm)
E : the activation energy for viscous flow (J)
E/R : ratio of sintering activation energy to the gas constant (K -1)
ED : energy density (J/mm2)
h : slice/layer thickness (mm)
I : the intensity of the beam (W/mm2)
k : thermal conductivity (W/m K)
Mw : molecular weight (mg)
m : mass of material (kg)
P : laser power (W)
R : Scan radius (mm)
r : the radius of a sphere (mm)
s : laser scan spacing (mm)
SP : scan spacing (mm)
T : absolute temperature (oC or K)
Tcc : crystallization temperature (oC)
Tm : melting temperature (oC)
Tb : powder bed temperature (oC)
Tg : glass transition temperature (oC)
t : time (s)
w : laser scan vector length or part width (mm)
x : sintering neck radius (mm)
Abbreviation and Notations Page xxii

CHAPTER 1. INTRODUCTION
1.1 Background
Manufacturing industry is seeking innovative methods to accelerate new product

development and to manufacture products more rapidly, due to global competition
among companies striving to improve their business processes. In this situation, rapid
manufacturing has become an attractive option.
Rapid manufacturing has developed as an advanced technology used to shorten
the time to market. Such technologies are based on an additive approach, where a
component is built gradually by adding material layer by layer so as to create a
geometry of components in two-dimensional profiles, controlled directly from 3D
computer-aided design (CAD) solid model systems. This additive manufacturing
approach has several advantages. The technologies are able to fabricate parts rapidly
without the use of any additional tooling or moulds for any level of geometrical
complexity [Kruth et al, 1998; Pham, 2001; Levy, 2002]. In addition, it allows
reduced design and production lead-times and cost.
Many different commercial additive manufacturing technologies are available,
such as stereolithography (SLA), selective laser sintering (SLS), fused deposition
modelling (FDM), solid ground curing (SGC), and three-dimensional printing (3DP).
All these technologies have limitations concerning the types and properties of
materials that can be used [Gibson and Shi, 1997; Dutta et al, 2001; Pham, 2001].
Selective laser sintering (SLS) was developed in the late 1980s as a technique
for rapid prototyping (RP) using a wide range of materials, including metals (steel,
titanium, alloy mixtures and composites), polymers (nylon, glass-filled or other fillers,
and polystyrene), and foundry sand [Kruth et al, 2003; Levy et al, 2003]. Hence, SLS
is frequently used for prototyping, tooling and manufacturing purposes.
In order to manufacture products, acceptable material characteristics must be
available. The properties of materials, and mainly the mechanical properties of
products such as strength and stiffness, are critical issues regarding service loading
and operation [Caufield et al, 2007].
Chapter 1. Introduction Page 1

1.2 Research Gaps and Problems
In conventional manufacturing processes, the repeatability of mechanical properties is

known. However, in layer manufacture material properties are less well understood
and depend on the orientation of fabrication [Gibson and Shi, 1997], where the tensile
strength and density of specimens produced in different orientations with different
fabrication parameters may vary. In addition, Hague et al [2004] and Hou [2005]
concluded that SLS material can be considered as anisotropic, showing different
levels of strength and modulus in different orientations.
Another issue in the rapid manufacturing field concerns mechanical properties
and improvements. Zarringhalam, et al [2006] investigated the effects of various
processing factors on the microstructure and properties of SLS Nylon 12, analysing
different specimens of SLS Nylon 12 powder and parts produced using different
thermal histories and various machines configured with contrasting parameters.
Significant increases in elongation at break were found for refreshed and used powder
compared with virgin powder for parts made on different machines. Small variations
were also demonstrated in ultimate strength between virgin and refreshed material,
with increases in tensile strength for used powder. The values of Young’s modulus
also varied, even though the same machine and material were used to fabricate
specimens.
Used powder is the powder left from every run of the machine during the SLS
process. Some may have overflown from bins, some powder remains from the part
cake after its removal from the machine, and also powder is removed from the part
build and feed cylinders when the machine is emptied. However, in reusing this
powder, there are concern involving how many times the powder has been used and
where it was extracted from in the machine, since a certain amount of degradation
occurs due to increases in molecular weight which hinder the sintering process. Part
quality and indeed the ability to build is reduced when used powder is reused in
repeated builds (Gornet, T.J., et al, 2002).
Virgin powder is as supplied from the manufacturer. However, virgin powder is
currently expensive and it is more practical in industry to use a refreshed powder, in
which used powder is mixed predefined proportions of virgin powder. However, this
does not solve the problem of variations in properties.

It has been recognised that material properties are influenced by microstructure,

the formation of which depends on temperature and process duration. In addition, the
properties of a structure can be affected by different temperatures, times and
heating/cooling rates in heat treatment processes.
Several studies have investigated the heat treatment of polymer materials.
Zarringhalam and Hopkinson [2003] reported that using heat treatment in a
conventional heating chamber as post-processing can increase the tensile and impact
strength of SLS parts made from Nylon 12. In addition, Scobbo and Hwang [1994]
showed that after annealing PolyPhenyleneSulfide material, its modulus increased as a
result of an increase in crystallinity.
The effect of annealing on the thermal properties and crystalline structure of
Nylon 11 (polyundecaneamide) and Nylon 12 (polylaurolactam) was investigated by
Gogolewski [1979] and Gogolewski, et al [1980], who found that crystal size, melting
temperature and the heat of melting rose with increased annealing temperature.
Ramazani and Mousavi [2005] investigated vacuum annealing effects on the
mechanical properties of thermoplastic materials, polystyrene and Nylon 6. They
studied the heat treatment process at different vacuum levels and for different time-
temperature settings. They showed that heat treatment in a vacuum has a positive
impact on the mechanical properties of polymeric parts.
From the brief summary above of other research, it can be concluded that the
mechanical properties of laser-sintered material depend on the process parameters,
materials and orientation and position of design products in the envelope of the
machine when it is manufactured. Furthermore, some research has utilised the
technology of heat treatment to improve the characteristics of polymer material.
Questions arise therefore concerning techniques that could be applied after a
product is finished to improve its properties and characteristics. Can a heat treatment
approach serve this purpose?
For example, the mechanical properties of laser-sintered material might be
improved if the product is designed with closed and open hollow structures with
different wall thicknesses using a default setting for a filling and outline scanning
strategy. Then, if the heat treatment approach is applied, it may show significant and
consistent changes in properties.

Furthermore, if the filling and outline scanning strategy changed for product
designs with closed hollow structures but with different wall thickness, this might lead
to significant changes in mechanical properties and their consistency?
Therefore, in this study the mechanical properties of Nylon SLS parts are
investigated and methods for improving and making more consistent its mechanical
properties using heat treatment are evaluated.
1.3 Research Aim and Objectives
The main purpose of this study is to investigate post-processing methods that can be
used to improve and ensure the consistency of mechanical properties in layer
manufacture materials.
To achieve the aim of this study, the following objectives were determined:
1. To determine the mechanical performance of one SLS material (Duraform
PA), over a range of stress states.
2. To develop post-processing methods for the material.
3. To determine and understand the impact that the post-processing methods
have on the consistency of mechanical properties.
1.4 Research Scope and Limitation
In order to focus the investigation, further steps were determined to achieve the aim
and objectives of this research. The research map consists of the following steps
below:
1) Identifying a basic technology of additive manufacturing (AM).
a) Benchmarking on mechanical properties obtained by other researchers
from different SLS machines.
b) Identification of research gap and opportunity
2) Identifying current research into the heat treatment of polymer material.
a) Reviewing previous relevant studies in improving the tensile properties of
SLS fabricated parts.
b) Identifying a heat treatment approach to improve the tensile properties of
laser-sintered material.

3) Identifying current research on the tensile properties of laser-sintered products

with different wall thicknesses and possible processes to improve them with a
heat treatment approach.
4) Identifying current research on filling and outline scanning strategy and
possible processes to gain better tensile properties of laser-sintered material
with the aforementioned strategy.
In this research, was initially performed benchmarking for mechanical

properties obtained by others researchers from different SLS machines, and the tests
involved tensile, compression, flexural and shear tests without a heat treatment
process. However, further investigation of closed and open hollow structures with
different wall thicknesses, are limited to the tensile test only to determine the tensile
properties of heat-treated and non-heat-treated laser-sintered material. Moreover, the
same test is conducted to compare samples of closed hollow structures with different
filling and scanning outline strategies.
In this study, six samples are measured in each individual test for all
experimental work, whether for heat-treated or non-heat treated samples. Refreshed
powder is used for all samples manufactured, as this is applied in practice in industry
using SLS machines.
In initial heat treatment, a regular furnace is used. However, for further heat
treatment study, a vacuum furnace is used to avoid contaminated environments.
1.5 Structure of the Thesis
This thesis consists of seven chapters. The second chapter is a literature review that
considers the fundamentals of additive manufacturing technology and its applications,
and gives definitions of rapid manufacturing, selective laser sintering (SLS),
materials, and post processing.
Chapter three details the experimental studies and procedures used, including
the standards, test methods and analysis, and the mechanical testing equipment
employed in this study. The subsequent chapter then presents the experimental results
and a discussion of the benchmarking of the characterisation of mechanical properties.
Chapter five describes the heat treatment studies, showing details of the heat
treatment regimes applied with selective laser-sintered Nylon 12 and their effect on

mechanical properties. In addition, heat-treated and non-heat-treated specimens are

compared and the results of the characterisation are discussed.
Chapter six describes the heat treatment and mechanical properties of closed
and open hollow structures, with different wall thicknesses of samples with a default
filling and outline scanning strategy. Specimens with and without heat treatment are
compared and the results of the characterisation are discussed.
Moreover, the mechanical properties of closed hollow specimens with different
wall thicknesses that using different filling and outline scanning strategies are
explained. Then, these closed hollow specimens with different filling and outline
scanning strategies are compared and the results of the characterisation are discussed
for specimens without heat treatment.
Chapter seven consists of a general discussion of the work, considering the
implications of the results and outlining the conclusions of the study, before making
recommendations for future work.

CHAPTER 2. LITERATURE REVIEW
2.1 Introduction
This chapter describes the fundamentals of additive manufacturing technologies and

their systems development and application, and outlines a definition of rapid
manufacturing and its technology. In addition, it also describes selective laser
sintering (SLS) as the rapid manufacturing technology considered in this research
study, and the characteristics of Nylon 12 as the material studied. Moreover, heat
treatment as a post-processing method is also reviewed.
2.2 Additive Manufacturing (AM) and Applications
Additive manufacturing (AM) was introduced commercially in the late 1980s [Kruth
et al, 2005a] by 3D-Systems, as a stereolithography process. The first application of
this technology was to build solid products for prototyping purposes, hence the name
rapid prototyping (RP). The technology was the developed for tool making, known as
rapid tooling (RT), and to manufacture final (functional) parts rapid manufacturing
(RM).
AM is a set of techniques that have the ability to generate physical models of
free/form or complex-geometry parts directly from 3D CAD data by adding material
on a layer-by-layer basis until the final geometry is obtained, without the use of
tooling [Kruth et al, 1998; Dalgarno and Wright, 2003]. Some of the commercial
advantages gained from AM technology are: decreased cycle time and cost in design;
earlier detection and reduction in the potential for expensive design errors; decreased
tooling costs on short-run parts; reduced time in production and marketing; the part
geometry can be visualised in advance of production due to its physical existence
[Pham et al, 2004; Noorani, 2006]. AM technologies have been applied to produce
components or final parts using a wide range of different materials and methods. The
basic principles of additive manufacturing and its system development and application
are outlined in subsequent sections.
2.2.1 Basic principles
Since AM technologies were introduced, many different processes using different

materials have become available. However, although these processes vary, their basic
Chapter 2. Literature Review Page 7

operating principles are very similar [Pham, 2001; Cee Kai, 2003; Noorani, 2006], as
illustrated in figure 2-1.
All AM technologies require a solid 3D-model file to begin with, so that the first
step is to create a 3D CAD solid model of the design as input data. Then, by using any
solid modelling CAD software, this model is converted to a stereolithography (STL)
file, which is the “de facto” standard industry format [Pham, 2001; Gibson, 2002; Cee
Kai, 2003; Noorani, 2006]. Basically, the STL file format represents a set of triangular
facets to indicate the geometric model of the part‟s surface [Gibson, 2002]. The more
triangles are used, the more accurate a representation is given of the original part the
STL file is for. This STL file is then transferred to the AM system.
Figure 2-1 Principle processes of AM [Noorani, 2006]
In the following step, the STL format of the 3D model is reviewed by the AM
system software. It is then “split” into a stack of 2D cross-sections which are the
layers. In the final step, each layer is then processed and bonded to form a physical
3D solid part. The build platform descends in the Z direction after one layer thickness,
and layers are then added consecutively to build the part [Pham et al, 2004]. The 3D
solid part is then removed from the machine after being totally built for any post-
processing to be applied.

2.2.2 Development of AM processes
Additive manufacturing has been developed for a wide range of different applications
and materials. Additive manufacturing systems can be categorised in a variety of ways
depending on the physics of the process, the source of energy, type of material, or size
of parts [Kruth et al, 1998; Pham, 2001; Noorani, 2006]. According to the raw
material used, additive manufacturing systems can be divided into: liquid-based,
powder-based or solid-based processes, as shown in Figure 2-2 [Kruth et al, 1998;
Pham, 2001; Noorani, 2006]. The basic processes and operation of the most common
commercial AM systems based on this classification is reviewed in the following
section.
Additive Manufacturing
(AM)
Liquid Based Process Powder Based Process Solid Based Process
Figure 2-2 AM process classification [Kruth et al, 1998; Pham, 2001; Noorani, 2006]
2.2.2.1 Liquid-based process
The stereolithography apparatus (SLA) is the most common liquid-based process, and
was the first commercial implementation of the AM system introduced by 3D
Systems in 1988 [Cee Kai, 2003; Noorani, 2006].
The process of a liquid-based system begins with the raw material, a
photosensitive polymer in the liquid state which is then converted into a solid state
through a curing process. A schematic view of the SLA process is illustrated in figure
2-3 [Upcraft and Fletcher, 2003], where the three-dimensional part‟s model is built in
a vat. Based on the predefined path in slicing the model, the surface layer of resin is
cured or solidified selectively by a UV (ultraviolet) laser beam to form a solid layer.
When one layer of the part is formed, the elevation systems control causes the
elevator table to lower the part, and then a new layer of liquid resin is swept over it
and the process repeated until the complete part is built [Cee Kai, 2003; Noorani,
2006]. The part is then removed from the vat and cleaned of uncured material.

SLA technology can produce parts with a good surface finish, high degree of
detail and thin walls. Drawbacks include a limited range of materials (photopolymer);
and the needs for a support structure during processing and postcuring [Noorani,
2006].
Figure 2-3 Schematic view of stereolithography apparatus (SLA) [Upcraft and

Fletcher, 2003]
2.2.2.2 Powder-based process
Powder-based processes are a special category of solid-based processes that use

powder as raw material. A wide range of polymers, metals and ceramics may be used
[Kruth et al, 2005b]. Three-dimensional printing (3DP) and selective laser sintering
(SLS) processes are the most popular type in this category.
2.2.2.2.1 Three-dimensional printing (3DP) systems
This layer manufacturing technology was invented and patented by the Massachusetts
Institute of Technology (MIT). It utilises ink-jet printing technology directly for the
rapid and flexible production of prototypes, parts and tooling [Noorani, 2006]. Firstly,
a thin layer of powder material is spread over the surface of the powder bed. Next, by
using ink-jet printing technology, a binder material is “printed” into the powder where
the part model is to be formed and the powder particles are selectively bonded to form
the solid part layer. Then the platform is lowered, a fresh powder layer is spread and
deposited over the previous solid layer, and the binder prints again. This process is

continued until the part is completely developed [Noorani, 2006]. Figure 2-4 shows a
basic schematic of three-dimensional printing (3DP) systems. For a metal part, the
3DP process is used firstly to build a green-part, which is then post-processed with
debinding, sintering and infiltrating in a furnace.
Figure 2-4 Schematic view of 3DP systems [Upcraft and Fletcher, 2003]
2.2.2.2.2 Selective laser sintering (SLS)
Selective laser sintering (SLS) was developed and patented by the University of Texas
at Austin and was commercialised by the DTM Corporation (later bought by 3D
Systems) [Noorani, 2006].
Selective laser sintering (SLS) builds a 3D solid part model by laser-fusing or
sintering a layer of powdered raw material. Firstly, the powder is transferred to the
build cylinder platform from the feed powder container through a counter-rolling
cylinder or blade. When a thin layer of the heat-fusible powder is spread and
deposited over the build platform, a concentrated heating laser beam selectively
sinters the powder, fusing the powder particles and forming a solid part. The powder
is spread again for the next layer and the process is repeated, so that layers of powder
are deposited and sintered until the part building is complete. Unsintered material
remains loose and can act as a support for the next powder layer and object under
fabrication. When the object is completely formed, the piston is raised to elevate the

object. Excess powder is brushed away and final hand finishing may be carried out.
Figure 2-5 illustrates a schematic of the SLS process.
Selective laser sintering is described in greater detail in section 2.3 below.
Figure 2-5 Schematic view of selective laser sintering (SLS) [Kruth et al, 2005a]
2.2.2.3 Solid-based process
Solid-based systems begin with the build material in a solid (non-powder) state,
which may include material in the form of wire, a roll, laminates or pellets. The most
common systems are fused deposition modelling (FDM) and multi-jet modelling
(MJM) systems.
2.2.2.3.1 Fused deposition modelling (FDM)
Fused deposition modelling (FDM) technology was developed by the Stratasys

company and a patent was issued in 1992 [Noorani, 2006]. FDM generates three-
dimensional parts using an extrusion process. A schematic process is shown in figure
2-6. To develop a 3D solid part, a filament of thermoplastic is extruded from a heated
dispensing nozzle and deposited onto a platform. The nozzle is laid across the X-Y
plane to form a thin 2D layer. To build each layer, the dispensing nozzle deposits the
outline of the layer first and then the next slice of the part is deposited. As each layer
is built up with hot filament, it bonds to the material in the previous slice to develop a
3D solid part. To build up the support structure, when needed, another nozzle is used
with different material [Pham, 2001; Noorani, 2006].

There are certain advantages to FDM, such as the variety of materials which can
be used its ease of material changeover and fast build speed on small/hollow
geometries; but requiring no postcuring. However, the surface finish is not as good as
in conventional parts there may/be weak properties in the Z axis and for large/dense
parts building speed is slow [Noorani, 2006].
Figure 2-6 Schematic view of fused deposition modelling (FDM) [Upcraft and
Fletcher, 2003]
2.2.2.3.2 Multi-jet modelling (MJM) systems
As shown in figure 2-7, the principle process of the multi-jet modelling (MJM)
system is similar to the technology used in an ink-jet printer, but applied in three
dimensions. To start the process, a small droplet of thermoplastic polymer material is
sprayed through the print head of a linear array jets or spray nozzle. To build 3D solid
parts layer-by-layer, the print head is moved in a similar fashion as a line printer. A
moveable platform is used to build the part. This lowers after each layer is developed,
where the hot droplet of material bonds to the previous layer of the part [Upcraft and
Fletcher, 2003].

Figure 2-7 Schematic view of multi jet modelling (MJM) [Upcraft and Fletcher, 2003]
2.2.3 Advantages and disadvantages of AM technology
Compared to subtractive methods, one of the critical advantages of all additive

manufacturing (AM) technologies is the ability to create parts or components with
significant unrestrained geometries. AM allows a geometric freedom not possible with
conventional methods, such as for variable wall thickness, zero draft angles and no
split line [Hague et al, 2003a; Hague et al, 2003b; Hopkinson and Dickens, 2003].
However, it is not yet possible to have total freedom, and there are still limitations of
feasibility and resolution in terms of geometry in AM systems. For instance, in SLS
systems, small features such as holes, cylinders and thin walls less than 0.5mm cannot
be built properly, because the enclosed loose powder is melted by surrounding heat.
Also, scanned features may/be insufficient in strength to resist the forces applied
during powder deposition, and wall thickness is limited by the laser beam spot size
[Kruth et al, 2005b]
There are other disadvantages too, such as the speed of production not being
comparable to conventional methods, and the surface finish and accuracy of the parts
not always being as good as in conventional technology [Dalgarno and Wright, 2003;
Castle Island‟s, 2010]. Post-processing, such as support removal and finishing
[Dalgarno and Wright, 2003; Kruth et al, 2005b; Castle Island‟s, 2010] is then
required. This can significantly increase costs and may be time-consuming, especially

when numerous parts are produced. Furthermore, due to the constraint of the building
envelope dimensions of AM machines, the size of products is limited.
Additive manufacturing produces parts or components directly from CAD-
driven data, which means that the product can theoretically be manufactured without
tooling. However, due to the restrictions of current processes and materials this is still
difficult or impossible to achieve. The possibility of the elimination of tooling could
shorten the time to market and reduce costs [Hague et al, 2003b; Hopkinson and
Dickens, 2003; Castle Island‟s, 2010]. In addition, the removal of tooling in the early
phases of the product development process saves time and money. Another possibility
is that parts and products could be produced using materials and design parameters
not feasible in the past [Hague et al, 2003b; Hopkinson and Dickens, 2003; Castle
Island‟s, 2010].
The material used is another disadvantage of each additive manufacturing (AM)
process [Kruth et al, 1998]. The availability of materials for rapid manufacturing is
limited compared to conventional manufacturing technologies. In addition, it may be
difficult or impossible to recycle complex materials.
Using additive manufacturing makes it possible to manufacture parts with
optimal material properties based on mechanical, thermal, optical and other
characteristics. Material can be selected for its variability, and then, by depositing the
material physically, its original properties can be optimised or changed. Using data
acquired from finite element analysis (FEA), additive technology offers the possibility
to use multiple materials, as well as to control the local geometric meso-structure (the
fine-scale geometric features and associated visual effects in fabrication, for instance
contour / profile / sketch) and micro-structure of a part. This means that the
functionality of a part can be optimised in ways that are not possible with other
common manufacturing methods [Kruth et al, 1998; Castle Island‟s, 2010].
In addition, using additive manufacturing technology decreases lead times, it
has the ability to produce parts on demand, saves transportation or shipping costs
when parts can be produced in or near to the place where they are needed or at the
point of use, in the quantity required [Hopkinson and Dickens, 2001; Castle Island‟s,
2010].

2.2.4 The application of AM Systems
Rapid prototyping (RP), rapid tooling (RT) and rapid manufacturing (RM) are three
categories of AM application which have been introduced in manufacturing processes
[Pham, 2001, Kruth et al, 2005b].
2.2.4.1 Rapid prototyping (RP)
Additive manufacturing (AM) was first used for developing prototypes of functional
products. Functional prototypes are usually used for the visualisation of the concept to
verify and justify design details; the verification and optimisation of the design of the
proposed part and/or assemblies in order to meet the requirements of the
form/fit/function; for design review and assessment; for marketing, promotion and
advertising; and for communication with customers. By using rapid prototyping, these
purposes can be fulfilled relatively rapidly, since no tooling is required.
2.2.4.2 Rapid tooling (RT)
In the product development cycle, tooling is a critical issue that must be considered.
Mould-making for prototypes and production is a time-consuming, high-risk and
expensive task due to requirements of geometrical complexity, high accuracy and
tolerance [Dalgarno and Stewart, 2001]. As a result, tooling preparation may be the
slowest and most costly phase in the product development cycle. The tool must be
durable, wear-resistant, and produce a quality surface finish. As reported by Pham
[2001], the introduction of rapid tooling has significantly reduced the manufacturing
time of prototypes, pre-production and, in some cases, full production tooling.
Rapid tooling is categorised into direct and indirect methods. Indirect rapid
tooling uses rapid prototype master patterns to produce a moulding before making a
tool, whereas direct rapid tooling uses the rapid prototyping machine directly to
manufacture the actual core and cavity mould inserts [Noorani, 2006]. Indirect
methods are more commonly used than direct methods, but direct tooling offers more
opportunities in time and cost savings. There are several problems with rapid tooling,
for instance, accuracy, durability, mould finishing and hardness [Pham, 2001].

2.2.4.3 Rapid manufacturing (RM)
Recently, additive manufacturing has begun to be applied to the manufacture of

structural or functional end-use parts or components directly. This technology is
continuously developing for a wide area of materials, including metals, polymers, and
ceramics for many applications, such as in medical applications and other customised
functional applications. This trend, known as rapid manufacturing (RM), is likely to
be an attractive technological approach in the future, and is considered in detail in the
next section.
2.2.5 Rapid manufacturing (RM)
2.2.5.1 Definition and basic process
Rapid manufacturing has been defined as “the manufacture of end-use products using
additive manufacturing techniques (solid imaging)” [Levy et al, 2003]. Other sources
describe rapid manufacturing as producing useable products or components directly
by using additive fabrication technology [Castle Island‟s, 2010], or “a technique for
manufacturing solid objects by the sequential delivery of energy and/or material to
specified points in space to produce that solid. Current practice is to control the
manufacturing process by computer using a mathematical model created with the aid
of a computer” [Wikipedia, 2009a]. According to Hopkinson et al [2006], rapid
manufacturing is defined as “the use of a computer-aided design (CAD)-based
automated additive manufacturing process to construct parts that are used directly as
finished products or components”.
2.2.5.2 Process chain of rapid manufacturing
Even though the basic process of rapid manufacturing (RM) is similar to that
mentioned in section 2.2.1, in this section the process is described in more detail as a
rapid manufacturing chain. Figure 2-8 illustrates the process chain of rapid
manufacturing, showing the relationships between phases and which parts or steps of
rapid manufacturing can be used to aid engineering design, in the design phase of
product realisation.

Moreover, the final step of the process chain, the post-processing phase, also
shows that there is an opportunity to heat-treat the component to improve its material
characteristics. This is one of the objectives developed in this research study.
Figure 2-8 Process chain of rapid manufacturing
The basic process chain of rapid manufacturing consists of the following steps
[Cee Kai, 2003; Venuvinod, 2004; Noorani, 2006]:
1. Created of CAD model of the design:
First of all, a CAD solid model is created. Solid modelling software is
commercially available, such as IDEAS, ProEngineer, SolidWorks,

Unigraphics, CATIA, or many others, which can be used to create the 3D

model part. In this phase, model design can be analysed and tested virtually,
for instance regarding the failure analysis of the design prototype of the
product.
2. Conversion of the design model to the stereolithography (STL) file format:
CAD software uses various algorithms to represent solid objects, so that this
conversion must be conducted since the STL format has been adopted as the
standard for rapid manufacturing technology. The STL format represents a
3D surface as an assembly of planar triangles.
3. Slicing of the STL file into two-dimensional (2D) cross-sectional layers:
In this phase, the STL file is made ready for construction. By using a pre-
processing program, the user adjusts the orientation, location and size of the
model. The pre-processing program is normally associated with a particular
type of layer manufacture machine.
4. Manufacture of the part:
The machine constructs the part layer by layer automatically. Generally, the
thickness of each layer, depending on the build technique, is in the range
between 0.01 mm to 0.5 mm [Gibson and Shi, 1997].
5. Post-processing (clean and finish the model):
Post processing is the last phase. This work involves removing the part from
the machine and detaching any supports that were applied. The component
may need surface treatment and minor cleaning, depending on the process
and application used. Furthermore, as a post-processing action, heat
treatment to modify and characterise the material can also be used to
improve its properties.
2.2.5.3 Issues in RM systems
As mentioned in the previous section, there are several reasons for the use of rapid
manufacturing. However, one of the most interesting reasons for adopting this
technology arises from its advantages in the area of design, rather than the
manufacturing approach used [Hague, 2006].
During the last few decades it has been recognised that geometry is a critical
factor in design analysis and the development of products, mainly due to cost factors.

Within rapid manufacturing (RM), geometry is not a limiting factor, and freedom of
design can be applied. It is claimed that this will reduce lead times and overall
manufacturing costs [Hague, 2006].
The range of material, used in RM systems is still limited compared to
conventional manufacturing. Since this technology has developed, the choice of
possible materials and processes has grown, but not widely enough to support all
materials-driven product design.
2.3 Selective Laser Sintering (SLS)
At the present time, SLS is a commonly used technique in rapid manufacturing for
prototyping, tooling and manufacturing purposes [Levy et al, 2003] using a wide
range of polymer, metal and ceramic materials [Kruth et al, 2003]. It was developed in
the late 1980s as a technique for rapid prototyping (RP). In this section, SLS is
described in more detail.
2.3.1 Definition and classification
Selective laser sintering (SLS) is an additive manufacturing process that it used to

fuse small particles of powdered materials (polymer, metal or ceramic) into a solid
mass of desired complex 3D shape, using thermal energy supplied by a focused high
power laser beam. Then, by scanning cross-sections generated from a CAD file of the
component, the laser selectively fuses and sinters powdered material on the surface of
a powder bed, for the first slice of the desired pattern of the object. After each cross-
section is scanned, the powder bed is lowered by one layer thickness. Then, using a
roller mechanism, a new layer of material is applied on top, and the building process
is repeated until the building part is complete [Kruth et al, 2005a; Wikipedia, 2009b].
A schematic of the SLS process is shown in figure 2-5.
From the definition and schematic process described above, the SLS process is
based on two principles, as follows [Cee Kai, 2003; Noorani, 2006]:
a. Consolidation occurs by sintering when the laser beam hits the powder
material. The interaction between the laser beam and powder material used
increases the temperature toward its softening and melting point, resulting in
particle bonding, fusing the particles and forming a solid.

b. The build of components is completed layer by layer, in which each layer of

the building process contains the cross-sections of one or many components.
Using a roller mechanism, the next layer is then deposited on top of the layers
built previously. Again, that powder layer is then sintered by laser to the
previous layer.
In SLS systems, the density of parts is affected by the packing density of the
powder. Therefore, mechanical properties are usually better when the packing density
is higher. However, this also is influenced by the machine parameters and scan pattern
used [Gibson and Shi, 1997; Cee Kai, 2003; Noorani, 2006].
2.3.2 Important processing parameters
As shown in Figure 2-9, process variables which have an effect on SLS processing
can be categorised into four types: controlled, machine specific and geometry-specific
parameters, and material properties. Each of these factors has a different effect on
sintering during the process of SLS component formation [Nelson, 1993; Agarwala et
al, 1995].
Controlled parameters
Laser scanner
Laser power Material properties
Mechanical layering of powder Viscosity
Atmospheric control Surface tension
Air flow Particle size distribution
Heater (bed temperature) Particle shape
Absorptivity/reflectivity
Machine specific parameters Thermal conductivity
Laser type Specific heat
Scan radius Emissivity
Melting temperature
Geometry-specific parameters
Scan vector length
Figure 2-9 Process variables which affect sintering in the SLS process [Nelson, 1993;
Agarwala et al, 1995]
This section describes fabrication parameters directly related to the properties of

SLS material. These are: part bed temperature, fill laser power, laser beam speed, scan

spacing and slice thickness [Gibson and Shi, 1997; Cee Kai, 2003; Venuvinod, 2004;
Noorani, 2006] as illustrated in Figure 2-10.
Laser
Beam speed Scanning path
Scan spacing
Layer
thickness
Sintered layer
Figure 2-10 Processing parameters in SLS process
a. Part bed temperature (Tb)
The part bed temperature (Tb) is the temperature at which the powder material sits in
the part cylinder. In the part cylinder, powder material is heated to the part bed
temperature before the laser scanners move, in order that laser power and distortion in
the sintering process can be reduced. In other words, the need for laser power is
reduced when the part bed temperature is higher, and the difference in temperature
between pre- and post-sintering is reduced. Part bed temperature varies depending on
the material. For an amorphous polymer, the temperature is normally set near the
glass transition temperature (Tg), and for a perfectly crystalline polymer it should be
set 3-40C below its melting temperature (Tm), due to it having no glass transition
temperature [Gibson and Shi, 1997].
b. Fill laser power (P)
The fill laser power (P) is the power used to heat the powder material during
scanning. The maximum laser power of machines is different based on the type and
manufacturer.
Fill laser power is influenced by laser beam speed, the diameter of the laser
beam, the density, specific heat, and latent melting heat of the powder [Gibson and
Shi, 1997].

In SLS technology, different kinds of lasers are used. CO2 and Nd-YAG lasers
are commonly used in SLS machines. These operate at different wavelengths of
10.6µm for CO2 lasers and 1.06µm for Nd-YAG lasers. The absorption of laser light
depends on material absorptivity, transmissivity and particle size [Nelson, 1993]. The
laser wavelength should be adapted to the powder material to be sintered, due to the
fact that laser absorption changes greatly according to the material and the laser light
wavelength [Tolochko et al, 2000]. Normally laser absorption depends on laser
wavelength, the condition of the sintered materials, surface geometry, ambient gas,
and temperature [Tolochko et al, 2000]. Powder material can usually absorb laser
energy better than solid material, due to multiple reflection and laser beam absorption
trapped in the powder pores [Kruth et al, 2003].
CO2 lasers with a wavelength of 10.6µm are suitable for polymer powders. On
the other hand, Nd-YAG lasers with 1.06µm wavelength are better for metals [Kruth
et al, 2003].
c. Laser beam speed (BS)
Laser beam speed is defined as the linear velocity of the laser beam across the powder
surface of the material [Nelson, 1993]. This machine-specific parameter can be
adjusted to suit the material being processed.
d. Scan spacing (SP)
The distance between the two nearest parallel scan vectors is called the scan spacing
(SP). Scan spacing influences laser beam size and energy density. The scan spacing
must be smaller than the diameter of the laser beam on the part bed, in order that a
cross-section of the sintered area can be built completely. Figure 2-11 illustrates scan
spacing in the SLS process [Venuvinod, 2004].

SP
2R
Figure 2-11 Scan spacing in the SLS process [Venuvinod, 2004]
e. Slice/layer thickness (h)
Slice thickness (h) is another important parameter that must be considered, because it
can cause a stair-step effect. Slice thickness is used to determine the powder material
thickness of each layer in the part. The range of slice thickness is normally set from
0.01–0.5mm, depending on the type of SLS machine [Gibson and Shi, 1997]. Slice
thickness will influence building time and surface roughness. Surface roughness can
be reduced by using smaller slice thickness, but the building time of the part will then
be longer. The penetration depth of the powder material limits the largest slice
thickness. The penetration depth of powder material is influenced by laser power,
energy density, particle size, powder density, specific heat and thermal conductivity
[Venuvinod, 2004]. Therefore, a suitable slice thickness can be set depending on the
material and the laser power available.
f. The energy density (ED)
One of the most important factors that influences the quality of laser sintering process
is energy density. This indicates how the amount of energy per uniform volume
delivered to the powders changes with process parameters.
The energy density is a function of fill laser power (P), beam speed (BS) and
scan spacing (SP). Therefore, the important relationship between processing
parameters can be defined as in equation 2.1 [Gibson and Shi, 1997]:
P
Energy density (ED) = (Joule/cm2) ............................................. (2.1)
BS * SP
where: P is fill laser power (watt)
BS is laser beam speed (cm/s)
SP is scan spacing (cm)

From the energy density equation, two parameters are most commonly altered to
gain greater energy density. These are fill laser power (P) and laser beam speed (BS)
as they are the two most influential parameters. For Scan spacing (SS), as mentioned
in section 2.3.2 (d) must be smaller than the diameter of the laser beam, or as an
alternative, it is advised that scan spacing should be 3-6 times smaller. This is in order
for each point on the powder bed to receive multiple laser exposures and thus the
sintered area can be built completely.
Alterating the settings of these parameters will affect the ability of the material
to be built successfully and the quality and properties of the parts manufactured.
For example, the density and tensile strength of an SLS fabricated part increase
as a result of increasing fill laser power, decreasing laser beam speed and decreasing
scan spacing. Beside of this, the viscosity of the material also decreases, which in turn
lowers the porosity and this then increases density [Williams and Deckard, 1998;
Caulfield et al, 2007]. On the other hand, due to increased liquid flow if the fill laser
power is too high, shear stresses between layers are created and the parts may bend or
become distorted [Beaman et al, 1997; Caulfield et al, 2007].
There is an optimum energy density that is needed to achieve the maximum
density and strength, which is different for each different material. If the optimum
energy density is exceeded, this results in the degradation of mechanical properties
and an increase in porosity and the surface degradation of the resulting laser-sintered
material [Ho et al, 1999; Caulfield et al, 2007].
Research has shown that varying the energy density can cause variations in the
density and tensile strength of polyamide sintered material [Gibson and Shi, 1997;
Tontowi and Childs, 2001]. This is supported by research conducted by Ho et al
[1999], who investigated the effect of energy density on the mechanical properties of
polycarbonate powder. Further discussion regarding the relationship between energy
density and mechanical properties is given in section 2.3.3 below.
2.3.3 Melting viscosity, surface tension and particle size
The other important parameters in the SLS process are melting viscosity, surface
tension and particle size. The relationships among these parameters for viscous
sintering was firstly established by Frenkel in 1945 [Ajoku et al, 2006b], whose
formula states that the energy generated during viscous flow is equal to the energy

gained by the reduction in the surface area. Frenkel then used an equation to explain
the neck formed in a combination of two adjacent particles, n, as shown in equation
2.2 and Figure 2-12.
3t
2
 x
   ............................................................................................. (2.2)
r 2r o
where: x = half the thickness of the neck, r = radius of a sphere,  = surface tension of
the material, t = time needed for sintering and 0 = melting viscosity.
Figure 2-12 Frenkel‟s sintering model [Ajoku et al, 2006b]
Frenkel analysed the initial stages of the sintering process in crystalline particles
and stated that, when high temperature is applied this could serve as a catalyst which
will change the crystal form of the structures. Increasing the temperature of the
particle resulted in an increase in surface contact area between adjacent particles, and
the volume of viscous flow between the particles can cause or eliminate pores. It can
be assumed that the sintering that occurs between two particles is formed completely
when x/r = 0.5 [Rosenzweig and Narkis, 1981].
From this theory, using equation 2.2, Frenkel illustrated the role of temperature
during the sintering process and confirmed that using higher temperatures cause an
increase in necking radius between sintered particles. Greater bonding between the
particles occurs and pores are reduced when the necking radius increases.
Figure 2-13 Cross-section of particle-particle bonding in x, y and z axes (Dx > Dy > Dz)
[Ajoku et al, 2006b]

In addition, Ajoku et al [2006b] investigated the bonding between particles in

each different fabrication orientation. This phenomenon was observed when they
investigated the effect of build orientation and end-of-vector (EOV) effect on the
mechanical properties of laser-sintered components. A 3D Systems Vanguard SI laser
sintering machine was used and a laser scanning strategy in the x scan vector only was
applied.
They illustrated the cross-section of particle-particle bonding in the x, y and z
axes fabrication orientations, and the findings showed that the diameter of the neck in
thex axis (Dx) is greater than the diameter of the neck in the y axis (Dy), and then (Dx)
and (Dy) are greater than the diameter of the neck in the z axis (Dz), as illustrated in
Figure 2-13.
Polymer powder melts when an infrared laser sinters the component, where the
particles bond together. The laser scans as a vector on a layer to build up the first
layer of component in the x direction and forms a neck with diameter Dx, as shown in
Figure 2-13(a). After finishing the first vector in this direction, the laser switches off
automatically and an increment occurs in the y direction that allows the laser to scan
the next vector. The bonding in the previous vector is shown in Figure 2-13(a), and
the bonding between particles in the x direction is quite similar. But particles in the
previous vector in the y direction will have cooled and, therefore the bonding between
these particles will be less than between those in the x direction, with a smaller
necking diameter Dy (Figure 2-13(b)). After this layer is completely sintered, then a
new layer of powder is applied. The particle bond in the previous layer will have
cooled to the point where the bonding between particles in different layers is less than
the bonding between particles in different vectors with an even smaller neck diameter,
Dz (Figure 2-13 (c)).
From their work, Ajoku et al [2006b] summarised that the mechanical properties
from different fabrication orientations are different (anisotropy). The highest tensile
and compression strength results from an x axis-fabricated part and the lowest results
from the z axis fabrication orientation. This agrees with the results of previous work
by Gibson and Shi [1997], in which the best tensile strength of a laser-sintered part
results from the test component built in the x axis orientation and the lowest value is
obtained from the z axis. However, for flexural strength, the highest value is obtained

from the y axis-fabricated component, followed by that in the x axis and the lowest
value for the z axis.
These findings also showed that the degrees of bonding between particles and
the subsequent bonding between layers are different in each direction of fabrication.
Bonding is also affected by the scanning strategy applied in each layer for different
fabrication orientations.
In selective laser sintering, the laser scans across the whole two-dimensional
area, which consists of parts being built. Greater time is taken for the laser to restart
scanning on a subsequent layer if the two-dimensional scan area is larger.
a. Melting viscosity (0)
The melt flow ratio or melt flow index (MFI) can be used to determine the melt
viscosity of a material. For polymer materials, the standard test of melt flow ratio can
be applied to measure its melt flow properties [ASTM D1238, 2004].
The MFI has units of gram/10min, determined from the mass of polymer
extruded during 10 minute of steady flow.
For instance, Kim and Creasy (2004) investigated the melt flow characteristics
of PA6 and PA6/clay nanocomposite, as shown in Figure 2.14. They found that the
melt flow of PA6 was 1.68 times greater than that of the PA6 nanocomposite.
Figure 2-14 Melt flow response of polyamide with and without nanoclay
reinforcement (Kim and Creasy, 2004)
For most polymers, the temperature dependence of viscosity is modeled with an

Arrhenius relationship as follows (Nelson et al., 1993):
 E 
 m  A exp   ……………………………….………………………. (2.3)
 RT 
where E is the activation energy for viscous flow and A is the frequency factor, R is
the gas constant and T is the absolute temperature (K).
The melt viscosity, m, of polymers follows an Arrhenius relationship with
temperature (T), typically decreasing by a factor of 2 for every 25 oC increase in
temperature at low shear rates (Nelson et al., 1993).
Equation 2.3 also suggests that, to consistently produce parts with uniform
quality, the size of powder particles and their distribution must be uniform to ensure a
uniformity of polymer molecule weight since melting viscosity is influenced by
average molecular weight (Mw).
The melting viscosity and molecular weight of polymer material follow the
relationship given by Shi et al [2004] as shown in equation 2.4:
0 = k(Mw)n ................................................................................................. (2.4)
where 0 is the melting viscosity, Mw is molecular weight, k is a constant, and n is

another constant value that depends on the critical molecular weight value MC. When
the molecular weight is less than MC, n = 1, if it is or greater than MC, n = 3.4.
Equation 2.4 indicates that melting viscosity varies with change in molecular
weight. So the quality of the SLS product is affected by the molecular weight via the
melting viscosity [Shi et al, 2004]. More discussion of this relationship is given in
section 2.3.4
b. Surface tension ()
“Surface tension is the driving force for sintering, where the viscous forces must be
overcome in order to allow powder sintering” [Ajoku et al, 2006b]. An interaction
between molten and non-molten particles occurs during the sintering process, where
the molten particle flow freely to attract non-molten powder particles. The powder
melts as large amounts of energy go through the surface of the powder bed. The
molten particles flow freely to minimise surface energy within the powder particles.
As necks form between particles, surface tension causes the material from the two
particles to flow together [Nelson, 1993]. The surface tension of polymer powder is
about 0.02-0.03 N/m [Shi et al, 2004].

c. Particle size
The particle size of the powder material is a factor that must be considered in
achieving the required surface smoothness and feature definition of the product. The
precision and density of SLS parts is also affected by particle size, which may be
different when different methods are used in producing the powder. For polymer
material, particle sizes between 75-100µm are suitable, where a larger size of particle
can cause a bigger step/layer thickness effect but a smaller size is difficult to spread
[Shi et al, 2004].
2.3.4 Process parameters and material characteristics
Various research studies have been conducted regarding the relationship between
process parameters and the characteristics of sintered material. Some of the most
important of these are discussed in this section.
Zarringhalam and his colleague investigated the effects of various processing
factors on the microstructure and properties of SLS Nylon 12 [Zarringhalam et al,
2006]. They analysed different specimens of SLS Nylon 12 powder and parts
produced from different thermal histories, using different machines which were
configured with contrasting parameters. Significant increases in elongation at break
were shown for refreshed and used powder compared with virgin powder on parts
made from different machines. Small variations between virgin and refreshed material
were also demonstrated in ultimate strength, with a considerable increase in tensile
strength for used powder. The values of Young‟s modulus also varied, even though
the same machine and material was used to fabricate the specimen.
Hague and Gibson assessed the influence of process parameters on the tensile
strength of thermoplastic material with different fabrication orientations [Gibson and
Shi, 1997; Hague et al, 2004]. These authors showed how the material properties from
the laser sintering process were influenced by build orientation, indicating fluctuations
in tensile strength and density of specimens produced in different orientations with
different fabrication parameters. Furthermore, Hague et al [2004] and Hou [2005]
concluded that SLS parts can be considered as anisotropic material, having achieved
different results for strength and modulus from different build orientation.
In addition, it has been recognised that, in the SLS process, both material
properties and fabrication parameters have a great influence on mechanical properties

[Gibson and Shi, 1997; Caulfield et al, 2007]. The most important characteristics that
determine the potential suitability of a material are its glass transition temperature (Tg)
and melting temperature (Tm). Characteristics involving specific heat, thermal
conductivity, particle size and density are also important.
Gibson and Shi [1997] studied the influence of fabrication parameters on tensile
strength and part density for different variations of scan size, scan spacing and fill
laser power. The results illustrated that tensile strength and part density increase with
decreasing scan size and scan spacing and increasing laser power.
Tontowi and Childs [2001] showed that the powder bed temperature (the
temperature of ambient build powder) affects the density of sintered parts which can
then have an effect on their mechanical properties. It was reported that the density of
sintered components increased when the part bed temperature increased. This
relationship is illustrated in Figure 2-15.
Figure 2-15 Powder bed temperature-density relationship of Duraform Polyamide

[Tontowi and Childs, 2001]
In their investigation, Tontowi and Childs [2001] used Nylon 12 (Duraform

Polyamide) sample material to study the effect of powder bed temperature on the
density of sintered parts. They used powder bed temperatures of 174, 178 and 182oC
and at 182oC a fully (100%) solid density resulted where 970kg/m3 resulted from a
default energy density of 0.0284J/mm2. Meanwhile, the density of SLS fabricated
parts was around 96% when the powder bed temperature was reduced to 178oC at the
same energy density. This confirms that higher bed temperature also affects the
strength of SLS fabricated parts. It also confirms that the control of bed temperature
can be used to control part density.
The effect of the properties of polymer materials on the quality of SLS parts
was investigated by Shi et al [2004] and molecular weight, molten viscosity,
crystallisation rate and the particle size of powder all had an impact. The work
initially focused on the relationship of melt viscosity and molecular weight, aiming to
increase density and improve mechanical properties. One possibility for increasing
density is to maintain a steady temperature flow inside the chamber during the SLS
building process. Moreover, the precision of SLS products is greatly influenced by the
crystallisation rate, which is itself strongly related to crystallinity. The precision and
density of SLS products are also affected by particle size, and sizes of polymer
material between 75-100µm are suitable for the SLS process, whereas larger particles
cause a bigger step effect and lower density, while smaller ones are difficult to spread.
Figure 2-16 shows the relationship between melting viscosity and molecular weight as
found by Shi et al [2004].
Figure 2-16 Relationship between the molecular weight and melting viscosity of
polystyrene [Shi et al, 2004]
Figure 2-16 shows that there is a significant increase in molecular weight with
increasing melting viscosity which confirms to Frenkel‟s model. Frenkel‟s equation
suggests that the radius of the neck between two adjacent particles is closer to the
particle size, this leads to a smaller cavity in the sintering part which means that x/r
increases as 0 decreases [Shi et al, 2004].
In research into the effects of energy density on the morphology and properties
of selective laser-sintered polycarbonate, Ho et al [1999] reported that the structure
can be built more densely when a high energy density of laser beam is applied. On the

other hand, when an excessively high energy density is applied, degradation of the
polymer will occur and decrease its density. In addition, density directly affects the
tensile strength of the specimen.
The tensile strength of sintered components can be improved by increasing laser
energy density, as shown in Figure 2-17. The maximum density of 1056 kg/m3 was
obtained at an energy density of about 0.09 J/mm2, while slight reductions in density
occurred with further increases in energy density (Figure 2-17a). The same trend was
shown in the relationship between energy density and tensile strength (Figure 2-17b).
(a)
(b)
Figure 2-17 Variation in density (a) and tensile strength (b) with energy density
[Ho et al, 1999]

The effect of bed temperature on the sintering process of polycarbonates has

been reported by Shi et al [2007], who stated that a sintered specimen warps when the
temperature is too low. When the warp is larger, the sintering cannot work normally,
due to specimen movement caused by the powder spreading roller. On the other hand,
the powder in the area around the specimen will agglomerate and become difficult to
remove when the bed temperature is too high.
It has been recognised that the processing used to fabricate components by layer
manufacture creates a structure different from normal components. This is because in
layer manufacturing, and especially SLS, the material properties depend on the
orientation of fabrication [Gibson and Shi, 1997; Caulfield et al, 2007].
2.3.5 Advantages and disadvantages of SLS
Compared to other AM technologies, SLS has several advantages as well as

disadvantages. The key advantage of the SLS process is that it can make functional
parts in actual final materials. However, the surface finish and accuracy of the parts
are not quite as good as those of other technologies, even though the material
properties can be quite close to those of the intrinsic materials.
Furthermore, the initial cost of the systems is high; they require various
peripherals and facilities, such as hard wire power, and large amounts of floor space.
Maintenance and operation costs are another disadvantage of SLS technology, since
the powder must be stored appropriately and recycled for further use.
2.4 Polymer characteristics
2.4.1 Introduction
This section describes the basic structure and classification of polymers and their
properties. Since SLS machines were introduced, polymers have been the most
popular materials for processing. This is principally because of the low energy that is
required for processing this type of material. In addition, polymeric materials have
low surface energies and high melt viscosities relative to metals and other high
surface energy inorganic crystalline materials. This eliminates the trend seen in metals
for the molten material to have lower energy due to the formation of spherical
droplets that can cause poor surface quality [Beaman, 1997].

2.4.2 Polymer classification
Polymers can be divided into two categories, namely crystalline and non-crystalline
(amorphous) structures [Noorani, 2006]. As illustrated in figure 2-18, in crystalline
structures, the atoms and molecules form in a repeated and regular form, while in non-
crystalline (amorphous) structures the atoms and molecules are arranged locally and
randomly without any overall to order or structure.
Figure 2-18 Amorphous and crystalline structure of a polymer [Noorani, 2006]
Crystalline and amorphous polymer structures respond differently to changes in

temperature, particularly in their phase change from the liquid to solid states. As
shown in figure 2-19, when a crystalline material is heated, thermal expansion occurs
uniformly until the material changes into a liquid phase when it reaches its melting
temperature (Tm), at which point the rate of increase in volume increases [Gibson and
Shi ,1997].
Figure 2-19 Variation of volume with temperature for amorphous and crystalline
polymers [Gibson and Shi ,1997]

Conversely, in an amorphous structure, thermal expansion does not change

suddenly between the solid and liquid states at the melting point. The material
expands at a constant rate until reaching the glass transition temperature point (Tg), at
which the solid begins to expand at a higher rate. Between Tg and Tm , the material is
in a transition/non-equilibrium phase [Kridli, 2006]. This condition is specified by Tg
in polymers, which depends on the cooling rate.
Glass and polymers are normally amorphous materials, even though some
plastics also have a semi-crystalline structure as an arrangement of crystalline and
amorphous structure. A plastic has a Tg above room temperature, while a rubber has a
Tg below room temperature [Baker and Mead, 2002].
Compared to crystalline materials, the density of amorphous materials is lower
[Kridli, 2006]. Crystallinity also affects the mechanical and physical properties of
materials [Kohan, 1995; Baker and Mead, 2002; Noorani, 2006].
All polymers are produced by chemical linkages between monomers (relatively
small molecules) to make polymers (very large molecules) that are packed in many
monomers as a multiple repeating molecular („mer‟) structure [Noorani, 2006 and
Kridli, 2006].
Polymer molecular chains can be subdivided into four chain structural groups:
linear, branched, cross-linked and networked, as illustrated in figure 2-20 [Griskey,
1995; Kohan, 1995; Kridli, 2006; Noorani, 2006]. The properties of polymers are
influenced by the molecular chain structure along with the type of monomer they are
made of.
A linear polymer is a long chain that is not straight; it is made up of several
backbone atoms and many mers are joined together to form a linear polymer. Some
examples of linear polymers include acrylics (polymethylmethacrylate), nylon
(polyamide), and polyethylene.
Branched polymers are similar to linear polymers; however, they have side
chains growing out of the main molecule. These branches can vary in number and
length. Each „block‟ varies in packing efficiency and the crystallinity. Compared to
linear polymers, branched polymers have lower densities but higher strength. The
ABS copolymer is an example of this category.
The third type of molecular chain is a cross-linked structure, whose chain has a
3D-like structure in which two or more chains acting as macromolecules are joined
together by side chains/branches that are bonded covalently to the bordering chain
structure. Generally with this chain structure, the strength, stiffness, brittleness and
hardness of polymers are higher. Examples of this structure are epoxies and phenolics.
The final polymer structure is the networked chain. Networked chain structures
have more covalent bonds compared to cross-linked chains and are usually more rigid
and stronger. Epoxies, acrylates, amorphous silica (glass) and phenolics are some
examples this type of chain structure.
Figure 2-20 Polymer chain structure (a) linear, (b) branched, (c) cross-linked, (d)
networked [Noorani, 2006].  represents a mer
2.4.3 Nylon 12 (Duraform Polyamide)
Among the polymers used in rapid manufacturing and selective laser sintering,
nylon is the one of the most popular. Known commercially as polyamides, there are
several types of nylon that are commercially available, such as: Nylon 6; Nylon 6,6;
Nylon 6,10; Nylon 6,12; Nylon 11; and Nylon 12. If only one monomer is involved
(such as lactam or an amino acid), the nylon is identified with only one number, such
as nylon 6 or nylon 12. If two monomers are used, the nylon will be identified using
two numbers, for example nylon 6,6 or nylon 6,12 [Baker and Mead, 2002]. Nylon 12
is a material commonly used in selective laser sintering (SLS).
Since its early development by Carothers, nylon has recently become an
important thermoplastic material [Baker and Mead, 2002]. It is a crystalline polymer
that has high modulus, strength, and impact properties a low coefficient of friction and
resistance to abrasion. In their backbones, all nylon materials contain the amide (-

CONH-) linkage chain structure, even though they have a variety of properties. Figure
2-21 shows their general structure.
Figure 2-21 Structure of nylon [Baker and Mead, 2002]
Nylon 12 has twelve carbon atoms combined with amide groups (-CONH-).
This nylon 12 molecule chain structure is shown in figure 2-22 [Kohan, 1995; Pham,
2008a]. The chain flexibility of semi-crystalline polymers is increased by the presence
of the amide group (-CONH-) [Kohan, 1995; Pham, 2008a]. Therefore, the nylon 12
backbone is relatively flexible and the long irregular linear molecule chains could
easily transform into a regular order of a folded chain structure during the
crystallisation process.
Figure 2-22 Nylon12 molecule chain structure [Kohan, 1995; Pham, 2008a]
2.4.3.1 Mechanical and physical properties
Table 2-1 shows the mechanical properties of polymers produced by the two main
manufacturers, 3D systems and EOS GmbH, including Duraform PA, while table 2-2
shows the physical properties of some polymers produced by the 3D Systems
manufacturer. These materials are available in powder form and are specifically
designed for laser sintering, tending to fuse or sinter when heat is applied. These are
predominantly polyamide-based thermoplastics, with a more recently released
thermoplastic elastomer. Therefore, the properties of these powdered materials are
similar to those of basic materials. However, the tensile test results in table 2-1 mostly

depend on the process parameters for each material and SLS machine used, and the
standard applied.
Table 2-1. Mechanical properties of polymers
Mechanical Properties
(Tensile)
Manufacturer Material Type of Material
Strength Modulus
(MPa) (MPa)
3D Systems - DuraForm PA Polyamide powder 44 1600
[2010b] (Nylon 12)
- DuraForm GF Glass-filled
polyamide powder 26 4068
- DuraForm AF Polyamide
aluminium powder 35 3960
- DuraForm Flex Elastomeric plastic
powder 1.8 7.4
- DuraForm EX 48 1517
- CastForm PS Polystyrene powder 2.84 1604
- Duraform FR100 Plastic 32 1880
EOS GmbH - PA 2200 Polyamide powder 45 1700
[2010] (Nylon 12)
- PA 3200 CF Glass-filled
polyamide powder 48 3200
- PrimeCast100 Polystyrene powder 1.2 1600
- Somos 201 Flexible TPE - 15.5

Table 2-2. Physical Properties of SLS Powder of 3D Systems [3D systems, 2006b,
2010b; Wohler, 2006]
Material DuraForm DuraForm DuraForm DuraForm CastForm
PA AF Flex GF
Type of Polyamide Polyamide Elastomeric Glass-filled Polystyrene
material powder aluminium plastic polyamide powder
powder powder powder
Density- 0.59 0.89 0.44 0.84 0.46
Tap
(g/cm3)
Average 58 Not 85 48 62
particle available
size
(µm)
Specific 0.97 Not Not available 1.4 0.86
gravity available
at 20oC
Melting 184 Not 192 185 Not
Temp. available available
(0C)
Benefit Excellent Cast Rubber-like Good heat Compatible
heat and aluminium flexibility resistance; with
chemical appearance; and increased standard
resistance; high functionality; stiffness foundry
high stiffness; create practices;
surface excellent coloured low ash
quality surface components content
finish detail by standard
and wear infiltrants
resistance
2.4.3.2 Melting, crystallisation and density
Polymer crystallisation takes place in two different stages, nucleation and growth
[Wunderlich, 1980; Young, 1983]. Nucleation describes the process of the formation
of crystal nuclei through molecular twisting at low temperatures, and this is followed
by crystal growth, normally through the addition of further lengths of chain.
The change in linear dimensions of the growing entities of crystallisation at a
given temperature is normally linear with time. This relationship can be shown as
equation 2.4 [Young, 1983]:
r = vt ........................................................................................................... (2.4)

where: r = spherulite radius, v = growth rate and t = time

Equation (2.4) is usually valid until the spherulites become so large that they
contact each other. As the linear dimensions of the lamellae changes, an equation of
similar form for solution crystallisation tends to be observed. Figure 2-23
demonstrates the relationship between crystal growth rate and crystallisation
temperature for different fractions of poly(tetramethyl-p-phenylene) siloxane,
showing that growth rate is influenced by crystallisation temperature until it reaches a
certain peak.
Figure 2-23 Relationship of crystal growth rate and crystallisation temperature Tc for
different fractions of poly(tetramethyl-p-phenylene) siloxane (molar mass given in
gmol-1) [ Young, 1983]
From this study, it was also shown that for, the polymer material, the growth
rate at the crystallisation temperature near the melting temperature is relatively low.
However, this will rise when the cooling rate is greater than before. Along with this, it
was also found that, for a given polymer, the growth rate is influenced by the
molecular mass, where the growth rate (v) increases as the molecular mass (M) is
reduced. This is possible due to two factors that affect the peak of the growth rate.
The first is that when the crystallisation temperature is lower, the thermodynamic
driving force for crystallisation will increase. However, the viscosity will increase
when the crystallisation temperature is reduced. As a result, the transfer of material to
the growth point will be troubled and, at this point, the growth rate reaches a peak and
then decreases even though the driving force continues to increase, as the
crystallisation temperature is reduced.

Furthermore, by making some assumptions about the crystallisation process, the

nucleation and growth of spherulites in a polymer liquid can be analysed, as described
by Young [1983]. Nucleation will start when a polymer melt of mass Wo is cooled
below the crystallisation temperature and when spherulites have been given a period
of time to grow. Assuming that homogeneneous nucleation occurs, then the number of
nuclei formed (per unit time per unit volume, i.e. the rate of nucleation) at a given
temperature is constant, N. As a results, the total number of nuclei formed in time
interval dt will then be N Wodt/ρL ,where ρL is the density of the liquid polymer. After
a period of time t, these nuclei will have grown into spherulites of radius r. The
volume of each spherulite will be be 4πr3/3, or using Equation 2.4, 4πv 3t3/3. If the
density of the spherulite material is ρs then the mass of each spherulite will be 4πv 3t3
ρs /3. A time t, and when grown from the nuclei formed in the interval dt, then the
total mass of spherulite material, dWs , will be given by:
dt
dWs  43 v 3t 3  s NWo ……………………………………………… (2.5)
L
Then, after a period of time t, the total massa of spherulitic material formed from all
nuclei is given by:
4v 3  s NWo t 3
t
Ws   dt ………………………………………………… (2.6)
0
3 L
after integration, equation 2.6 becomes:

Ws Nv 3  s t 4
 ……………………………………………………… (2.7)
Wo 3 L
Alternatively, the equation 2.7 can be expressed in terms of the mass of liquid, WL ,
remaining after time t, since Ws + WL = Wo
WL Nv 3  s t 4
i.e.  1 ………………………………………………… (2.8)
Wo 3 L
For the overall crystallisation process, Equation 2.8 could be modified to
account for the impingement of the spherulites, because it is only valid in the initial
stages of crystallisation. The overall volume of the system is reduced during
crystallisation due to the occurrences of the centres of spherulites moving closer to
each other. However, it can be shown that, when impingement is taken into account,
WL / Wo is related to t through an equation of the form:

WL
Wo
 
 exp  zt 4 ………………………………………………………. (2.9)
Equation 2.9 is generally known as an Avrami equation. When t is small, equations

2.8 and 2.9 have the same form. For types of nucleation and growth other than those
considered here, the Avrami equation can be expressed as:
WL
Wo
 
 exp  zt n …………………………………………………… .. (2.10)
where n is called the Avrami exponent.
According to Young [1983], the history of specimen process affects the

characteristics of polymer melting, especially rystallisation temperature. Besides that,
the heating rate at which it is heated is also influential. So the characterisation of a
single melting temperature for a polymer sample is unachievable, since melting
commonly takes place over a range of temperatures.
A function of the mass per unit volume, density is a function of the weight of
individual molecules and the way that they pack [Brydson, 1999]. Higher density can
result when the polymer crystallises more efficiently due to molecular packing.
2.4.3.3 Degree of crystallinity
The degree of crystallinity is influenced by many factors during the crystallisation

process [Anderson et al, 2003] such as thermal history. It can be calculated using the
following equation [Mateva et al, 1997]:
 H m 
XC     100% ………………………………………………... (2.11)
 H PA 12 
where Hm is an experimentally determined enthalphy of melting, and HPA-12 is the

melting enthalpy per gram of 100% crystalline material. The value of HPA-12 used for
Nylon 12 is 209.3J/g based on the measurement of Gogolewski [1980].
2.4.3.4 Effect of crystallinity on mechanical properties
Crystallinity affects the mechanical and physical properties of materials [Kohan,

1995; Baker and Mead, 2002; Noorani, 2006]. Bessel et al [1975] investigated the
effect of the crystallinity of polymerised Nylon 6 on its mechanical properties. As
shown in figure 2-24, the modulus and failure stress increased when the crystallinity

rose, while strain to failure decreased. The fracture behaviour of the Nylon 6 changes
from brittle to a ductile material with variations in percentage crystallinity. The
material was relatively brittle with an elongation at break (EaB) around 16%, when
crystallinity was 44%. However, at lower crystallinity around 37-42% the material
became ductile with an EaB from 50-80%. Moreover, at 32% crystallinity the material
became highly ductile with the EaB reaching 240%.
Zarringhalam [2007] investigated the relationship between crystallinity and
degree of particle melt on the mechanical and thermal properties of SLS Nylon 12
using different settings for building parameters. The results showed that crystallinity
and degree of particle melt have a definite relationship with mechanical and thermal
characteristics. The values of crystallinity and degree of particle melt varied
depending on the build parameters, and it was concluded that tensile strength, tensile
modulus and elongation at break fluctuated with the different parameters. The
processing and core peak height of heat flow and melt temperature differed as a result
of different settings of the build parameters.
Figure 2-24 Stress-strain behaviour as a function of crystallinity for Nylon 6

polymerised at 453 K [Bessel et al, 1975]
For the degree of particle melting, a high core peak indicates a large amount of
un-melted material, and a lower core peak represents more completely melted
material [Hopkinson et al, 2009; Zarringhalam et al, 2009].

2.5 Heat Treatment of Polymers
To alter the characteristics of a material, such as its mechanical properties, heat

treatment is an option in manufacturing process solutions. Heat treatment can be
applied not only in metallurgical processes, such as to steel, but also with polymers.
The heat treatment processes for polymers and non-polymers seem similar.
However, the main difference is the temperature used. Depending on the purpose of
the heat treatment, polymer materials are usually heat-treated at a lower temperature
than for non-polymer materials and so for the latter a furnace with higher temperature
is needed.
The heat treatment of polymer material can reduce porosity and increase density
and crystallinity, resulting in improved mechanical properties. Meanwhile for non-
polymer material such as metal it can also be used to reduce residual stress and
improve surface hardness.
Other points that distinguish heat treatment for polymer and non-polymer
materials are based on the cooling media and soaking time.
2.5.1 Basic process
Heat treatment is a technique in which the material is heated to below its

melting temperature in order to improve its properties. Heat treatment in polymer
science can be applied to improved and more stable crystal structure, normally the
aiming to improve impact strength [Wunderlich, 1976]. Figure 2-25 illustrates the
heat treatment cycle during the process.
Temp. (oC)
Time (Hours)
Heating up Holding temp.-time Cooling

(soaking)
Figure 2-25 Heat treatment cycle

The process is started by increasing the temperature of the oven or furnace at a

certain rate, then heating the workpiece up to a selected point, normally below the
melting temperature of the material. When this temperature is reached, it is held for a
certain duration as required by the design. After that, the specimen is cooled until it
reaches room temperature, inside or outside the furnace.
The heat treatment process can be conducted in an inert gas or a controlled
atmosphere. To minimise or avoid surface oxidation, the process can be performed at
lower temperatures [Kalpakjian and Schmid, 2006].
The heating temperature is selected depending on the type of material and its
melting temperature. Also the duration of soaking must be determined.
From the heat treatment cycle above, it can be concluded that three steps that
must be considered during the heat treatment process, amely:
1. Heat treatment of the workpiece to a specific range of temperatures in a
furnace
2. Soaking in which temperature and duration time of the work piece heated.
3. Cooling in air or in the furnace
As mentioned previously, the heat treatment of polymer materials has not in

standardised, and so in this study information was collected from metal heat treatment
as an initial reference point, to carry out and to suit the polymer heat treatment
process.
Along with the above factors, the effect of distortion should also be considered.
2.5.1.1 Heating temperature / time
Heating temperature can be selected depending on the type of material and its melting
temperature. Besides that, the time for soaking is determined based on the expected
result of the process. For comparison, in metal heat treatment, there are several factors
that must be taken into consideration in determining heat-up and soak times, as shown
in Figure 2-26.

Figure 2-26 Main factors influencing heat-up and soak time [Totten, 2006]
In addition, Figure 2-27 shows another way to determine the thermal soak time
for a packed load.
Figure 2-27 Thermal soak time for a packed load [Totten, 2006]
2.5.1.2 Heating media
The heating medium is another factor that must be considered when a heat treatment
process is designed. Usually, a furnace is selected based on the capacity needed. The
next choice concerns the atmosphere the heat treatment will be conducted in.
Alternative furnaces are available on the market such as those fitted with inert gas or
vacuum. Currently, in most furnaces start and end temperatures, heating rate and
holding time are set as required using digital control.
The vacuum furnace is another alternative for the heat treatment process. Heat
treatment in a vacuum has several advantages. It can prevent surface reactions such as

oxidation or decarburisation remove surface contaminants such as oxide films and

residual traces of lubricants add a substance to the surface layers of the work or
remove dissolved contaminating substances [Techpro, 2009].
2.5.1.3 Heating and cooling rates
Heating and cooling rates are also important. If the rate to slow, heat treatment will
take more time. However, when it is too fast, it will affect the homogeneity of the
structure, therefore, it is recommended to select slow heating/cooling rates.
2.5.1.4 Cooling medium
As shown in table 2-3, several cooling media are commonly used in the heat treatment
of Nylon, each having advantages and disadvantages.
Table 2-3. Advantages and disadvantages of heat treatment fluid [Kohan, 1995]
Fluid Advantages Disadvantages

Gas
General No clean up Poor heat transfer
Air Low cost Surface oxidation,
Nitrogen No oxidation discolouration
Breathing hazard
Liquid
General Heat transfer Post-anneal cleaning
Boiling water Low cost, clean Plasticisation
Mineral Oil Low cost High temperature limits
Glycerin Water soluble Low boiling point: 182oC
Silicon Oil Chemically inert Oxygen permeable
Polyglycol Water soluble Limited stability,
use antioxidant
Hot wax
General -- --
Paraffin Modest cost Wax residue
Glyco wax S932 -- Wax residue
2.5.1.5 Distortion
The variations in the manufacturing process can cause distortion which creates an
unexpected or inconsistent change in size or shape [Narazaki, M., 2006]. This must be
carefully checked when the product needs the same dimensions after heat treatment.

Due to the improvement in properties, the post-processing analysis of distortion

needs to be taken into account to ensure the consistent geometry of the end product.
2.5.2 Review of heat treatment of polymers and its impact
Many studies investigating polymeric materials have considered thermal condition

and stability, microstructure, and the impact on the microstructure on the mechanical
performance. In addition, experimental work effect of selective laser sintering (SLS)
on material properties has also been conducted.
Zarringhalam and Hopkinson (2003) reported that using heat treatment as post-
processing can improve the tensile strength and impact strength of SLS parts made
from Nylon 12. In these studies, conventional heating in a Zwick temperature
chamber was used to heat treat the specimen samples at different temperatures (155 to
183oC) and for varying times (1 to 3 hours).
It was reported that impact strength increased by 60% compared to the standard
value for the material after heat treatment at 183oC for 3 hours. Tensile strength
improved from 45.79 MPa to 52.69 MPa by 15% at these settings, whereas it
increased to 47.24 MPa (3%) at 175oC, and to 47.70 MPa (4%) at 180oC for the same
duration. In addition, the tensile modulus increased from 1796.17 MPa to 1904.47
MPa, 1863.19 MPa, and 1916.08 MPa when heat treated at 175oC, 180oC, and 183oC
respectively for 3 hours. Lower values of impact strength, tensile strength and tensile
modulus were reported when heat treated at 155oC, 165oC, and 175oC for 1 hour.
The effect of annealing on the thermal properties and crystalline structure of
Nylon 11 (polyundecaneamide) and Nylon 12 (polylaurolactam) was investigated by
Gogolewski [1979] and Gogolewski, et al [1980] respectively. It was reported that
crystal size, melting temperature and the heat of melting rose with increased annealing
temperature. For Nylon 11, the melting temperature and melting heat increased from
190oC and 10cal/g to 211oC and 20cal/g when treated at 180oC for 2000 hours,
whereas for Nylon 12 the values increased from 174oC and 10cal/g to 187oC and
23cal/g when treated for 1000 hours at 169oC.
Scobbo and Hwang [1994] studied the effect of annealing on Poly(Phenylene
Sulfide) using dynamic mechanical analysis (DMA). The annealing was performed in
a vacuum oven for a constant duration of 4 hours, but in different temperatures of 160,
180, 200 and 220oC. It was found that, as the annealing temperature increased, the

modulus also increased as a result of the increase in crystallinity. Moreover, the glass
transition temperature increased with crystallinity. It was concluded that more of the
amorphous material can gradually crystallise as the annealing temperature increases.
The overall percentage crystallinity increased slightly with annealing, and modulus
behaviour was influenced by thermal history.
Tregub, et al. [1993] investigated the effect of the degree of crystallinity on the
static mechanical properties and fatigue performance of poly(etherether ketone) PEEK
matrix composite material. A higher degree of crystallinity resulted in improved
mechanical properties, and strength and flexural modulus were considerably higher
for 35% compared to 10% crystallinity.
The effects of vacuum annealing on the physical properties of thermoplastics
were studied by Ramazani and Mousavi [2005]. Their studies investigated the vacuum
heat treatment of polystyrene at temperature between 60 to 90oC and of Nylon 6
between 160 to 190oC; and for durations between 24 to 72 hours with different
degrees of vacuum (700 to 900mbar). It was found that significant impacts on the
mechanical properties of materials occured when heat treatment was conducted at
higher levels of vacuum at high temperatures. The tensile strength of polystyrene
increased by about 10% when heat treated at 800C or less at a vacuum level of 700
mbar. When nylon 6 was treated at about 190oC (near the melting temperature of
about 220oC) for a short time in 900 mbar vacuum, tensile strength increased from
6.90 kgf/mm2 to 7.56 kgf/mm2 and the tensile modulus also increased from 190.12
kgf/mm2 to 223.88 kgf/mm2.
Heat treatment to improve the mechanical properties of the polymer
polyhydroxyamide was also performed by Chang, et al [1998]. It was reported that the
values of the strength and initial modulus of the material were improved by increased
the annealing temperature and time respectively.

Table 2-4 summarises the heat treatment studies on some relevant materials,
nylon and polymer materials.
Table 2-4. Summary of Heat Treatment Studies of Nylon and Polymer Materials
Proxi
Study Temp. Atmos- Time
Materials mity Outcome
by (0C) phere (Hours)
to Tm
1 Polystyrene 60  90 N/A vacuum 24  72 At vacuum level (700
mbar), tensile strength and
modulus increased by
about 10% when heat
treated at 80oC or less
Nylon 6 160  190 30 vacuum 24  72 At 190oC tensile strength
(220) and modulus increased at
vacuum level (900mbar)
2 Nylon 12 155  183 1 air 13 At 183oC for 3 hours,
laser (184) ultimate tensile strength
sintered and tensile modulus
increased
3 Polyhydro- 250  500 N/A vacuum 0.59 Strength and initial
xyamide modulus increased when
(PHA) annealing temperature and
annealing time increase
4 Poly(Pheny 160  220 65 vacuum 4 Modulus, crystallinity,
lene (285) glass transition
sulfide) temperature increased
5 Poly(ether 25  150 N/A air 40 Higher crystallinity, give
ether minu higher flexure modulus
ketone) tes and strength
(PEEK)
6 Nylon 11 175  180 40 air 5 At 180oC and 2000 hours,
6 (220) 2000 crystal size, melting
temperature, heat of
melting increased
Nylon 12 165  173 36 air 10 At 169oC and 1000 hours,
(209) 2500 crystal size, melting
temperature, heat of
melting increased
Note: 1. Ramazani and Mousavi [2005]

2. Zarringhalam and Hopkinson [2003]
3. Chang, et al [1998]
4. Scobo and Hwang [1994]
5. Tregub, et al [1993]
6. Gogolewski [1979] and Gogolewski, et al [1980]
N/A = Not available

2.5.3 Review of geometry impact and scanning strategy on mechanical

properties of SLS and SLM
It was noted that most researchers specialising in selective laser sintering (SLS) or
selective laser melting (SLM) investigated and focused on the effect of processing
parameters on mechanical properties. However, even though it has been studied
extensively, there is still limited information on the effect of geometry on mechanical
properties and thermal properties. In addition, studies that were conducted focused on
the geometrical impact on mechanical properties using solid “dog bone” and solid
round bar specimens with different setting parameters.
Research by Majewski and Hopkinson (2011) was carried out using solid “dog
bone” specimens to investigate the geometrical impact on mechanical properties.
They used tensile test specimens that were built in three different orientations in YX,
YZ and ZY direction (where the first letter denotes the axis parallel to the longest
dimension, and the second letter denotes the axis parallel to the second longest
dimension) with six different thickness 2, 3, 4, 5 and 6mm. The results showed that at
any build orientations, the section thickness had no significant effects to the tensile
properties or on the repeatability of the properties. Thus, there was no significant
effect on any mechanical properties on nylon-12 laser sintered materials.
In another study conducted by Hitt et al (2011), they investigated the effect of
section thickness on fracture toughness. The study involved using a number of
specimens made with different range of thicknesses. It was reported that an increase
of around 48 per cent in the energy is required to initiate crack growth per unit area
with increasing thickness for SLS nylon 12 specimens. Meanwhile, in the density
observation there was a small increase (around 5 per cent) as the sample thickness
increased from 2 to 10 mm each time.
The scanning strategy is another important factor besides process parameters to
consider as described in section 2.3.2 when fabricating components using SLS and
SLM. To optimise this function, the user or operator can select the scanning strategy
before manufacturing the laser sintered parts. This could be a “fill only” or a “fill and
outline” option. In the most recent established strategy, the laser beam does not only
scan the entire cross section but also outlines its contour.
Three scanning strategies in SLS and SLM were introduced, which includes
roster scan, across scan and outline scan applied in processing the parts [Yusoff and

Thomas, 2008]. The first one is roster scan, which involves scanner moving parallel
to the X axis. The second one is called an across scan, where the scanning strategy is
applied in perpendicular direction to each layer. The third scanning strategy is „outline
scan‟. This is where the outer perimeter of the area is outlined.
Recently, research study on different scanning strategy in SLS and SLM became
an interest due to the purpose of scanning itself in building the parts and material.
From several investigations regarding this issue, the research finding shows and
confirms that there were changes on their mechanical and other properties of laser
sintered product and materials.
The influence of the laser scan strategy on grain structure and cracking
behaviour in SLM powder-bed fabricated nickel super-alloy was investigated by
Carter et al (2014). In this study, default scan strategy (Concept Laser M2 SLM) and
the concept of a „band‟ heating effect across each „island‟ has been compared.
„Moving point‟ heating was suggested and supported by Electron Back Scattered
Diffraction (EBSD) evidence. An EBSD map from a sample formed using a simple
„back-and-forth‟ strategy was also presented and a dramatic difference in grain
structure and crystallographic orientation was found. It was confirmed that the
condition undergone by the sample strongly influences the grain structure of the final
material of specimen.
Bo, Qian et al (2012) stated in their research that laser scanning strategy can be
used to determine the relationship of deformation magnitude and the shape of the
residual stress profiles. However, in order to optimize manufacturing and reduce the
deformation of manufactured metal layers, helix scan strategy and progressive scan
strategy for selective laser melting technology was implemented, as illustrated in
Figure 2.28. They have found that helix scan strategy was more able to reduce the
deformation of melted layer compared to progressive scan strategy.

Figure 2-28 Illustration of technique and specifies of helix and progressive scanning
strategy (a) the part model, (b) the random five scan layers with progressive scan
strategy, (c) the random five scan layers with helix scan strategy [Bo, Qian et al,
2012]
The effect of laser scanning strategy and energy density on orange peel
phenomenon of selective laser sintering process (SLS) was studied by Yussoff and
Thomas (2008). In addition, Pham et al (2008) investigated the same issues that
impacted the surface finish. They used recycled Polyamide 12 (PA12) powder on the
manufacture of a Selective Laser Sintered (SLS) part.
Yusoff and Thomas (2008) found that by using recycled PA12 material, a poor
and unacceptable surface finish was obtained. Then, they developed a strategy to
eliminate the coarse surface texture known as “orange peel” by experimenting it on
the specimens and controlling the most important SLS process parameters.
Pham et. al (2008b) used experimental design optimisation (DOE) to improve
part of surface finish by observing the factors of laser scanning speed, scan spacing,
laser power, scanning strategy and part bed temperature on the specimens . Their
studies confirmed that SLS process parameters could enhance the surface quality
making it better while using the proper scanning strategy.
The experimental results suggested that by implementing an effective laser
scanning strategy and energy density value, the surface quality would be enhanced.
Proposed method by Yusoff and Thomas (2008), states that a threshold MFR value of
18 is needed for the PA12 and any values lower than that created parts of poor and
unacceptable surface quality.

Yadroitsev et al (2007) studied on parametric analysis of the selective laser

melting process. They used different power and velocity (P/V) ratio to show the
different influences of it on the properties. Their study reported that larger P/V ratio
means larger re-melted line known as the “vector”. Along with this, they also studied
whether the shifting of consecutive single vectors had any impact on the process of
forming the first layer. Different strategies were implemented to form objects with
inner structures that are less than 1mm.
Ma, Liang and Bin, Hongzan (2007) studied the effect of laser scanning pattern
on temperature, residual thermal stresses and distortion. They found that a layer
created by a fractal scanning patterned processed by a moving laser beam has reduced
amount of distortion. They concluded that in comparison with the „S‟ scanning
pattern, a more symmetrical temperature field and smaller distortion can be obtained
by using fractal scanning pattern. In addition, their results demonstrated residual
thermal stresses have insignificant effect on distortion in contrast to transient thermal
stresses. Figure 2.29 shows the illustration of scanning strategy that investigated by
Ma, Liang and Bin, Hongzan (2007).
(a) (b)
Figure 2-29. Scanning strategy (a) with “S” scan pattern and (b) with Fractal scan
pattern [Ma, Liang and Bin, Hongzan, 2007]
Kolosov et al (2006) studied homogeneity aspects in selective laser sintering

(SLS). They examined the impact of the laser scanning strategy on the quality of
sintered structure. They found that both sintering precision as well as the inner
sintering quality is influenced significantly by the hatch distance. However, this fact
has been often ignored in earlier studies.

Other related research regarding scanning strategy was also investigated by

Yang et al (2003). They observed the relation between fractal scanning path
generation and control systems for selective laser sintering (SLS). From the studies,
they found that in comparison with the linear scanning path, the samples possessed an
improvement in physical performance when using the fractal scanning.
Bugeda et al (1999) studied on numerical prediction of temperature and density
distributions in selective laser sintering processes. Their investigation was based on a
finite element model which has been developed for the 3D for simulations of single
track SLS process. The model takes into account the thermal and sintering
development in the process along with the temporary nature of the problem.
Stamp et al (2009) investigated the development of a scanning strategy for the
manufacture of porous biomaterials by selective laser melting. Their aim was to
develop a manufacturing process that is cost efficient as well as producing structures
that meet the requirements. In order to do this, they developed a „beam overlap‟
technique to create porous structures in commercially pure titanium using SLM.
2.6 Summary
The concept of additive manufacturing (AM) has been reviewed and discussed in this
chapter. AM technology has developed using an additive approach that differs from
conventional manufacturing processes. Selective laser sintering (SLS) is one of these
technologies, using powder as raw material. Currently, SLS has become an option in
the type of manufacture of end use products or components known as rapid
manufacturing. Rapid manufacturing is a rapid process concept aiming to match
market and customer demand rapidly.
However, this process produces inconsistent mechanical properties, even when
using the same materials and the same machine. The mechanical properties of
components are influenced by processing parameters, build orientation and the type of
machine used. Scanning strategy is another factor that needed to consider when
produce the part. One of the most common materials used in the SLS process is Nylon
12, commercially known as Duraform PA [3D Systems, 2006b]. Nylon 12 is a
thermoplastic polymer which has a semi-crystalline microstructure. Crystallinity is the
main factor that influences the properties of polymers, and particularly their

mechanical properties. Better mechanical properties, such as strength, modulus,

hardness, and density, can be obtained when crystallinity is higher.
The degree of crystallinity of many polymers can be increased by using heat
treatment to improve the molecular bonding in their chain structure. Nylon 12 has a
semi-crystalline structure, and so it has the potential to be heat treated.
In rapid manufacturing and the SLS process, the final phase is post-processing.
This phase is used to improve the quality of the product, for instance its surface
quality or dimensional accuracy, as well as other finishing work as required.
From the fundamental theory and previous work in this area, the following
conclusions can be drawn:
 Selective laser sintering is a technology developed for rapid manufacturing.
 Nylon 12 is one of the most popular polymers used in selective laser sintering
(SLS).
 The material properties of Nylon 12 produced by SLS are anisotropic. Its
characteristics also vary in different fabrication orientations. The mechanical
properties are inconsistent.
 Crystallinity is an important factor affecting material properties, especially
strength and modulus. In addition, scanning strategy is another alternative
factor to alter the tensile properties.
 Heat treatment can be used to increase the degree of crystallinity.
2.7 Engineering context of research work
The main purpose of this research study is to understand the mechanical properties of
Nylon SLS parts as an important step in aiding engineering design as one stage of the
product realisation process. Then, post-processing methods are developed that can be
used to improve the characteristics of the layer manufacture material. A heat treatment
process is developed with the aim to generate consistent mechanical properties in the
processing of Nylon 12 powders using selective laser sintering (SLS).
Analysis of rapid manufacturing has shown that there is a lack of data for
materials used in rapid manufacturing systems. Such data is critical in the selection
processes, particularly regarding failure analysis. Laser-sintered material produced
using SLS has no consistent mechanical properties. As a consequence, to improve and
generate more consistent mechanical properties in the process of producing objects

from Nylon 12 powder using SLS, the development of post-processing can be

considered an alternative solution.
Therefore, in this research, heat treatment was selected as a post-processing
method to improve the mechanical properties of laser sintered products. Figure 2-30
shows a flow diagram of the research into the heat treatment of nylon 12 laser sintered
products conducted in this study.
Furthermore, this study also investigates the impact of fill and outline scanning
strategy on the mechanical properties of closed hollow structures with different
section thickness and solid specimens.

Start
Nylon 12 laser sintered Level 1

component
(Heat treatment workpiece) Workpiece preparation
(Post processing of SLS)
Parameter setting Level 2

 Heating rate
 Heating temperature
Parameter setting
 Soaking time
selection
 Cooling rate
Heating furnace Level 3

 Conventional furnace
Heating furnace
 Inert gas or vacuum furnace
selection
Selection of cooling Level 4

medium
 Gas Cooling medium for
 Liquid polymer heat treatment
 Vacuum
Performance, analysis and Level 5

evaluation
 Mechanical properties
 Thermal characteristics Performance, analysis
 Microstructure and evaluation
 Distortion
Finish
Figure 2-30 Flow diagram of heat treatment of nylon 12 laser sintered product

CHAPTER 3. EXPERIMENTAL STUDIES AND

PROCEDURE
3.1 Introduction
This chapter gives details of the tests used to characterise mechanical properties and
material conditions. Section 3.2 focuses on the SLS machine and material, while the
subsequent section describes the specimen design and building for test samples. The
mechanical testing equipment, and experimental test selection and procedures are
described in sections 3.43.5. The heat treatment processes are then discussed in
section 3.6 and the statistical tools employed for data analysis are explained at the end
of this chapter.
3.2 SLS Machine and Material
A Sinterstation VanguardTM SLS machine from 3D Systems, as shown in figure 3-1

installed at the University of Leeds, was used to produce test samples in this study.
The machine has a build volume as follows: width (X) = 370 mm, depth (Y) = 320
mm, and height (Z) = 445 mm. The machine has a maximum scan speed of 10,200
mm/sec. and was equipped with a 100 Watt CO2 laser. More detailed specifications
for the machine can be found at the 3D Systems website [3D systems, 2006a].
In this research, the study is focused on the investigation of heat treatment
effect as a post processing and also the effect of different scanning strategy on their
material characteristics of laser sintered product, rather than effect of different setting
process parameters.
As described in the chapter 2 (literature review), setting process parameters of
the SLS machine would influence the characteristics of laser sintered product,
therefore to simplify and to reduce the risk of this consequence, in this research study,
the default setting parameters was used to produce the test samples. Table 3-1 shows
the default setting process parameters that were used in fabricating test specimens.
Duraform Polyamide (PA) material was chosen for the samples to be tested,
because it is commonly used in industrial products. This material is commercially
manufactured and supplied by 3D Systems Inc., and Table 3-2. shows the its typical
properties [3D systems, 2006b].
Chapter 3. Experimental Studies and Procedure Page 60

Figure 3-1 VanguardTM SLS System [3D Systems, 2006a]
Table 3-1. Process parameters used for specimen fabrication

Parameters: Setting:
1. Fill laser power 41 watt
2. Scanning speed 1778 mm/sec
3. Scan size 0.4 mm
4. Scan spacing 0.36 mm
5. Slice thickness 0.1 mm
6. Bed temperature 172oC
7. Outline laser power 6 watt
Table 3-2. Duraform PA material properties [3D Systems, 2006b]

Properties Units Test Method
Powder properties:
 Density g/cm3 ASTM D4164 0.59
 Particle size average m Laser Diffraction 58
 Specific gravity (20 0C) g/cm3 ASTM D792 0.97
 Moisture absorption (23 0C) % ASTM D570 0.41
Thermal properties:
 Melting point: Tm
0
C DSC 184
Mechanical properties:
 Tensile strength MPa ASTM D638 44
 Tensile modulus MPa ASTM D638 1600
 Tensile elongation at break % ASTM D638 9
 Flexural modulus MPa ASTM D790 1285

3.3 Specimen Design for Testing
To create the CAD model and to generate the STL file for the test specimens, the
Ideas 3D modelling software was used. The specimens were produced in a range of
different fabrication orientations.
The basic building orientations are defined as the x orientation as the laser
scanning orientation; the y orientation the plane perpendicular to scanning but in the
scan plane; and the z orientation perpendicular to the x-y plane and manufactured
along the short axis scan. Figure 3-2 shows the orientations used to fabricate the
specimens for the bend/flexural and shear tests. The specimens built in the SLS
machines were placed in 3 build orientations which are the x-axis, y-axis and z-axis
orientations, with 2 variations of each orientation. Thus specimens test were produced
in a total of 6 orientations in the XY, XZ, YX, YZ, ZX and ZY orientations as shown
in Figure 3-2. The same orientations with a single variation were used to manufacture
the specimens for tensile and compression tests were identified as X, Y and Z for each
x, y and z direction. Six samples were built of each particular test specimen.
Scan Vector
ZX
YZ
ZY
YX
Y
XZ
XY
X
Figure 3-2 Orientation of fabrication of test specimens

3.4 Mechanical Testing Equipment
A INSTRON 4505 Universal Testing Machine, with max load of 100kN (in tension
and compression), a maximum stroke of 1000 mm (approximate), and a maximum
cross head speed of 20 mm/min., was used to carry out tensile, compression, shear
and bending/flexure tests at room temperature.
The Instron Strain Gauge (25mm) extensometer was used to measure the strain
or change in length. Moreover, these changes were recorded for each incremental
change for continuous loading and computerised by using the software systems
provided in the universal testing machine that was used to do the test.
3.5 Experimental tests selection
To characterise the materials, five basic types of mechanical tests were chosen.
Tensile, compression and shear tests were carried out in order to characterise basic
mechanical behaviour. Besides these, four- and three-point bend/flexure tests were
used to investigate the structural responses of the material to more complex stress
states.
Standard tests for layer manufactured structures are not yet available. Therefore,
British and ASTM standards were reviewed, in particular those used for testing
reinforced plastic composites which have similar structures to materials produced by
SLS using layer manufacturing technology.
The standards used in defining the tests were:
1. BS EN ISO 527-1:1996 for tensile testing [BSI, 1996].
2. ASTM Designation: E 9-89a: 2000 [ASTM, 2000] for compression testing,
and consulting ASTM Designation: E111-97 [ASTM, 2003].
3. BS EN ISO 15310: 2005 for shear test [BSI, 2005].
4. The four-point bend test and the three-point bend test based on standard BS
EN ISO 14125:1998 [BSI, 1998] and consulting ASTM D6272-02 [ASTM,
2004].
3.5.1 Tensile test
The tensile tests were carried out to measure the tensile properties, include tensile
strength and tensile modulus.
As shown in figure 3-3, the specimen was a round bar, which at the gauge
length has smaller diameter. The design of this specimen was based on the tensile
specimen used on the Hounsfield Tensometer testing machine. However, BS EN ISO
527-1:1996 was used to determine the tensile properties. The detailed dimensions of
the specimen are illustrated in Figure 3-3 and to fabricate it, the technique mentioned
in section 3.3 was used. For this initial work, the specimen was manufactured for each
different orientation in the X, Y and Z directions. Meanwhile, for heat treatment
investigation, the tensile specimen was produced in the Y orientation in the y direction
only (see chapter 5 and 6).
L1 = 54.4mm
10.0mm
 10.1mm  15.3 mm
L = 50.4mm
Figure 3-3 Tensile test specimen drawing
The specimen was placed on the testing equipment using a chuck as a gripper,
as shown in figure 3-4a. The cross-head speed was set at 2mm/minute based on the
BS EN ISO 527-1:1996 standard and the specimen was loaded continuously until it
fractured. The tensile strength at break and tensile modulus were measured for all
specimens. Figure 3-4b shows examples of fractured specimens after testing.
(a) (b)
Figure 3-4 Tensile test specimens: (a) on testing machine, (b) after testing

a. Tensile strength
All stress values are defined on the basis of the initial cross-sectional area of the
specimen, and therefore the tensile strength is given by the maximum stress calculated
using equation 3.1 by letting F equal the maximum load.
F
 ........................................................................................................ (3-1)
A
where  is stress (MPa), F is load / the measured force (N), and A is the initial cross-
sectional area of the specimen (mm2 ).
Tensile strength,  M , is the maximum tensile stress sustained by the test
specimen during a tensile test (MPa).
b. Tensile modulus
Tensile modulus is determined by measuring the slope on a stress-strain diagram

within the elastic range of deformation movement:
 2  1
E ……………………………………………………………… (3-1)
 2  1
where E is tensile modulus, 2 is stress at strain ε2 and 1 is stress at strain ε1.
The strain was calculated using equation 3.2:
L
 ............................................................................................................................................................ (3-2)
L
where ε is strain value expressed as a dimensionless ratio or percentage (%), L is
gauge length (initial length) of specimen (mm),ΔL is changes in length between

gauge marks (mm). As shown in figure 3-5, the slope was taken on the straight line
portion of the stress-strain curve. The slope for the line is b/a.

40 b b
a
35
a
30
Stress (MPa)
25
20
15
10
-5
0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07
Strain
Figure 3-5 Slope measurement on stress-strain curve
3.5.2 Compression test
Compression tests were conducted to determine the compression modulus. The

compression specimen design was based on the test specimen used in the standard
ASTM Designation: E9-89a: 2000 [ASTM, 2000] for methods for the compression
testing of metallic materials. The dimensions of the specimen with a L/D
(length/diameter ratio) of 1.5 were taken, with diameter (D) 30 mm, and length (L) 45
mm. The detailed dimension of the specimen used with the test geometry is shown in
Figure 3-6. The same technique was used to fabricate the specimen as is mentioned in
section 3.3.
L D
Figure 3-6 Compression test specimen drawing
The test machine with a cross-head speed of 0.3mm/min was used in testing.
Tests progressed until the strain reached 10%. Figure 3-7 shows a compression test

specimen before testing and figure 3-8 shows it being tested. The equations used for
the compression tests were those employed for tensile tests analysis.
Figure 3-7 Compression test specimens for X, Y, and Z orientations
Figure 3-8 Compression specimen in the test equipment
3.5.3 Shear test
Shear tests were carried out to determine the shear modulus properties of the
specimens. The same technique was used to manufacture the specimens as mentioned
in section 3.3. The specimen design was based on standard BS EN ISO 15310 [BSI,
2005], as shown in Figure 3-9, on a square specimen of dimensions 150mm in length
(L), 150mm width (W), and 4mm height (H).

L L
W H
Figure 3-9 Shear test specimen drawing
In order to carry out the shear tests, a jig was designed based on the BS EN ISO
15310 standard [BSI, 2005] as shown in Figure 3-10. This jig was fixed to the
INSTRON universal testing machine. Figure 3-11 shows a test being carried out. All
the dimensions of this jig are shown in the appendix.
280
240
154
280
95
Figure 3-10 Jig fixture for shear test

Figure 3-11 Shear specimen on the test equipment
According to the BS EN ISO 15310 standard [BSI, 2005], the following

equation can be used to calculate the shear modulus:
3   L W  K
G  .............................................................................. (3-3)
4 1000 H 3
F2  F1
where: 
w2  w1
and where G is the plane shear modulus (GPa), L is the length of the specimen (mm),
W is specimen width (mm), H is its thickness (mm), K is the geometric correction
factor = 0.822; F1 and F2 are the corresponding loads (Newton); and w1 and w2 are the
deflections (mm).
Shear modulus can be calculated using the loads F1 and F2 at the deflections w1
and w2 respectively, as shown in Figure 3-12. By defining the values of w1 and w2,
then the deflection value recorded on the computer can be used to gain values of F1
and F2. The test progressed until the deflection reached 20mm, with a cross-head
speed of 1 mm/min.

Load
F2
F1
w1 w2
Deflection
Figure 3-12 Load/deflection curve
3.5.4 Four-point bend/flexure test
Four-point bend/flexure tests were carried out to measure the flexural properties of the
specimens. The specimen design was based on the EN ISO 14125:1998 standard
[BSI, 1998], and a rectangular bar (beam) of 80mm length, 10mm width and 4mm
depth was fabricated using the same technique as mentioned in section 3.3. Standard
ASTM D6272-02 (ASTM, 2004) was used as a guide for the determination of the
properties. The loading set-up for the test is shown in Figure 3-13.
P/2 P/2
d
b
L/4 Load Span L/4

Support Span, L
P/2 P/2
Figure 3-13 Four-point bend/flexure test specimen loading set-up
Based on the ASTM D6272-02 standard, the rate of crosshead motion, R, can be
determined using equation 3-4:
0.167 ZL2
R ........................................................................................... (3-4)
d

where R is the rate of cross-head motion (mm/min), L is the support span length (65
mm), d is the depth of the beam (mm), Z is the rate of straining of the outer fibres
(mm/mm), and min Z = 0.01. The rate of cross-head motion used for the test was R =
1.76 ≈ 2 mm/min. This cross-head speed is recommended in the BS EN ISO 14125:
1998 standard.
The tests were carried out after specimen were placed on the testing equipment
using a jig designed by a previous researcher [Yubero, 2004], as shown in figure 3-14.
Figure 3-14 Four-point bend/flexure test specimen being tested
a. Flexural strength
For a load span of one half of the support span as shown in Figure 3-13 and as
confirmed by the ASTM D6272-02 standard (ASTM, 2004), for relatively small
deflections at any point along the load deflection curve, the flexural stress can be
determined using equation 3-5:
3PL
S  ................................................................................................. (3-5)
4bd 2
where S is the stress (MPa), P is the load at a given point on the load deflection curve
(N), L is the support span length (mm), b is the width of the beam (mm), and d is the
depth of the beam (mm).
Then, the flexural strength of the specimen is equal to the maximum stress in
the outer fibres at the moment of break, which is calculated by letting P equal the load
at the moment of break. However, if the resultant deflection occurrs at more than 10%

of the support span length, the maximum stress from equation 3-6 is applied (ASTM,
2004):
3PL   10.91Dd 
S   1   .................................................................. (3-6)
4bd 2   L2 
where P, L, b, and d are the same as for equation 3-5, but D is the maximum
deflection of the centre of the beam (mm); and the maximum strain, r, is determined
using equation 3-7:
4.36Dd
r ............................................................................................... (3-7)
L2
The value of deflection at the centre of the beam (D) must be calculated and
was estimated through analysis of beam deflection. The value of deflection recorded
on the computer was the deflection value at the loading point which resulted from the
machine test.
The deflection at any point on the beam is based on the loading diagram (Figure
3-13) and can be determined using equation 3-8 (William, 1994):
P 3    3Pl 3 
3 3
l  3l 
EI y ( x)  x    x     x    x  l      ................ (3-8)
3
12   4  4   64 
where E is the modulus of elasticity, I is the second moment of area, P is the load, l is
the support span length, and x is the loading position at a given point.
From equation 3-8, by substituting x = l/4, the value of the deflection at the
loading point can be calculated as:
 Pl 3
yl / 4  ............................................................................................. (3-9)
96 EI
and then by substituting x = l/2, the value of the deflection at the centre can be
calculated as:
 11Pl 3
yl / 2  .......................................................................................... (3-10)
768EI
Then, dividing equation 3-10 by 3-9, the ratio for deflection at both points can be
determined. Therefore, the value of deflection at the centre was calculated as 1.375
times the value of the deflection at the loading point.

b. Flexural modulus
The flexural modulus was determined using the same procedure as for measuring
tensile modulus as shown in Figure 3-5.
3.5.5 Three-point bend/flexure test
Three-point bending tests were also carried out to determine the flexural properties of
the specimen. The same technique was used to fabricate the specimen as mentioned in
section 3.3. The specimen design based on the EN ISO 14125:1998 standard [BSI,
1998] was a rectangular beam of 80mm length, 25mm width and 5mm depth. The
same standard was used to determine the properties. The loading set-up for the test is
shown in Figure 3-15, with a support span length of 65 mm. The cross-head speed
used was 1 mm/min.
d
Load Span b
L/2 L/2
P Support Span, L P
Figure 3-15 Three-point bend/flexure test specimen loading set-up
The tests for three-point bending were carried out after the specimen was placed
on the testing equipment using a jig designed by a previous researcher [Yubero,
2004], as shown in figure 3-16.

Figure 3-16 Three point bend/flexure test specimen being tested
a. Flexural strength
As shown in Figure 3-15, the load in the three point bending test is applied at the mid
support span and so will cause maximum deflection at the mid point of the beam.
According to the BS EN ISO 14125:1998 standard [BSI, 1998], the following
procedure can be used to calculate the flexural strength and strain of the specimen.
The flexural stress f is given by the following equation [BSI, 1998]:
3PL
f  ................................................................................................. (3-11)
2bd 2
where  f is the flexural stress in megapascals (MPa), P is the load in newtons (N), L
is the span length in millimetres (mm), d is the thickness of the specimen, in

millimeters (mm), and b is the width of the specimen in millimeters (mm)
And the strain in the outer surface of the specimen can be calculated as follows:
6sd
 ...................................................................................................... (3-12)
L2
where s is the beam mid-point deflection in millimetres (mm)
b. Flexural modulus
The flexural modulus was determined using the same procedure as for measuring
tensile modulus as shown in Figure 3.5.

3.6 Differential Scanning Calorimetry (DSC)
To analyse thermal conditions and crystallinity, a Mettler Toledo Differential

Scanning Calorimetry (DSC) 823e was used, as shown in Figure 3-17.
PC control Cooler
DSC
Figure 3-17 DSC Mettler Toledo 823e
The degree of crystallinity and melting point were measured for samples
prepared by cutting a small amount of material from the tensile test specimens (after
the tests were carried out). The amount of material used for each DSC run was
~70.1mg. Each sample for measurement was positioned in a 100µl aluminium
crucible before being placed in a heat flux cell.
The measurements were made under nitrogen flow with a Mettler FP90
controller connected to a FP85 heat flux cell. A heating rate of 10 oC/minute was used
to heat the sample from 30 to 220oC, with then a hold of 2 minutes. After that, the
instrument was set up to cool at the same rate from 220 to 30 oC, and a melting and
crystallisation thermogram recorded.
To determine the degree of crystallinity and melting point, software analysis
was used with the value of HPA-12 as the melting enthalpy per gram of 100%
crystalline. A value of HPA-12 had to be inputted and the value used was 209.3J/g,
based on the measurement of Gogolewski et al [1980]. Then the value of degree of
crystallinity or melting point can be obtained.

3.7 Statistical analysis tool

Confidence intervals were used to estimate the reliability of results. A confidence
interval gives an estimated range of values which is likely to include the unknown
population average, the estimated range being calculated from a given set of sample
data.
Minitab software was used to calculate confidence intervals, using the function:
CONFIDENCE (alpha, standard_dev, size), in which alpha is the significance level
used to compute the confidence level, standard_dev. is the population standard
deviation for the data range and is assumed to be known and size is the sample size.
In this study, the confidence interval was chosen to be 95 %, and therefore
 X  M 
2
alpha is 0.05. The function of standard deviation is S  , where s is the

N 1
standard deviation, X is the value from the each test results. M is the average of all test
results from all samples and N is the sample size.
Furthermore, in order to analyse the result of the experiment, the analysis of
variance (ANOVA) is also used. ANOVA is used to identify the performance of
group of parameters under investigation. The purpose of ANOVA is to investigate the
parameter, whose combination to total variation is significant. One-way analysis of
variance (abbreviated one-way ANOVA) is a technique used to compare means of
three or more samples (using the F distribution). This technique can be used only for
numerical data.
Source of Degrees of Mean square

Sums of squares (SS) F
variation freedom (DF) (MS)
Treatments
Error
Total

In ANOVA, sum of squares or SS is calculated by where I
is the number of experimental unit, and m is the mean. The mean square of a factor
(MSj) or error (MSe) is found by dividing its sum of squares with its degree of
freedom. The three DFs are calculated from the summaries. Then the MSs are
calculated and a ratio determines F. The F-value of each design parameter is found by
dividing MS treatment with MS error. The calculated probability (p-value) says the
probability of rejection the null hypothesis in case the null hypothesis holds. In case
of , where α (0.05) is chosen significance level (0.05), the null hypothesis or Ho
is rejected. On the other hand, in case of then the result is valid and Ho is
accepted [Ostertagová and Ostertag, 2015].
3.8 Scanning Electron Microscope (SEM).

For morphological analysis, Scanning Electron Microscope (SEM), Hitachi S2400 SEM
with an Oxford Instruments Isis 200 ultra-thin window X-ray detector was used in this
study.

CHAPTER 4. MECHANICAL PROPERTY

CHARACTERISATION
4.1 Introduction
The purpose of this chapter is to present the results of the tests of mechanical properties
on Selective Laser Sintered Nylon 12 (Duraform PA) carried out to provide a
benchmark of mechanical performance. The specimens were produced based on those
described in section 3.3, in which different fabrication orientations were defined. All
specimens were produced using the commercial Sinterstation Vanguard SLS machine at
Leeds University, as described in section 3.2.
To characterise material performance, the responses of the material to load in
different conditions need to be on determined, such as in tensile, bending and shear
testing. Tensile testing is used to measure the resistance of a material to tensile load, and
flexural testing is used to measure the durability in terms of wear, pressure, or damage
of a material to withstand bending load, while shear testing is used to measure shear
resistance of a material. All these tests are important in knowing how the material will
behave in service conditions.
The effects of different fabrication orientations on the mechanical properties of
specimens were characterised, as described in the next section. For each test result
obtained, an analysis on variance (ANOVA) was carried out to investigate the
significance of variations in the mechanical characteristics of specimens according to
orientation.
For the initial work described in this chapter, all specimens were tested as non-
heat-treated samples.
4.2 Tensile test
Tensile tests were carried out for three different fabrication orientations, X, Y and Z.
For each orientation, six samples were tested. The values of stress and strain were
calculated using equations 3.1 and 3.2. The stress-strain graphs were then plotted as
shown in Figure 4-1.
Chapter 4. Mechanical Property Characterisation Page 78

Orientation - X Orientation - Y
50 50
40 40
Stress (MPa)
Stress (MPa)
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0.12 0 0.02 0.04 0.06 0.08 0.1 0.12
Strain Strain
Orientation - Z
50
40
Stress (MPa)
30
20
10
0
0 0.02 0.04 0.06 0.08 0.1 0.12
Strain
Figure 4-1 Stress-strain graphs for each of the tensile tests for the different part
orientations.
From these stress-strain graphs, tensile strength was determined from the
maximum stress value. The results for specimens fabricated in different orientations are
summarised in Table 4-1. The table also show the average values and confidence
intervals for each orientation of the specimens. Sample number one in the Z orientation
failed when it was tested, and therefore no value is given.
Table 4-1. Tensile strength results (MPa)
Sample Average 95 % Confidence
Orientation
1 2 3 4 5 6 (MPa) interval
X 28.98 32.75 32.00 30.55 26.45 31.90 30.44 ±1.89
Y 44.88 42.11 42.44 43.70 43.60 37.99 42.45 ±1.92
Z - 34.63 36.16 35.02 34.56 32.25 34.52 ±1.25
The confidence intervals in table 4-1 show that there were small variations in
tensile strength within the specimens fabricated in the same orientations. For specimens
fabricated in different orientations (Figure 4-2) there are significant differences in
tensile strength. The ANOVA results also indicate that the differences in average
strength for different orientations were significant, where the p value is 0.00 smaller
than 0.05.

The X orientation has the lowest mean tensile strength, while Z is slightly higher.
However, the Y orientation gives the highest values of mean tensile strength, which is
consistent with other research findings (Gibson and Shi, 1997; Hague et al, 2004).
50
Tensile Strength (MPa) 40
30
20
10
0
X Y Z
Orientation
Figure 4-2 Tensile strength of different fabrication orientations (distribution for all
specimens)
The tensile modulus for each test was calculated from the initial slope of the
straight line on the stress-strain curves shown in Figure 4-1. The slope was measured at
strain value by constructing a linear fit line on selected data, as shown in Figure 4-3
which illustrates the plot for the Y orientation. The tensile modulus for each orientation
was found from the linear fit equation, and the results are summarised in Table 4-2.
Orientation - Y
10
y1 = 989.66x - 1.1986
8 y2 = 1045x - 1.244
Stress (MPa)
6 y3 = 1009x - 1.1376
y4 = 987x - 0.997
4
y5 = 1055.5x - 1.4435
2 y6 = 1027.4x - 0.488
0
0 0.002 0.004 0.006 0.008 0.01
Strain
Figure 4-3 Linear fit line on stress-strain relationship of specimen in tensile test for the
Y oriented parts
Table 4-2. Tensile modulus results (MPa)
Sample 95 %
Average Confi-
Orientation
1 2 3 4 5 6 (MPa) dence
interval
X 853.20 876.15 870.21 866.18 765.09 938.84 861.61 ±44.78
Y 989.66 1045.00 1009.00 987.00 1055.50 1027.40 1018.93 ±22.81
Z - 964.06 934.17 921.35 912.02 841.10 914.54 ±39.89

It can be seen from the confidence intervals in table 4-2 that the tensile modulus
also varies within each orientation. The results shown in Figure 4-4 give the highest
mean tensile modulus value in the Y orientation (1018.93 MPa), while the lowest is in
the X orientation (861.61 MPa) which represents 15.4% of variation in the different
build orientations. From the ANOVA results, it is also confirmed that the p-value is 0.00
< 0.05, meaning that tensile modulus is influenced by build orientation.
1200
Tensile Modulus (MPa)
1000
800
600
400
X Y Z
Orientation
Figure 4-4 Tensile modulus of different fabrication orientations (distribution for all
specimens)
Tensile elongation at break was determined by calculating the strain value at the
point where the break was found, then multiplying it by 100 (%), as shown in Figure 4-1
for each different fabrication orientations, and the results are summarised in table 4-3.
Table 4-3. Tensile elongation at break (%)

Sample Avera 95 % Confidence
Orientation
1 2 3 4 5 6 ge (%) Interval
X 6.80 8.30 6.20 7.60 7.50 8.00 7.40 ±0.62
Y 11.50 8.60 10.70 11.20 10.30 6.50 9.80 ±1.53
Z - 6.10 6.90 6.20 6.10 5.50 6.16 ±0.44
From table 4-3, the results show that the tensile elongation for the Y orientation
has the highest of average value followed by those X and for Z. This result is slightly
different from the results for average tensile strength, in which the Z orientation had a
slightly higher value than that for the X orientation. However, the tensile elongation for
the Y orientation was still the highest for both average tensile elongation and average
tensile strength. As shown in Figure 4-5, for specimens fabricated with different
orientations, there are significant differences in tensile elongation at break. This

observation is confirmed by the statistical analysis (ANOVA) where the p-value is less
than 0.05.
Elongation at break (%)

12
10
8
6
4
2
0
X Y Z
Orientation
Figure 4-5 Tensile elongation at break of different fabrication orientations (distribution

for all specimens)
As shown in Figure 4-1 and Figure 4-5, the variations in the results of elongation
at break (strain) almost 100% in the Y-orientation and only 15% in the Z-orientation.
Meanwhile the variations in the results of tensile strength (Table 4.1 and Figure 4-2) and
tensile modulus (Table 4.2 and Figure 4-4) are not bigger than 18% and 7% for Y
direction; 12% and 14.6% for Z direction, respectively.
When the direction of loading during tensile testing is perpendicular to the scan
direction, the average elongation at break was found to be greater. Loading for specimen
in Y and Z are perpendicular to the direction of scan, while for X is it parallel to the
direction of scan. The scanning laser in the perpendicular specimen is only traveling at
the same width of the specimen. Thus this could cause faster fusing of scan lines. In
comparison, for parallel specimen the laser must travel the full length of the specimen
before scanning the next line. Thus delaying the fusion time. As a result, the difference
in time between scan lines causes different particle-to-particle bonding during sintering
and therefore causes a variation in material properties depending on the direction of
loading during tensile testing [Nelson J.A., et al, 2014].
Besides that, the differences of the variations might be due to that all areas of the
powder bed were used to position specimens with different quality of powder used
whilst they were being built.

4.3 Compression test
Compression tests were carried out on the three different fabrication orientations X, Y
and Z. The properties of six samples were evaluated for each orientation of the
specimens. The values of stress and strain were calculated using similar equations to
those used for tensile stress and strain. The stress–strain graphs were plotted as shown in
Figure 4-6 for the X, Y and Z orientations.
Orientation - X Orientation - Y
70 70
60 60
Stress (MPa)
Stress (MPa)
50 50
40 40
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0 0.02 0.04 0.06 0.08 0.1
Strain Strain
Orientation - Z
70
60
Stress (MPa)
50
40
30
20
10
0
0 0.02 0.04 0.06 0.08 0.1
Strain
Figure 4-6 Stress-strain graphs for each of the compression tests for the different
orientation of parts
Using the same procedure as applied for tensile properties, the compression
modulus of each orientation was determined by measuring the slope of the straight line
on the stress-strain curve. As mentioned in section 3.5.2, the compression tests were
progressed until the strain reached 10%. The slope was measured at strain by
constructing a linear fit line on selected data, as shown in Figure 4-7 which illustrates
results for the Y orientation. The compression modulus for specimens in each
orientation (X, Y and Z) was found from the linear fit equation, and the results are
summarised in Table 4-4.

Orientation - Y
y1 = 1402.6x - 2.608
12
y2 = 1192x - 3.0973
10 y3 = 1487.4x - 2.4135
Stress (MPa)
8 y4 = 1352.7x - 3.0283
6
4
y5 = 1382x - 2.8811
2
y6 = 1447.2x - 2.0092
0
0 0.002 0.004 0.006 0.008 0.01
Strain
Figure 4-7 Linear fit line on stress-strain relationship of specimen in compression test
for the Y-oriented parts
Table 4-4 shows the compression modulus from the specimens produced with
different orientations.
Table 4-4. Compression modulus results (MPa)

Sample 95 %
Orienta Average Confi-
-tion 1 2 3 4 5 6 (MPa) dence
interval
X 1036.10 1162.90 857.16 1310.10 1369.10 1226.80 1160.36 ±150.88
Y 1402.60 1192.00 1487.40 1352.70 1382.00 1447.20 1377.32 ±82.08
Z 882.44 1115.10 1298.50 1221.70 1243.00 1191.00 1158.62 ±118.62
As shown by the confidence intervals in Table 4-4 and Figure 4-8, the average
value of compression moduli in different fabrication orientations vary. The Y
orientation still gives the highest average value (1377.32 MPa) and the X and Z
orientation values (1160.36 MPa and 1158.62 MPa) are similar. From the ANOVA
results, it can be seen that the p-value is 0.036 which is less than 0.05, meaning that
compression modulus also differs with the orientation of the specimen.
Compression Modulus (MPa)
1600
1400
1200
1000
800
600
400
X Y Z
Orientation
Figure 4-8 Compression modulus of different fabrication orientations (distribution for

all specimens)
4.4 Shear test
Shear tests were carried out on six samples each of the six different fabrication
orientations to evaluate the material properties in each individual orientation. Figure 4-9
shows load-deflection plot for the YZ and ZY orientations. Results for other orientations
(XY, XZ, YX, ZX), can be found in appendix A.
Orientation - YZ
Orientation - ZY
100 100
80 80
Load (N)
Load (N)
60 60
40 40
20 20
0 0
0 5 10 15 20 25 0 5 10 15 20 25
Deflection (mm) Deflection (mm)
Figure 4-9 Load-deflection graphs for each of the shear tests for the YZ and ZY oriented
parts
The shear modulus of a specimen was calculated by using equation 3-3 as

described in section 3.5.3, in which loading F1 at deflection w1 and loading F2 at
deflection w2 were used for the calculation. Here, by defining w1=2mm and w2=6mm,
and then by interpolating the load from the data recorded, F1 and F2 can be determined.
The data was recorded until the deflection reached 20 mm. Table 4-5 shows the shear
modulus the specimens produced in different orientations.
Table 4-5. Shear Modulus Results (GPa)

Sample 95 %
Average
Orientation Confidence
1 2 3 4 5 6 (GPa)
interval
XY 0.59 0.38 0.46 0.35 0.31 0.44 0.42 ±0.08
YX 0.62 0.60 0.54 0.66 0.54 0.55 0.58 ±0.04
ZX 0.62 0.58 0.56 0.57 0.59 0.59 0.59 ±0.02
XZ 0.56 0.54 0.56 0.50 0.49 0.52 0.53 ±0.02
YZ 0.51 0.53 0.56 0.58 0.57 0.51 0.54 ±0.02
ZY 0.48 0.47 0.42 0.45 0.48 0.48 0.46 ±0.02
Table 4-5 and Figure 4-10 shows that the highest average shear modulus values
are achieved in the YX and ZX orientation while the lowest values are in XY and ZY.
The highest and lowest average values differ by 28.8%. This shows that the shear
modulus is also influenced by build orientation, as confirmed by the p-value from the

ANOVA result which is less than 0.05. Figure 4-10 illustrates the distribution of shear
modulus value for all specimens from different fabrication orientations.
0.80
Shear Modulus (GPa)

0.60
0.40
0.20
0.00
XY YX ZX XZ YZ ZY
Orientation
Figure 4-10 Shear modulus of different fabrication orientations (distribution for all
specimens)
4.5 Three-point bending/flexure test
Three-points bending tests were carried out on the six different fabrication orientations.
To evaluate the properties for each individual orientation, six samples were tested.
Figure 4-11 shows the full test results of the three point bending/flexure test for the YX
and ZX orientations. For other results in the XY, XZ, YZ, ZY orientation, see in the
appendix A.
Orientation - YX
70 Orientation - ZX
60 70
60
Stress (MPa)
50
Stress (MPa)
40 50
40
30
30
20
20
10
10
0 0
0 0.04 0.08 0.12 0.16 0 0.04 0.08 0.12 0.16
Strain Strain
Figure 4-11 Stress-strain graphs for each of the three point bending tests for the YX and
ZX oriented parts
To determine the flexural modulus properties of the specimen, the three-point

bending test was carried out using the procedure as described in section 3.5.5.
Using a similar procedure as that to determined tensile modulus as described in
section 3.5.1, the flexural modulus was determined. The slope was measured at a strain

value by constructing a linear fit line on selected data, as shown in Figure 4-12 for the
ZX orientation. Results for all specimens are summarised in Table 4-6.
Orientation - ZX
14 y1 = 1450x + 0.1631
12 y2 = 1552.2x - 0.4861
y3 = 1530.2x - 0.1758
Stress (MPa)
10
y4 = 1450.8x + 0.0063
8
6
4 y5 = 1453.7x - 0.2647
2 y6 = 1420x + 0.136
0
0 0.002 0.004 0.006 0.008 0.01
Strain
Figure 4-12 Linear fit line on stress-strain relationship of specimen in three point
bending/flexure test for the ZX-oriented parts
Table 4-6. Three point flexural modulus results (MPa)
Sample 95 %
Average Confi-
Orientation
1 2 3 4 5 6 (MPa) dence
Interval
XY 1428.60 1150.20 1179.40 1336.20 1033.00 847.50 1162.48 ±166.85
YX 1224.20 1401.00 1444.10 1336.10 1417.30 1005.20 1304.65 ±133.20
ZX 1450.00 1552.20 1530.20 1450.80 1453.70 1420.00 1476.15 ±41.87
XZ 989.08 1107.20 998.92 866.58 1059.00 1153.10 1028.98 ±81.04
YZ - 1562.50 1597.00 1551.60 1547.70 1537.80 1559.32 ±20.02
ZY 1487.70 1518.60 1452.60 1412.50 1428.90 1433.70 1455.67 ±32.12
Figure 4-13 shows a graphical representation of the flexural modulus results.
1600
Flexural Modulus (MPa)
1400
1200
1000
800
600
400
200
XY YX ZX XZ YZ ZY
Orientation
Figure 4-13 Three point flexural modulus of different fabrication orientations

(distribution for all specimens)
As shown in Table 4-6 and Figure 4-13, the flexural modulus of the different
fabrication orientations also varies. From the average value, the YZ orientation has the
highest average modulus, followed by ZX, ZY and then YX. From the ANOVA results,

it can be confirmed that the p-value is 0.000, which is less than 0.05 and means that
modulus varies with different orientations.
The high YZ orientation average value is thought to be caused by the scan factor
in this orientation being short. According to Gibson and Shi [1997], the average
sintering temperature is uniform when the scan vector is short. This condition can
provide better mechanical properties in the sintered part. The XZ orientation then has
the lowest mean modulus (1028.98MPa) due to the long scan vector.
4.6 Four-point bending/flexure test
Four-point bending tests were carried out on the six different fabrication orientations.
For each individual orientation, six samples were tested to evaluate the material
properties. Figure 4-14 shows the full test results of the four point bending/flexure tests
for the XY and YZ orientations. Other results can be seen in appendix A.
Orientation - XY
Orientation - YZ
60 60
50 50
Stress (MPa)
Stress (MPa)
40 40
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0 0.02 0.04 0.06 0.08
Strain Strain
Figure 4-14 Stress-strain graph for each of the four point bending tested for the XY and
YZ oriented parts
The flexural modulus from the four-point bending test was determined using the
same procedure as described in section 3.5.4. During the test the specimens did not
fracture, except for one YZ specimen. This happened due to a slip of the specimen
between the spans, resulting in the specimen not breaking and generating an error in the
results.
The slope was measured at a strain value by constructing a linear fit line on
selected data, as shown in Figure 4-15 for the XY orientation. The results for all
specimens are summarised in Table 4-7.

Orientation - XY
16 y1 = 1493.1x + 0.8091
14 y2 = 1672x + 0.3151
12
Stress (MPa)
y3= 1566.8x + 0.2384
10 y4= 1706.3x + 0.5099
8
6
4 y5= 1510.2x + 0.0674
2 y6 = 1662.8x - 0.1643
0
0 0.002 0.004 0.006 0.008 0.01
Strain
Figure 4-15 Linear fit line on stress-strain relationship of specimen in four point
bending/flexure test for the XY-oriented parts
Table 4-7. Four point flexural modulus results (MPa)

Sample 95 %
Average
Orientation Confiden-
1 2 3 4 5 6 (MPa)
ce Interval
XY 1493.10 1672.00 1566.80 1706.30 1510.20 1662.80 1601.87 ±72.46
YX 1466.40 1437.30 1322.40 1480.40 1475.40 1293.50 1412.57 ±66.34
ZX 1573.90 1538.60 1665.20 1656.30 1510.60 1659.70 1600.72 ±54.77
XZ 1326.90 1202.80 1276.60 1326.80 1163.40 1289.90 1264.40 ±53.79
YZ 1460.90 1408.10 1418.00 1460.10 1522.60 1508.00 1462.95 ±36.89
ZY 1535.90 1748.90 1668.30 1619.30 1689.60 1709.90 1661.98 ±60.27
The results for flexural modulus of the different fabrication orientations are shown
graphically in Figure 4-16.
1800
1600
1400
1200
1000
800
XY YX ZX XZ YZ ZY
Orientation
Figure 4-16 Four point flexural modulus of different fabrication orientations

(distribution for all specimens)
From Table 4.7 and Figure 4-16, it can be seen that the highest average flexural
modulus value is in the ZY orientation, followed by XY and ZX. This also shows that

flexural modulus varies with the orientation of specimens. This observation is supported
by the results from the ANOVA where the p-value is less than 0.05.
A part with a position in the shorter laser scanning direction will produce a high
flexural modulus, whereas a part in the longer laser scanning direction will have a low
flexural modulus. This is shown in Figure 4-16. The laser scanning direction was
described in Figure 3-2 and section 3.3.
4.7 Summary, Benchmarking and Discussion
4.7.1 Summary of test results
The mechanical properties of Nylon 12 laser sintered components has been

characterised, and the results for all modulus values are shown in Figure 4-17.
Flexural-3P (MPa)
Flexural-4P (MPa)
Shear (MPa)
Compression (MPa)
Tensile (MPa)
0 500 1000 1500 2000

X Y Z XY YX ZX XZ YZ ZY
Figure 4-17 Summary of result for tensile, compression, shear and flexural modulus
(three- and four-point) of nylon 12 laser sintered material in the X, Y, Z, XY, YX, ZX,
XZ, YZ and ZY orientations
Tensile strength, tensile modulus, tensile elongation at break and compression

modulus were highest in the Y orientation.
In the other tests conducted, the modulus values in different orientations showed
some variation. However, the values of the shear modulus were always smaller than
other modulus values, even for different orientations.

4.7.2 Comparison of present results with those of previous research
The values of modulus and strength resulting from the tests were compared with the
values found by other researchers [Hague et al, 2004; Yubero, 2004; Hou, 2005] and
reported by 3D Systems [2006b], as shown in Tables 4-8 to 4-11, and Figures 4-18 to 4-
22. Even though other researchers used different names for each of their specimen
orientations, the general directions of how the specimens are placed are similar.
Table 4-8. Comparison of results of tensile and compression modulus of Duraform PA

(SLS) material
Tensile Modulus (MPa) Compression Modulus (Mpa)
Data from
X Y Z X Y Z
This research 861.61 1018.93 914.54 1160.36 1377.32 1158.62
Hou [2005] 1582.89 1547.34 1507.52 1572.01 1574.18 1328.94
3D Systems
1600 -
[2006b]
Hague et al Flat Upright - -
[2004] 2047.0 1817
Horizontal Vertical Horizontal Vertical
Yubero [2004]
1575.46 1316.29 1518.55 1405.36
2100
1800
1500
1200
900
600
300
0
Y
s
3D p t
st t
Z
iz Z
Ve tal
l
X
ca
Sy righ
a
em
Fl
on
rti
U
or
H
Orientation
Kamil Hou Yubero R.Hague 3D System
Figure 4-18 Comparison of results of tensile modulus test
Figures 4-18 and 4-19 shows that the test results for tension and compression in
this study were low compared to those reported by other authors. However, from the
results as a whole, it is clear that there are significant variations between researchers.

1800
Compression Modulus
1600
1400
1200
(MPa)
1000
800
600
400
200
0
X Y Z X Y Z l al
ta
on rtic
o riz Ve
Orientation H
Kamil Hou Yubero
Figure 4-19 Comparison of results of compression modulus tests
It can be seen that the value of tensile modulus does, in general, vary with
orientation, as does the compression modulus. Both sets of results illustrate that the
highest values are in the Y orientation, while the values in the X or Z directions differ
only slightly.
In addition, table 4-9 illustrates the flexural modulus results from the three-point
bend/flexure tests found in this research and by other researchers.
Table 4-9. Comparison of results of flexural modulus of Duraform PA (SLS) Material

(Three-point test)
Flexural Modulus (3 Point) (MPa)
Data from
XY YX ZY XZ YZ ZX
This research 1162.48 1304.65 1476.15 1028.98 1559.32 1455.67
X Y Z - - -
Hou [2005]
1515.9 1747.0 1386.92 - - -
3D Systems
1285
[2006b]
It should be noted that 3D Systems conducted the test based on the ASTM D790
(three-point bend/flexure test) standard, while this research (and also Hou’s) used BS
EN ISO 14125:1998. Moreover, table 4-10 shows the results of the four-point
bend/flexure test with different fabrication orientations.

Table 4-10. Comparison of results of flexural modulus of Duraform PA (SLS) material

(Four-point test)
Flexural Modulus (4 Point) (MPa)
Data from
XY YX ZY XZ YZ ZX
This research 1601.87 1412.57 1600.72 1264.40 1462.95 1661.98
X Y Z - - -
Hou [2005]
1727.77 2052.13 1398.04 - - -
Yubero Horizontal Vertical
[2004] 1443.91 1347.09
1800
1600
1400
1200
1000
800
600
400
YZ
s
ZY
Y
XY
YX
Z
XZ
ZX
em
st
Sy
Orientation
3D
Kamil Hou 3D Systems
Figure 4-20 Comparison of results of test of flexural modulus (three-point)

2200
2000
1800
1600
1400
1200
1000
800
600
XY YX ZX XZ YZ ZY X Y Z al al
o nt rtic
iz Ve
Orientation or
H
Kamil Hou Yubero
Figure 4-21 Comparison of results of test of flexural modulus (four-point)
From tables 4-9 and 4-10 and Figures 4-20 and 4-21, it is also clear that the
values of flexural modulus for different fabrication orientations are different. The
highest value from the three point test is in the YZ direction, while the ZY orientation
has the highest value in the 4-point test. The lowest values from both sets of result are in
the XZ direction.
Furthermore, table 4-11 and Figure 4-22, compare the tensile strengths found in
this research with findings by other researchers.

Table 4-11. Comparison of tensile strength results

Tensile Strength (MPa)
Data from
X Y Z
This Research 30.44 42.45 34.52
Hou [2005] 41.93 42.36 37.35
3D Systems
44
[2006b]
R.Hague [2004] 48.7 - 40.9
Horizontal Vertical
Yubero [2004]
44 34.4
In this research, the highest tensile strength is in the Y direction, while the values
of X and Z are similar, as was noted for tensile modulus.
50
40
30
20
10
0
X Y Z X Y Z al al at ht s
o nt rtic Fl rig tem
p
or
iz Ve U Sys
H
3D
Orientation
Kamil Hou Yubero R. Hague 3D Systems
Figure 4-22 Comparison of tensile strength results
It can be concluded that the variations in results between researchers are greater
than the variations between fabrication orientations in the X, Y or Z directions. It is
clear that there are significant variations between researchers, which means that the
results are inconsistent even using the same materials. In addition, there is a significant
factor to account for in the general repeatability of the results, such as sample
preparation and test conditions.
The test results as a whole also show that both strength and modulus vary between
different fabrication orientation. Therefore the material is anisotropic, and this is a
second significant factor that needs to be taken into account.
4.7.3 Discussion
The mechanical property test results have demonstrated that the part built in the Y
orientation showed highest modulus values, while the parts produced in the X and Z

orientations showed lower values. This is thought to be the result of the different
densities of the build material. In the Y direction, a short scan vector causes higher
density, which gives improved mechanical properties. Generally, for laser fabricated
material a single spot laser is focused to consolidate the powder grains. However, when
the scan vector is short; this single laser spot that is moving back and forth very rapidly
with a high level of overlap between each scan line could potentially heat an entire
‘band’ or ‘length of motion’ of the materials instantaneously. The cooling rate and
direction of heat flow caused the adjacent materials to have less difference in thermal
gradient during solidification and generate a more homogenous material [Carter et al,
2014]. Therefore, shorter scan vectors can offer uniform temperatures in sintering,
resulting in good mechanical properties. This is as reported by Gibson and Shi [1997].
The same trend was also observed for the shear modulus. Shorter laser scanning
in YX yielded the highest modulus, while longer scanning in the XY direction produced
a lower modulus. However, for the part produced in the YZ orientation the modulus
values were higher than those in other orientations in the three-point bending test and
for the four-point bending test. This is according to the direction of loading applied to
each specimen, which also has an effect on the mechanical properties.
The results also indicated that there exist some variations in the modulus between
other researchers. Since different standard is used by different researchers, slight
changes in term of layer thickness, height/thickness of the samples geometry, the state
of the material used, and the samples position within the build envelope during build
will have impact on the mechanical properties [Rusenberg and Schmid, 2014;
Rusenberg et al, 2014]. As materials used (polyamide Nylon-12) during SLS process is
susceptible to material ageing which is dependent on, job height and position of the
samples [Rusenberg and Schmid, 2014; Rusenberg et al, 2014].
In addition, the results obtained indicated that the variations in the different tests
in this analysis could be due to the different geometries of the specimen samples.
From the results presented in this chapter and also from the literature review, it is
clear that the mechanical properties of laser-sintered material vary and are not consistent
across machines or fabrication orientation. Besides that, variations also occur due to the
process parameters used for powder material.
For crystalline materials, such as Nylon 12, they are often brittle. This property
makes it difficult to get yield point, but easy to get the ultimate tensile strength (UTS).

Yield stress is mostly used in design application, mainly for ductile material. Yield
strength is an important factor to consider during design to ensure safety factor is
maintained. Meanwhile UTS is used for quality control of products that are
manufactured, as it easy to carry out the test.
Therefore, in this study, the ultimate tensile strength is being used as a criterion
for comparing the performance of tensile specimens and not the elastic modulus or the
yield stress.

CHAPTER 5. HEAT TREATMENT STUDIES

5.1 Introduction
This chapter is to present details of the heat treatment regimes applied to selective
laser sintered nylon 12 and their impact on the material’s mechanical properties.
Section 5.2 describes heat treatment selection factors and related matters. Following
sections are the descriptions of samples preparation and characterisations. The
furnaces and equipment used in this work, procedures and results are reported in
section 5.5. At the end of the chapter, a discussion is presented. All samples used in
this chapter were solid structure samples.
As initial work, in chapter 4, characterisation was carried out to provide a
benchmark of mechanical performance. It is to show the responses of the material to
load in different condition, such as the resistance of a material to tensile load, bending
load and shear load.
In practice, humidity and temperature, the effect of type of loads, strain and
creep play a big role in the properties, characteristics and behaviour of a plastic
material and the moulded product performance. However, since it is expensive to test
all the available properties of a plastic material, the test are carried out to ensure that
the properties meet the minimum requirement needed for our intended purpose.
Therefore, in chapter 5 and 6, it was decided to do the tensile test rather than
other tests, to minimise the research cost. It was believed that with the tensile
properties results obtained, it can be used to observe the different characteristics of
laser sintered specimens with and without heat treatment; and from different scanning
strategy applied.
5.2 Heat Treatment Selection
As discussed in the literature review, several researchers have investigated the use of
heat treatment to improve materials’ properties, in particular; nylon and polymer.
These heat treatment studies were conducted using different setting of
temperature, time and environment. In general, it can be seen that better mechanical
properties were obtained when the heat treatment was carried out close to the melting
temperature. All these studies used materials produced using conventional processes,
except for the nylon 12 SLS material used by Zarringhalam and Hopkinson [2003].
Chapter 5. Heat Treatment Studies Page 97

In their investigation, Zarringhalam and Hopkinson [2003] conducted

conventional heating in a Zwick temperature chamber. In this research, heat
treatments in air and heat treatments in vacuum conditions has been assessed. Both
conditions were studied in order to establish the optimum condition for heat treatment
of SLS Nylon 12.
Heat treatment at lower temperature and duration is preferable from practical
perspective. Therefore, in this research, the temperature was selected for 140 oC with
two hours heat treatment duration as initial work. This was later increase to 180oC
with 16 hours duration. Then, the heat treatment was conducted with the same
temperature for 100 hours duration. Two heat treatment environments in air and
vacuum were selected to find out the effect of heat treatment on the mechanical
properties of SLS made parts. A slow heating rate of 2-3oC/minute was selected in
order to achieve homogeneous heating conditions.
5.3 Specimen Preparation
Tensile test samples were produced using the SLS Sinterstation Vanguard at Leeds
University, Leeds (see section 3.2.). These specimens were used for comparing the
properties with and without heat treatment. The tensile test specimens were fabricated
in three different built orientations; X,Y and Z using specimen design and processing
conditions as described in section 3.3 and 3.4.
5.4 Characterisations
In this study, three characteristics were measured using heat treated and non heat
treated specimens. Tensile tests were used to examine the mechanical properties,
while DSC analysis was used to determine the crystallinity and melting temperature.
Furthermore, the optical microscopy was used to observe the microstructure of the
specimens.
5.5 Experimental Work and Procedures
5.5.1 Heat Treatment in Air

A conventional heating oven, the “BINDER”, was used to heat treat the tensile
specimens from room temperature with a heating rate of 2oC/minute up to 140oC.
Then the temperature was held constant for 120 minutes before cooling at the natural

rate of the furnace to 27oC. The specimens were placed in the oven as shown in figure
5-1.
Figure 5-1 Heating oven and tensile specimen
5.5.1.1 Tensile Test Results
Tensile tests were carried out as described in section 4.4 on all three different
fabrication orientations. Six samples for each orientation were tested in order to
evaluate the tensile properties. The values of stress and strain were calculated using
equation 3.1 and 3.2. Figure 5-2 shows the stress-strain curves of heated and non-
heated specimens for all orientations while table 5-1 summarises the tensile strengths
obtained. The curves indicated that the material under-goes plastics failure, which
typical of behaviour of polymer materials.

Orientation - X Orientation - X
50 50
40 40
Stress (MPa)
Stress (MPa)
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0 0.02 0.04 0.06 0.08 0.1
Strain Strain
(a) (b)
Orientation - Y Orientation - Y
50 50
40 40
Stress (MPa)
Stress (MPa)
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0 0.02 0.04 0.06 0.08 0.1
Strain Strain
(c) (d)
Orientation - Z Orientation - Z
50 50
40 40
Stress (MPa)
Stress (MPa)
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0 0.02 0.04 0.06 0.08 0.1
Strain Strain
(e) (f)
Figure 5-2 Stress-strain graphs for each of the tensile tests in X, Y and Z orientation
(a), (c) and (e) without heat treatment; (b), (d) and (f) with heat treatment.
Table 5-1. Tensile Strength Results (MPa) – With and Without Heat Treatment
Sample 95 %
Average
1 2 3 4 5 6 (MPa)
Interval
X 40.1 37.7 41.7 32.6 36.6 29.7 36.4 ±3.62
Xht 40.9 37.1 42.6 43.0 27.9 41.8 38.9 ±4.63
Y 29.4 29.2 26.4 25.6 25.6 26.7 27.2 ±1.38
Yht 27.5 28.9 29.9 30.7 25.0 30.2 28.7 ±1.70
Z 39.7 33.3 33.1 36.4 37.6 40.3 36.7 ±2.45
Zht 44.0 39.0 42.2 36.9 37.8 44.0 40.7 ±2.52

The graphical comparison of the results with and without heat treatment in air
can be seen in figure 5-3.
50

40
30
20
10
0
X Xht Y Yht Z Zht
Orientation / Sample
Figure 5-3 Tensile strength of different fabrication orientations with and without heat
treatment (distribution for all specimens). The suffix ht indicates heat treatment.
From Table 5-1, it can be seen that the average tensile strength in the X, Y and
Z orientations after heat treatment in air improved slightly. The average increases are
6.8% in the X orientation, 5.71% in Y and 10.59% in Z. However, the overlaps in
95% confidence intervals mean that, at a 95% confidence level, the differences are not
statistically significant.
The tensile modulus for each test was calculated from the initial slope of the
straight line on the stress-strain curves and this is summarised in tables 5-2.
Table 5-2. Tensile Modulus Results (MPa) - With and Without Heat Treatment
Sample 95 %
Orien Averag
Conf.
tation 1 2 3 4 5 6 e (MPa)
Interval
X 1093.7 1065.4 1069.5 998.6 1030.5 984.6 1040.4 ±34.5
Xht 1079.8 1077.8 1124.0 1136.4 919.2 1111.8 1074.8 ±63.8
Y 874.1 846.0 814.4 790.8 789.3 839.1 825.6 ±26.8
Yht 824.0 871.1 871.1 903.7 786.8 902.0 859.8 ±36.8
Z 1092.7 1013.9 1009.8 1078.4 1061.9 1034.3 1048.5 ±27.5
Zht 1121.7 1122.6 1143.4 1082.9 1089.2 1167.0 1121.1 ±25.5

The graphical comparison of the tensile modulus results for Duraform PA

(Nylon 12) with and without heat treatment in the air environment is illustrated in
figure 5-4.
1400
1200
1000
800
600
400
X Xht Y Yht Z Zht
Figure 5-4 Tensile modulus result with and without heat treatment (distribution for all
specimens). The suffix ht indicates heat treatment.
Table 5-2 shows that the average tensile modulus in the X, Y and Z orientations
with heat treatment increased slightly compared with those without heat treatment.
The increases are 3.31% in the X orientation, 4.14% in Y and 6.93% in Z. Again,
there are overlaps in the 95% confidence intervals for the X and Y orientations, but
not for the Z orientation. Therefore, at a 95% confidence level, the increase is
significant in Z but not in X and Y.
Tensile elongation at break for each test was determined by calculated from
strain value at point where the break found as shown in Figure 5-2 for each different
fabrication orientation, with and without heat treatment, then multiplied with 100(%)
and the results is summarised in table 5-3. Figure 5-5 illustrates the graphical
comparison of the tensile elongation at break results.
Table 5-3. Tensile Elongation at Break (%)

Sample 95 %
Avera
1 2 3 4 5 6 ge (%)
Interval
X 7.20 6.50 8.10 4.70 6.60 3.90 6.17 ±1.26
Xht 7.20 5.20 8.30 7.40 4.20 7.90 6.70 ±1.30
Y 5.90 6.40 5.60 5.80 5.60 5.40 5.78 ±0.28
Yht 6.30 6.10 6.80 6.20 5.20 6.00 6.10 ±0.42
Z 6.80 4.60 4.70 5.70 5.60 7.90 5.88 ±1.02
Zht 8.70 5.40 7.50 4.60 5.00 7.00 6.37 ±1.29

10
Elongation at Break (%)

8
0
X Xht Y Yht Z Zht
From Table 5-3, it can be seen that the average tensile elongation at break
improved slightly, in which the average increases are 8.6% in the X orientation, 5.5%
in Y and 8.3% in Z after heat treatment in air. However, the differences are not
statistically significant at a 95% confidence level, as it shown that there are overlaps
in 95% confidence intervals.
5.5.1.2 Distortion Performance
In order to make sure that there were no geometric changes, the dimensions of each
specimen for different fabricated orientations were measured before and after heat
treatment using a vernier calliper. All dimensions that measured are shown in Figure
5-6 that illustrates the solid tensile specimens.
 b1
 b2  b3
L1
L
Figure 5-6 Solid tensile specimen

For each dimension L, L1, b1, b2 and b3, it was measured three times at three
different positions and then an average was calculated. To determine the distortion
before and after heat treatment the equation (5.1) was used and the results was
summarised in tables 5.4 – 5.6.

dimension after  dimensionbefore

Distortion =  100% …………. (5.1)
dimensionbefore
Table 5-4. Distortion for Heat Treated Sample (X Orientation)

Orientation Distortion (%)
/Sample L L1 b1 b2 b3
X1ht 0.00 0.00 0.33 0.33 0.00
X2ht 0.00 0.00 0.00 0.33 0.00
X3ht 0.13 0.18 0.33 0.33 0.00
X4ht 0.13 0.18 1.00 0.33 0.00
X5ht 0.00 0.00 1.00 0.66 0.00
X6ht 0.00 0.00 0.00 0.99 0.00
Table 5-5. Distortion for Heat Treated Sample (Y Orientation)

Y1ht 0.00 0.00 0.00 0.33 0.00
Y2ht 0.00 0.00 0.33 0.33 0.00
Y3ht 0.00 0.00 0.00 0.33 0.00
Y4ht 0.00 0.00 0.00 0.00 0.00
Y5ht 0.00 0.00 0.00 0.33 0.00
Y6ht 0.00 0.00 0.00 0.33 0.00
Table 5-6. Distortion for Heat Treated Sample (Z Orientation)

Z1ht 0.00 0.00 0.00 0.33 0.99
Z2ht 0.00 0.00 0.00 0.33 0.99
Z3ht 0.00 0.00 0.00 0.33 0.33
Z4ht 0.00 0.00 0.33 0.00 0.33
Z5ht 0.00 0.00 0.00 0.33 0.00
Z6ht 0.00 0.00 0.00 0.00 0.00
From table 5-4 to 5-6 it was concluded that there were no significant
dimensional changes.

5.5.1.3 DSC test results
Thermal characterisations were conducted on six samples of each individual

orientation. Figure 5-7 shows DSC plots of specimens in X orientation, without heat
treatment in air, which were obtained. DSC analysis focused on melt region between
130oC and 210oC to determine crystallinity and melting temperature. All others DSC
plots (test) results can be seen in Appendix.
Figure 5-7 DSC plots for X orientation specimen sample without heat treatment
Figure 5-7 shows that all of the DSC plots show a similar trend with single peak
and the results from other test samples were equally consistent. From DSC plots, the
degree of crystallinity can be determined (see section 3.6).
All the values of degree of crystallinity of heat treated and non-heat treated
specimens material in the X, Y and Z orientations are shown in tables 5-7 and Figure
5-8.
Table 5-7. Crystallinity Results (%) – With and Without Heat Treatment
Sample 95 %
Average
Orientation Conf.
1 2 3 4 5 6 (%)
Interval
X 24.8 27.3 27.6 27.3 27.2 26.3 26.8 ±0.86
Xht 27.4 25.8 25.9 25.9 25.6 25.9 26.1 ±0.51
Y 25.3 27.1 28.1 26.7 24.3 27.7 26.5 ±1.18
Yht 27.9 25.9 25.9 28.1 28.2 28.8 27.5 ±1.00
Z 28.6 25.3 25.4 27.4 26.9 26.3 26.7 ±1.00
Zht 25.8 20.8 26.7 22.5 25.7 25.9 24.6 ±1.88

Table 5-7 and Figure 5-8 shows that the degree of crystallinity is not
significantly changed by heat treatment.
Degree of Crystallinity (%)

40
30
20
10
0
X Xht Y Yht Z Zht
Orientation / Treatment
Figure 5-8 Crystallinity results with and without heat treatment (distribution for all
specimens). Suffix ht indicates heat treatment.
The melting peak temperature was also determined from the DSC plots and the
results are tabulated in tables 5-8 and illustrated in figure 5-9.
Table 5-8. Melting Peak Temperature (oC) – With and Without Heat Treatment
Sample Avera 95 %
Orientation ge Conf.
1 2 3 4 5 6
(oC) Interval
X 183.6 184.0 183.7 183.7 184.3 183.1 183.7 ±0.32
Xht 184.1 184.7 186.3 184.5 182.5 184.0 184.4 ±0.98
Y 183.9 183.3 184.7 183.7 183.7 182.5 183.7 ±0.58
Yht 183.6 183.2 184.7 183.6 184.6 183.0 183.8 ±0.59
Z 183.8 184.4 183.6 185.7 182.9 186.1 184.4 ±1.00
Zht 185.2 184.0 183.1 183.0 182.8 183.6 183.6 ±0.70
200
Melting Peak Temp. (C)
180
160
140
120
100
X Xht Y Yht Z Zht
Orientation/Treatment

Tables 5-8 and Figure 5-9 show that there is no significant change in the value
of the peak melting temperature due to heat treatment in air.
5.5.1.4 Optical Microscopy

In order to observe the microstructure of the specimens with and without heat
treatment, an optical light microscope was used. An Olympus BX60M optical
microscope fitted with an adjustable micrometer was set up to observe the specimen
and a JVC digital camera (KY-F55BE) fitted to the optical microscope and connected
to a PC was used to capture the images.
The samples were prepared using a cold mounting technique. 16ml of Araldite
resin and 2ml of Vantico hardener were mixed together and then poured into a mould,
and left at room temperature. The samples were released from the mould after 24
hours.
To analyse microstructure polished flat surfaces were required. The specimens
were ground using silicon carbide grinding paper, starting from a low grit number
(320), and then using progressively finer paper (grit number 500 and 1200). The
specimens were then thoroughly polished using a diamond polisher of 3 micron grit.
This was then followed by the use of 1 micron grit as a final stage of the polishing
process to remove all scratches.
The microstructures of specimens with and without heat treatment then were
observed to compare the effect of the heat treatment in air, as shown in Figure 5-10.

(a). X3 (b). XA3
(c). Z2 (d). ZA2
(e)
Figure 5-10 Microstructure samples with and without heat treatment in air, with 5X
magnification: (a) sample no.3-X orientation without heat treatment; (b) sample no.3-
X orientation with heat treatment; (c) sample no.2-Z orientation without heat
treatment; (d) sample no.2-Z orientation with heat treatment; (e). The slicing direction
of the sections.

The micrograph of heat treated and non-heat treated specimens as shown in

Figure 5.10, the porosity seems not to have reduced much in size after heat treatment
of the specimens in air environment.
5.5.1.5 Heat Treatment in Air (at 180oC, time of 100 hours)

Heat treatment in air at temperature 180oC with 100 hours time setting was also
conducted using the drying oven. However, all specimens were burned/charred,
presumably by oxidation which occurred during heat treatment. Figure 5-11 shows
selected samples after treatment. Consequently, this approach was not considered
useful.
X6 X5
Y1
Y5
X4
Y6
Figure 5-11 Cut-off specimens that burned after heat treated in air atmosphere for
180oC and 100 hours.
5.5.2 Heat Treatment in Vacuum
In order to avoid oxidation, heat treatment in a vacuum environment was undertaken.

This was at 180oC for 16 and 100 hours, with a vacuum level of 1018mbar.
A vacuum oven, “Technico”, installed at Metaltech Ltd., Consett, Durham, was
used to heat treat the tensile specimens. The heat treatment was conducted with a
heating rate of 3.3oC/minute. This temperature was held for 16 or 100 hours before
cooling naturally in vacuum conditions to a temperature of 27oC. The specimens were
placed in the vacuum oven as shown in Figure 5-12.

Figure 5-12 Vacuum oven facilities and tensile specimens
To avoid the repetition of the tests, reduce build time and also reduce
experimental cost, tensile tests were carried out on one built orientation; Y-axis only.
In addition, if mechanical properties can be improved for specimens produced in a
single orientation, then the study can be extended further to other built orientation of
specimens.
5.5.2.1 Test results at holding time 100 hours
A. Tensile Tests Results
Using the same equipment and procedure as described in section 5.5.1.1, tensile tests
were carried out and Figure 5-13 shows the stress-strain plots with and without heat
treatment. The results shown in Table 5-9 and Figure 5-14 illustrate the tensile
strength with and without heat treatment.
50 50
40 40
Stress (MPa)
Stress (MPa)
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0 0.02 0.04 0.06 0.08 0.1
Strain Strain
(a) (b)
without heat treatment and (b) with heat treatment in vacuum for 100 hours

Sample 95 %
Average
1 2 3 4 5 6 (MPa)
Interval
Y 36.1 25.1 31.6 34.1 38.2 24.2 31.6 ±4.63
Yht 29.6 45.4 33.6 36.9 40.2 39.8 37.6 ±4.44
50
40
30
20
10
0
Y Yht
all specimens). The suffix ht indicates heat treatment.
Table 5-9 shows that the mean tensile strength increased, but that the difference
was not significant at a 95% confidence level.
Tensile modulus values for each individual specimen were also determined.
The results are summarised in Table 5-10 and Figure 5-15.
Table 5-10. Tensile Modulus Results (MPa) – With and Without Heat Treatment
Sample 95 %
Orienta- Averag
Conf.
tion 1 2 3 4 5 6 e (MPa)
Interval
Y 1020.6 797.1 924.5 1007.0 1060.6 712.5 920.4 ±110.7
Yht 1002.3 1209.4 1099.9 1161.1 1202.7 1208.8 1147.4 ±66.2
1300
1200
1100
1000
900
800
700
600
500
400
Y Yht

From Table 5-10, it can be concluded that the average tensile modulus of the
heat treated samples increased by 24.66% compared to the non-heated samples, and
that this increase is significant at a 95% confidence level.
For each individual specimen, tensile elongation at break was also determined
and the results are summarised in Table 5-11 and Figure 5-16.
Table 5-11. Tensile Elongation at Break (%) – With and Without Heat Treatment
Sample 95 %
Average
1 2 3 4 5 6 (%)
Interval
Y 7.50 5.70 7.00 7.30 8.10 5.80 6.90 ±0.77
Yht 5.30 8.00 5.70 4.90 5.90 5.70 5.92 ±0.87
9
8
7
6
5
4
3
2
1
0
Y Yht
(distribution for al specimens). The suffix ht indicates heat treatment.
From Table 5-11, it can be seen that the average tensile elongation at break of
the heat treated samples decreased by 14.20% compared to the non-heated samples,
but was not significant at 95% confidence interval.
B. Distortion Performance
For heat treatment in 100 hours, any changes in the dimensions of specimens before
and after heat treatment, as shown in table 5-12, were not significant.


Y1ht 0.08 0.11 0.00 0.00 0.00
Y2ht 0.13 0.18 0.06 0.07 0.00
Y3ht 0.11 0.15 0.13 0.06 0.07
Y4ht 0.15 0.20 0.07 0.00 0.03
Y5ht 0.03 0.04 0.03 0.03 0.00
Y6ht 0.05 0.07 0.07 0.00 0.03
C. DSC Test Results
Using the same procedure as described in section 5.5.1.3, the degree of crystallinity
was determined and all the results are shown in Table 5-13 and Figure 5-17.
Sample 95 %
Average
1 2 3 4 5 6 (%)
Interval
Y 27.9 25.9 28.2 28.0 24.5 25.5 26.7 ±1.28
Yht 27.0 28.2 30.0 29.5 29.9 24.9 28.3 ±1.62
Degree of Crystallinity (%)
40
30
20
10
0
Y Yht-100
Table 5-13 shows that the average degree of crystallinity of the heat treated
samples increased by 5.81% compared to the non-heat treated samples, but was not
significant at 95% confidence interval.
The procedure described in section 5.5.1.3 was also used to obtain the values of
melting peak temperature (Tpeak) with and without heat treatment in the vacuum
environment. The results are shown in Table 5-14 and Figure 5-18.

Sample 95 %
Average
Orientation o Conf.
1 2 3 4 5 6 ( C)
Interval
Y 184.5 183.4 183.1 184.8 182.4 183.6 183.6 ±0.71
Yht 185.1 187.5 186.4 186.9 186.3 189.1 186.9 ±1.08
200
180
160
140
120
100
Y Yht-100
Table 5-14 shows significantly increased values of melting peak temperature.
D. Optical Microscopy
To observe and analyse microstructure, the same equipment and procedure as

described in section 5.5.1.4 were used. Figure 5-19 shows the microstructure of some
specimens with and without heat treatment in the vacuum environment.

(a). Y2 (b). Y2V
(c). Y6 (d). Y6V

Figure 5-19 Microstructure samples with and without heat treatment in vacuum, with
5X magnification: (a) sample no.2-Y orientation without heat treatment; (b) sample
no.2-Y orientation with heat treatment; (c) sample no.6-Y orientation without heat
treatment; (d) sample no.6-Y orientation with heat treatment
From Figure 5.19 shown the micrograph of heat treated and non-heat treated
specimens and it seems that the porosity was not to have reduced much in size after
heat treatment of the specimens.
From overall heat treatment in vacuum with 100 hours holding time, it can be
summarised, the results show that heat treatment has increased the average melting
temperature and the average tensile modulus, with this change being significant at a
95% confidence level. The average tensile strength and the average crystallinity also
show an increase, but this result is not significant at a 95% confidence level.
However, the average tensile elongation at break shows a decrease.

5.5.2.2 Test results at holding time 16 hours
A. Tensile Tests Results
The tensile strengths of these specimens were determined from stress-strain plots as
shown in Figure 5-20 and all the results are shown in Table 5-15. The graphical
comparison of tensile strength results from specimens with and without heat treatment
is shown in Figure 5-21.
60 60
50 50
Stress (MPa)
Stress (MPa)
40 40
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16
Strain Strain
(a) (b)
without heat treatment and (b) with heat treatment in vacuum for 16 hours
Sample 95 %
Average
1 2 3 4 5 6 (MPa)
Interval
Y 45.9 47.0 47.7 46.1 47.5 44.7 46.5 ±0.91
Yht 45.6 50.2 48.3 46.6 46.8 48.7 47.7 ±1.34
50
40
30
20
10
0
Y Yht

Table 5-15 shows that the average tensile strength not increased after heat
treatment, even though there were slightly different around 2.57%, but at a 95%
confidence interval there were overlaps.
Tensile modulus values for each individual specimen were also determined,
Table 5-16. Tensile Modulus Results (MPa) – With and Without Heat Treatment
Sample 95 %
Orienta- Averag
Conf.
tion 1 2 3 4 5 6 e (MPa)
Interval
Y 1256.8 1238.3 1245.0 1225.2 1167.1 1294.2 1237.8 ±33.5
Yht 1225.1 1221.9 1272.0 1258.7 1267.4 1271.5 1252.8 ±18.6
1400
1200
1000
800
600
400
Y Yht
From Table 5-16, it can be concluded that the average tensile modulus of the
heat treated samples not increased significantly compared to the non heated samples.
The increase is only 1,19%. Again, there was overlap in the 95% confidence interval.
Tensile elongation at break for each individual specimen was also determined.
Table 5-17. Tensile Elongation at Break (%) – With and Without Heat Treatment
Sample 95 %
Average
1 2 3 4 5 6 (%)
Interval
Y 11.7 14.7 15.2 11.5 13.4 9.80 12.7 ±1.66
Yht 9.10 12.9 11.4 10.5 11.1 12.7 11.3 ±1.13

16
Elonagtion at Break (%)

14
12
10
8
6
4
2
0
Y Yht
(distribution for all specimens).The suffix ht indicates heat treatment
Table 5-17 shows that the average tensile elongation at break was decrease
around 11.3% after heat treated.
B. Distortion Performance
Dimensions before and after heat treatment for 16 hours were measured and the
results showed there is no significant impact, as shown in table 5-18.

Y1ht 0.01 0.02 0.00 0.03 0.03
Y2ht 0.03 0.00 0.03 0.03 0.03
Y3ht 0.00 0.00 0.00 0.03 0.00
Y4ht 0.00 0.00 0.03 0.03 0.03
Y5ht 0.01 0.00 0.06 0.00 0.03
Y6ht 0.00 0.02 0.00 0.00 0.03
C. DSC Test Results
Using the same procedure as described in section 5.5.1.3, the degree of crystallinity
was determined and all the results are shown in Table 5-19 and Figure 5-24.

Sample 95 %
Average
1 2 3 4 5 6 (%)
Interval
Y 25.4 25.4 24.3 25.4 27.1 26.0 25.6 ±0.74
Yht 28.5 25.8 28.2 27.9 28.9 27.3 27.8 ±0.87
Degree of Crystallinity (%) 40
30
20
10
0
Y Yht-16
Table 5-19 shows that the average degree of crystallinity of the heat treated
samples increased by 8.44% compared to the non-heat treated samples.
The procedure described in section 5.5.1.3 was also used to obtain the values of
melting peak temperature (Tpeak) with and without heat treatment for 16 hours in the
vacuum environment. The results are shown in Table 5-20 and Figure 5-25.
Sample 95 %
Averag
Orientation o Conf.
1 2 3 4 5 6 e ( C)
Interval
Y 182.6 183.5 183.7 182.9 183.9 183.5 183.35 ±0.39
Yht 185.1 183.7 184.1 184.4 184.6 184.4 184.36 ±0.38
200
180
160
140
120
100
Y Yht-16

Table 5-20 shows that there was no significantly increased values of melting
peak temperature of heat treated compared to non-heat treated specimens.
It can be concluded all the results show that heat treatment in vacuum with
holding time of 16 hours has significantly increased the average crystallinity, but not
significantly increased any of the other parameter.
5.6 Summary and Discussion
The results for heat treatment in air and vacuum atmosphere as determined in section
5.5.1 and 5.5.2 summarised in Figure 5-26 to 5-30.
Tensile Strength Comparison (with and without heat treatment)
(Vacuum, 180oC, 16 hours)
(Air, 140oC, 2 hours)
0 10 20 30 40 50 60
X Xht Y Yht Z Zht
Figure 5-26 Summaries of test results: average tensile strength with and without heat
treatment. The X, Y and Z indicate fabrication orientation. The suffix ht indicates heat
treatment. Error bars are 95% confidence interval
The overall conclusion from figure 5-26 is that none of the heat treatment
regimes has significantly increased the tensile strength. ANOVA also indicated that
the differences in average tensile strength with and without heat treatment for all heat
treatment conducted were not significant effect, with all p values greater than 0.05.

Tensile Modulus Comparison (with and without heat treatment)
0 200 400 600 800 1000 1200 1400

X Xht Y Yht Z Zht
Figure 5-27 Summaries of test results: average tensile modulus with and without heat
From Figure 5-27 we can conclude that the effect of the heat treatment on
tensile modulus was mixed, with significant improvements in two out of five
comparisons. This was confirmed by ANOVA for a p-value of 0.05.

Elongation at Break (with and without heat treatment)
(Vacuum, 180oC, 16
(Vacuum, 180oC,
0 2 4 6 8 10 12 14
X Xht Y Yht Z Zht
Figure 5-28 Summaries of test results: average elongation at break with and without
heat treatment. The X, Y and Z indicate fabrication orientation. The suffix ht indicates
heat treatment. Error bars are 95% confidence interval
From figure 5-28, it can be seen that the overall conclusion that none of the heat
treatment regimes has significantly effects on elongation at break. This observation is
also confirmed by ANOVA for a p-value of 0.05.

Crystallinity Comparison (with and without heat treatment)
0 5 10 15 20 25 30 35
Crystallinity [%]
X Xht Y Yht Z Zht
Figure 5-29 Summaries of test results: average crystallinity with and without heat
From Figure 5-29 it can be seen that the effect of heat treatment on degree of
cristallinity is small and was only significant in one comparison. ANOVA confirmed
this.

Melting Peak Temperature Comparison (with and without heat treatment)
0 20 40 60 80 100 120 140 160 180 200

Melting Peak Temperature [oC]
X Xht Y Yht Z Zht
Figure 5-30 Summaries of test results: average melting peak temperature with and
without heat treatment. The X, Y and Z indicate fabrication orientation. The suffix ht
indicates heat treatment. Error bars are 95% confidence interval
Figure 5-30 demonstrates the melting peak temperature value of samples that
heat treated and non-heat treated samples. The effect is small, and significant in two
of five comparisons. This observation is also supported by ANOVA for a p-value is of
0.05.
From Figure 5-26 to 5-30 shown that heat treatment in air has generally been
established to have an insignificant effect on the mechanical properties of the
specimens; therefore there was no difference in the degree of crystallinity between
specimens with or without heat treatment. Moreover, the melting peak temperatures
were also similar. Although some small increases overall were shown between heat
treated and non-heat treated components, but these were not statistically significant. It
is also clear that the heat treatment in air has not had any impact on the anisotropy
present in the samples. This contrasts with the results from Zarringhalam and
Hopkinson [2003].
This condition may be caused by the lower amount of heat melting in the air
environment, which was not enough to completely melt the particles, and to improve
the bonding between them. This did not show a different result when compared with
and without heat treatment of the specimens. This is also confirmed by observation of

the microstructure of the specimens, which were heat treated and non-heat treated
(Figure 5-10).
Heat treatment in vacuum with 16 hours holding time and at 180 0C heating
temperature shows that there was also no improvement in all aspects involving
tensile strength, tensile modulus and elongation at break, even though there were
slight differences, but not at a 95% confidence interval. However, there was a
significant difference with the degree of crystallinity and melting peak temperature
although the increase was small. This is because, there is a possibility that the initial
characteristics of the fully sintered samples were already high prior to heat treatment
and there was no impact from further heat treatment.
However, when heat treatment is carried out in a vacuum, particularly when the
samples were heat treated with 100 hours holding time; the modulus and melting peak
temperature improved as a result of increased crystallinity. But not the tensile strength
and elongation at break even though the values increased and decreased, respectively.
These results are comparable with those obtained in previous work [Scobbo and
Hwang, 1994; Chang et al., 1998; Ramazani and Mousavi, 2005].
The observation of microstructures of the specimens with and without heat
treatment in the vacuum environment shows very similar results compared to the
microstructure heat treated in the air environment. These figures (Figures 5-10 and 5-
19), explain that even though the amount of heat melting in the vacuum is higher than
the heat melting in the air environment, but the results still show that the particles did
not bond effectively.
From the results of the experiment conducted, it demonstrates that the
mechanical properties and thermal characteristics of SLS material are different for
each orientation of fabrication (in air heat treatment and in vacuum heat treatment).
This condition is caused related to the particular characteristics of SLS material.
These differences may be due to the fact that they are produced at different times, and
whether the same machine and material were used. As a result, in this research study,
each different experimental work used for fabrication of the specimens was produced
from the same batch as a control of material properties in order to be fairly compared.
Tables 5-4 to 5-6, 5-12 and 5-19 (distortion performances) show that even
though the position of placement of the specimens in furnace was changed, the
dimensions of all samples did not change. These conditions confirmed that the
position of the product did not influence the distortion performance. This situation
means that using heat treatment to improve mechanical properties and thermal
characteristics has no relative impact on the geometry of the parts.
In vacuum heat treatment, the specimens used were built in Y orientation only.
This condition has some impact on mechanical properties and thermal characteristics
but not crystalinity, as compared to X and Z orientations. In addition, the position of
building specimen in building area was another factor that affected the characteristics.
Furthermore, the setting up of building parameters can possibly influence the results.
From all the results, the comparison of the values of tensile strength and tensile
modulus were bigger when obtained from a vacuum heat treatment rather than the
results in air heat treatment. This is because of the higher amount of energy absorbed
(heat melting) when it occurs in vacuum heat treatment. Another possibility is that,
reaction in vacuum condition causes the polymer chain to be easily released as
compared to in air environment. In addition, heat treatment in air with higher
temperature and longer time has an effect on degradation and oxidation, as shown in
Figure 5-11.
The values of crystallinity and heat melting for specimens in X, Y and Z
orientation in air heat treatment are different, in which the highest values is in Z
orientation, whereas the lowest is in X orientation. This condition shows that in Z
orientation, more energy (melting heat) was used to melt the particles in order for
more bonding to form, which can improve the mechanical properties. This condition
can be confirmed by observation of the tensile strength and modulus of Z, which is
the highest as compared to the tensile strength and modulus of X and Y orientations.
Heat treatment can provide more absorption energy and time due to dependent
phenomena to get equilibrium condition. As compared to heat treatment in air
environment, heat treatment in vacuum environment can provide more energy
absorption and can avoid / eliminate oxidative degradation. Thermal/residual stress
that occurred during fast temperature in cooling and crystallisation, stress relaxation
phenomena, re-orientation of macromolecules can be removed which was completed
during cooling in machine process [Ramazani and Mousavi, 2005].
Compare to the previous results in chapter 4 with the results in chapter 5, in the
first set of results; the tensile strength in the Y-direction is greater than that in the Z-
and X-directions. But in the second and third set of results, the tensile strength in the
Y-direction is lower than that in the X- and Z-direction.

This is thought to be the result of different densities of the build material, as

confirmed by Ho et al [1999] the strength of sintered part is closely related to the
density. As stated on page 95, chapter 4, in the Y-direction, a short scan vector causes
higher density, which gives the higher tensile strength. The shorter scan vectors can
offer uniform temperatures in sintering [Gibson and Shi, 1997].
The laser power plays the role of energy density, which means that when laser
power is higher it would give the higher energy density. When higher energy density
was produced, the powder fused thoroughly which allowed a more dense structure to
be built. However, when lower laser power was used, there is a lower energy density
and therefore there was not enough energy to melt the powder particle to create a
better bonding.
Therefore, in the first set of results, the values show that there was enough laser
energy to melt the particles and make it bond better. The lower values in the second
and third set of results might be caused by insufficient laser power. The fracture tends
to occur between the polymer particles, when specimens built under a low energy
density [Ho et al, 1999].
For instance, the ultimate tensile strength results on page 100 are very different
from the earlier results. The Y values were much smaller than those in the X and Z
directions. As reported in heat treatment in the air with a temperature of 140oC for 2
hours in chapter 5, the Y value for tensile strength and tensile modulus was the lowest
compared to X and Z direction, and also the lowest compared to the results in chapter
4. This might be because of insufficient laser power when setting up the machine, due
to the inaccurate default power settings. With limited laser power, it was not high
enough to melt the particle completely and the limited fusion of the particle powder
resulted in weak bonding.
The same trend, on page 111, the tensile modulus in the Y-direction was
reported as being 920.4MPa, which was very different from the values of 1018 and
from 825 reported earlier in the thesis. This means that there was also the variation in
the tensile modulus. This also might be because of insufficient laser power when
setting up the machine, due to the faulty default power settings.
In addition, all these variations in the results is due to the test being done of
different specimens, which were produced from different batch of production with
different quality of the material used [Gibson and Shi, 1997; Zarringhalam, 2007].
During produce, the samples are often mixed with used powder as well as virgin
powder. However, it is unknown which material contained recycled powder and how
many times that powder has been used. The quality of the component is reduced when
recycled powder is used during production; particularly when the recycled powder has
been used for a number of times [Gornet, T.J., et al, 2002].
It was not anticipated that the values of tensile strength and tensile modulus in
the Y orientation would be the lowest when compared to X and Z in the tests of heat
treatment in air and in vacuum atmosphere. A comparison of the results (in Figures 4-
2 and 4-4 and Figures 5-3 and 5-4) would suggest that the Y orientation specimens
used in the chapter 5 were flawed in some way, meaning that the results in chapter 5
for these specimens should not be considered representative.
However, since the purpose of the experiment was to investigate the influence
of heat treatment, the result of the sample can still be used as part of the study.
It is of note that no special constraints were applied in this work to reduce
inconsistency. A mixture of old and new powder was used in fabricating specimens,
and all areas of the powder bed were used to position specimens, whilst they were
being built. Work by Zarringhalam and others [Gibson and Shi, 1997; Zarringhalam,
2007] has shown that this can reduce variability; however, such strategies were not
considered in this work as they were thought to represent unreasonable constraints for
the typical user.

CHAPTER 6. HEAT TREATMENT AND

MECHANICAL PROPERTIES OF CLOSED AND
OPEN HOLLOW STRUCTURES
6.1 Introduction
This chapter describes the effect of outline/cross-section thickness on properties of

laser sintered hollowed parts. In particular, their impacts on the material’s mechanical
properties of closed and opened end hollow structure, and their thermal properties.
Moreover, an in depth description of the impact of heat treatment, on different
thicknesses of laser sintered parts of closed hollow structure.
The purpose of performing this experiment was to investigate if any changes of
mechanical properties when their crystallinity increases after heat treatment as a post
process. This observation was conducted through different geometry, particularly on
different outline/cross-section thickness.
In addition, in Section 6.6 reports the scanning strategy studies that was applied
to improve the mechanical properties of closed hollow structure on different
thicknesses. The methodology applied in this chapter is as described in Section 6.2
and Section 6.3 for samples preparation and equipment used. All characterisations,
tests, procedures and results are reported and at the end of the chapter, a summary and
brief discussion is presented.
6.2 Specimens Preparation and Material
To observe the impact of different section thickness on the mechanical and thermal
properties of closed and opened hollow structure, the tensile test specimens were
produced with hollow structure in different section thickness and solid specimens. In
addition, a similar design of specimens of closed hollow structure was also produced
to observe the effect of heat treatment on their mechanical and thermal characteristics.
All specimens were produced using the SLS Sinterstation Vanguard SLS at
Leeds University, as described in section 3.2. Basically, tensile test specimens were
built with a similar design of specimen as described in section 3.3. However, to
achieve one of the objectives of this research, the samples were then modified with
different part geometry. Hollow structure with different section thickness 1mm, 2mm
and 3mm were used.
Chapter 6. Heat treatment and Mechanical Properties of closed and open hollow structures Page 129
(a)
(b)
(c)
Figure 6-1 Tensile test specimens drawing with shell and closed hollow structure with
different section thickness (a) 1mm, (b) 2mm and (c) 3mm
Figure 6-2 Tensile test specimens drawing with shell and open hollow structure with
2mm section thickness
Figure 6-1 showed the technical drawing of closed hollow structure, while
Figure 6-2 for opened hollow with a 2mm section thickness. The closed hollow
structure specimen contains un-sintered material trapped inside the hollow shell,
whereas the opened end hollow structure is the specimen without un-sintered
material; the un-sintered material is removed from inside the specimen after
completely built from SLS machine.
For comparison, the similar design of solid specimens as control specimens, as
reported in previous chapters (chapter 4 and 5); were also manufactured from this
same batch of material to observe any changes on the properties between the solid and
the hollow specimens. All these specimens were produced in the same batch as to
avoid and minimise any differences that will affect their properties in order to
maintain the accuracy (Zarringhalam et al, 2006).
For these studies, twelve specimens were produced for each of the 1mm, 2mm,
3mm section thickness of closed hollow structure and solid section, respectively. Six
of the specimens with individual different section thickness and solid were used for
the analysis of the impact of section thickness on mechanical and thermal
characteristics and also used as a reference of non-heat treated specimens. The other
six specimens with individual different section thickness and solid were used for
analysis of the impact of non-heat treatment and heat treatment on the different areas
in terms of mechanical and thermal properties. Moreover, an additional six specimens
of 2mm section thickness of opened hollow structure (Figure 6-2) were fabricated for
investigation of the effect of closed and opened hollow structure on their tensile
properties. This was produced from a different batch. The same Nylon 12 material
was used to produce all the 54 specimens using the default setting process parameters
and default fill scanning strategy, as described in section 3.2.
Due to the hollow geometry of the specimens as shown in Figure 6-1, the
specimens have two sections or areas. One section is sintered by the laser during
sintering process while the other section consists of trapped un-sintered powder in the
hollow section of the specimens. This is because the un-sintered powders were
entrapped in the shell hollow of the specimens.
All of the designed specimens were then placed in the building envelope of the
SLS machine as shown in Figure 6-3 and were produced using default setting with
built direction along the Y-axis orientation.
Figure 6-3 Specimens position and orientation in the Vanguard SLS machine
6.3 Equipments, Experimental Work and Procedures
The same experimental work and procedures as described in chapter 3 were also used
in this study. Universal testing machine was used to carry out tensile tests and DSC
equipment was used to measure the crystallinity and melting peak temperature. In
addition, the morphology/microstructure of the specimens was observed using
Scanning Electron Microscope (SEM).
Tensile tests were carried out as described in section 3.4. From those tests, the
behaviour of the layer material on different section thickness and solid specimens was
determined. Tensile tests were carried out on one built orientation; Y-axis only. As
mentioned in chapter 5, Y-orientation of specimens was used to avoid the repetition of
the tests, reduce build time and also reduce experimental cost. Moreover, if
mechanical properties can be improved for specimens produced in a single

orientation, then the study can be extended further to other built orientation of
specimens.
Each individual different section thickness and solid was tested for six samples
to evaluate tensile properties, as described in chapter 4. The values of stress and strain
were calculated using equation 3.1 and 3.2. The area of the sintered section was
determined by using the original formula for the determining the area of solid round
bar. To calculate the area of the solid round bar, the outer diameter (d o) is squared,
multiplied by phi (π) and then divided by four (π/4*d 2). However, to work out the
area of the hollow specimen, the formula has been modified by subtracting the inner
diameter (di) from the outer diameter (do) of the hollow structure specimen. Therefore,
the formula to find out the measurement of the area of hollow specimen is = π/4*(d o2-
di2). Figure 6-4 shows the cross section of hollow specimen.
Figure 6-4 Cross section of hollow specimen
The measurement of the outer diameter for solid and hollow structure specimens
was measured before the specimen were tested, meanwhile for the inner diameter for
specimens with hollow structure; it were measured (approximately) after the
specimens broke. The measurement was conducted three times at three different
positions and then an average was calculated.
The mechanical and thermal properties results were then analysed using
statistical analysis tool (Minitab software), in which the p value, average and
confidence interval can be determined, as described in chapter 3 section 3.7.
6.4 Impact of Section Thickness
6.4.1 Tensile Test Results for Non-heat treated Specimens
Stress strain graphs of all the specimens for 1mm, 2mm, 3mm, and solid are shown in
Figure 6-5 and Figure 6-6. In addition, Figure 6-7 shows the specimen of each section
thickness after tensile test. The tensile test results for different thickness of section,
closed and opened hollow structure and solid section in non-heat treated specimens
are shown in Figure 6-8 to Figure 6-13.
2mm Thickness
1mm Thickness 60
60
50
50
Stress (MPa)
40
Stress (MPa)
40
30
30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
Strain Strain
(a) (b)
3mm Thickness Solid
60 60
50 50
Stress (MPa)
Stress (MPa)
40 40
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
Strain Strain
(c) (d)
Figure 6-5 Stress-strain graphs for each of the tensile tests non-heat treated specimens
for closed hollow with section thickness (a) 1mm, (b) 2mm, (c) 3mm and (d) solid
2mm Thickness (open hollow)

60
50
Stress (MPa)
40
30
20
10
0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0.16 0.18
Strain
Figure 6-6 Stress-strain graphs for each of the tensile tests non-heat treated specimens
for open hollow with 2mm section thickness
Figure 6-7 Tensile test specimens (after testing) for closed hollow structure with
1mm, 2mm and 3 mm section thickness
Figure 6-8 Tensile strength result of different section thickness of closed hollow and
solid non-heat treated specimens. Error bars are at 95% confidence interval
Based on Figure 6-5 and Figure 6-6 all the specimens displayed a linear elastic
behaviour within the same region of strain values between 0.0 and 0.02. It can be
observed Figure 6-8 that the smaller thickness section created a lower average tensile
strength and the solid specimens promoted the highest average tensile strength. The
ANOVA confirms that the average tensile strength on each different section thickness
and solid specimens increased significantly, in which the p value was smaller than
0.05. This indication confirmed that section thickness have effect on average tensile
strength significantly.
The tensile modulus for each specimen with different wall thickness and solid
samples were then determined from the initial slope of the straight line on the stress-
strain curves as explained in section 3.5.1.b. Tensile modulus results for all specimens
then are shown in Figure 6-9.
Figure 6-9 Tensile modulus result of different section thickness of closed hollow and
solid non-heat treated specimens. Error bars are at 95% confidence interval
It was clear that the solid specimens promoted the highest average tensile
modulus, while the smaller section thickness specimens created a lower average
tensile modulus, in which the smallest average tensile modulus occurred when the
smallest thickness of 1mm section specimens were used. The ANOVA confirmed that
the p value was smaller than 0.05 when the average tensile modulus on each different
section thickness and solid specimens increased significantly. This indication showed
that the outline/cross-section thickness and solid have an effect on average tensile
modulus significantly.
The elongation at break for each specimen with different section thickness and
solid were then determined as described in section 3.5.1.b. Elongation at break results
for all specimens are illustrated in Figure 6-10.

hollow and solid non-heat treated specimens. Error bars are at 95% confidence
interval
The smaller the sections thickness, the lower the average elongation at break
obtained. While the solid specimens promoted highest average elongation at break.
However, for section thickness 1mm and 2mm, there was not significantly difference,
whereas for the 3mm section thickness and solid, it was significantly different.
The results of tensile properties of opened end hollow structure were obtained
and shown in Figure 6-11 to Figure 6-13. As described in previous section 6.2, these
result then compare to the closed hollow structure to observe if there is any different
impact of closed and opened end hollow structure.
Figure 6-11 Summaries of test results: average tensile strength of closed and open
Figure 6-11 illustrates the tensile strength for 2mm section thickness of open
hollow specimens. From Figure 6-11 and Figure 6-8 It can be seen that there was no
significant change in the average tensile strength between closed and open hollow
2mm section thickness specimens, as it was also confirmed by ANOVA in which the
p value was greater than 0.05.
Figure 6-12 Summaries of test results: average tensile modulus of closed and open
Figure 6-12 shows the average tensile modulus results of open hollow
specimens with 2mm section thickness. From these evidences, it can be observed that
there were no significant changes between the average tensile modulus of open
hollow specimen and closed specimen. This was confirmed by ANOVA for a p value
that was greater than 0.05.
Figure 6-13 Summaries of test results: average elongation at break of closed and open
Figure 6-13 shows the average elongation at break of closed and open hollow
specimens with 2mm section thickness. it can be seen that there was no changes of
elongation at break for open hollow structure, although there were small differences
of average at around 0.16 %, but not significant enough at a 95 % confidence level.

The ANOVA confirmed this by showing a p value that was greater than 0.05.
Figure 6-11 to Figure 6-13 show that there were no differences in the tensile
properties of the specimens with closed and open hollow structure. Even though there
were slight different in average tensile strength by 2.24 MPa, average tensile modulus
by 115.62MPa and average elongation at break by 0.16%, but they were not
significant at 95% confidence level. Therefore, it can be summarised that whether the
structure of the specimen were open or closed hollow, it had no effect on the
mechanical characteristics. These differences possibly occurred due to the fact that
both specimens of closed and open structure were manufactured in different batch of
production, as it was confirmed by Zarringhalam et al. [2006] in their research.
6.4.2 Crystallinity Characterisation Results for Non-heat treated Specimens

Using the same procedure as described in section 3.7, the degree of crystallinity of
sintered and un-sintered material from different section thickness specimens and
sintered material from solid specimens were determined and all the results are shown
in Figure 6-14. All these specimens were non-heat treated specimens.
Figure 6-14 Degree of crystallinity of sintered and un-sintered material of non-heat

treated specimens with different section thickness and solid. Error bars are at 95%
confidence interval
From Figure 6-14 it can be seen that the average degree of crystallinity for
sintered material were between the range of 23.65 % and 25.65 % for the all
specimens, even though there was slight difference for the specimens with 1mm,
2mm, 3mm section thickness and solid specimens, but the difference was not
significant. Moreover, as it can be observed that the degree of crystallinity for un-
sintered material of each specimen in each different section thickness were also not
significantly different. However, the average degrees of crystallinity of un-sintered
material were higher compared to sintered material. As indicated Figure 6-14, the
difference was not significant for each specimen in 1mm, 2mm, 3mm section
thickness and solid specimens.
6.4.3 Melting Peak Temperature (Tpeak) Results for Non-heat treated

Specimens
This section describes the impact of crystallinity on melting peak temperature (Tpeak)
of each specimen with different wall thickness and solid, for non-heat treated
specimens. This was determined from the DSC plot.
The comparison of melting peak temperatures for sintered and un-sintered
material, with different section thickness is illustrated in Figure 6-15.
Figure 6-15 Melting peak temperature sintered and un-sintered material of non-heat
treated specimens with different wall thickness and solid. Error bars are at 95%
confidence interval
Figure 6-15 show that there was no significant change in the value of the peak
melting temperature on the different section thickness of the specimens, for un-
sintered material, in which the temperature was between 189.41oC to 189.57oC.
However, for sintered material, the average melting peak temperature was
significantly different for each individual section thickness and solid specimens. The
average temperature were 184.51oC, 186.14oC, 183.55oC and 181.59oC for 1mm,
2mm, 3mm section thickness and solid, respectively. It was clear that the melting
temperatures for all section thickness and solid were not consistent and did not follow
a specific pattern. Furthermore, by comparing to the melting peak temperature for un-
sintered specimens, sintered specimens have lower melting peak temperature.
6.4.4 Morphology and Microstructure for Non-heat treated Specimens

To observe and analyse the morphology of the fabricated parts, a Scanning Electron
Microscope (SEM) as described in section 3.8 were used. Figure 6-16 shows the
microstructure of selected specimens that were observed by using SEM. In addition,
6-17 illustrates un-sintered material particle grain (without heat treatment).
(a). Y1 (b). Y1 middle
(c). Y1 outer edge (d). Y1 inner edge

Figure 6-16 Microstructure samples without heat treatment in vacuum; sample with
3mm section thickness (a) surface cutting, (b) in the middle, (c) outer edge (d) inner
edge of specimen
Figure 6-16 (a) and (b) show parts of the surface of specimens made with the
default setting of SLS machine. It can be seen that the specimens consists of porous
surface, which probably occurred during the sintering process. From Figure (c) and
(d), it can also be seen that the presence of rough surface may be due to ineffective
sintering or low degree of powder particle melting during the laser sintering.
Furthermore, there was an increase of serration in the surface texture and this was
possibly due to the un-sintered particle grain size (d).
Figure 6-17 Un-sintered material particle grain (without heat treatment)
Figure 6-17 shows the un-sintered particle grain without heat treatment. It can
be seen that the un-sintered particle have spherical shapes with diameter of between
60-80micron. Due to the condition of this is un-sintered material, it means that
during the SLS process, this particle was not affected by laser and remained like the
original powder of Nylon 12. Therefore, the SEM showed that between the particles,
there was no bonding amongst each other. However, when compared with the
previous Figure 6-16, all particles bonded as a result of laser sintering process.
6.5 Impact of Heat Treatment on Properties of Different Section Thickness
As further experimental work in this research study, impact of heat treatment on

mechanical and thermal properties on different wall thickness on closed hollow
structure were investigated. The same material Nylon 12 was also used and produced
in the same methods as the material used in the non-heat treated specimens. The
specimens were then heat treated in vacuum environment by using the same vacuum
furnace as described in sub-section 5.5.2. This experimental work was conducted to
observe if there was an impact on each individual different section thickness and solid
specimens after heat treatment in term of mechanical and thermal properties as
compared to non-heat treated specimens.
From the previous studies in Chapter 5, when specimens were heat treated for 2
hours at 140oC and for 16 hours at 180oC there was no significant effect on
mechanical and thermal properties as observed. Therefore, in this heat treatment the
temperature setting was selected directly at 180 oC (near melting point of Nylon 12),
for 100 hours with the vacuum level setting at -1018mbar (the maximum capacity of
the available vacuum oven).
6.5.1 Tensile Tests Results for Heat treated Specimens

Figure 6-18 shows the plot of stress-strain heat treated specimens for different section
thickness and solid. From this Figure, the tensile strengths of these specimens were
determined and all the results are shown in Figure 6-19.
1mm Thickness 2mm Thickness

60 60
50 50
Stress (MPa)
Stress (MPa)
40 40
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Strain Strain
(a) (b)
3mm Thickness
Solid
60
60
50
50
Stress (MPa)
40
Stress (MPa)
40
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0.12 0.14 0 0.02 0.04 0.06 0.08 0.1 0.12 0.14
Strain Strain
(c) (d)
Figure 6-18 Stress-strain plots for each of the tensile tests of heat treated specimens
with section thickness, (a) 1mm, (b) 2mm, (c) 3mm and (d) solid
and solid heat treated specimens. Error bars are at 95% confidence interval
Figure 6-19 shows the average of tensile strength obtained from all heat treated
specimens. It was clear that the average tensile strength of each different section
thickness and solid specimens were significantly different. This observation
confirmed that there is a correlation between thickness and tensile strength where
tensile strength increases from smallest thickness section to solid. The ANOVA
confirmed this with the p value of 0.05.
Tensile modulus values for each individual specimen after heat treatment was
determined and the results are summarised Figure 6-20.
and solid heat treated specimens. Error bars are at 95% confidence interval.
Figure 6-20 illustrates the average tensile modulus obtained after heat
treatment. This observation confirmed again that this a correlation between thickness
and tensile modulus where tensile modulus increases from smallest thickness section
to solid. The ANOVA confirmed that the p value was smaller than 0.05.
Elongation at break values for each individual heat treated specimen was
determined and the results are summarised in Figure 6-21.

hollow and solid heat treated specimens. Error bars are at 95% confidence interval
Figure 6-21 demonstrates that the average elongation at break for 1mm, 2mm,
3mm section thickness and solid specimens after heat treatment. This indicated that
they were not significantly different, even though there were small differences of the
distribution elongation at break among them in each individual section thickness.
However, compared to 1mm, 2mm and 3mm section thickness, the average elongation
at break of solid specimens was 11.68 % increased dramatically.
6.5.2 Crystallinity Characterisation Results for Heat Treated Specimens

Using the same procedure as described in section 3.6, the degree of crystallinity of
sintered and un-sintered material of heat treated specimens was determined and all the
results are shown in Figure 6-22.
Figure 6-22 Degree of crystallinity of sintered and un-sintered material result of heat
confidence interval
It can be seen from Figure 6-22 that the average degree of crystallinity for
sintered material were between the ranges of 27.30% and 29.99 %. Even though there
was slight differences as illustrated in, but they are not significant. It also shows that
the average degree of crystallinity for un-sintered material for all section thickness
was also not significantly different. The average degree of crystallinity ranges from
39.17 % to 40.73 %. However, the average degree of crystallinity of un-sintered
material is higher than the sintered material for all heat treated specimens.
6.5.3 Melting Peak Temperature (Tpeak) Results for Heat treated Specimens
The same procedure as described in section 3.7 was used to obtain the values of
melting peak temperature (Tpeak) with and without heat treatment in the vacuum
environment. The results are shown in Figure 6-23.
Figure 6-23 Melting peak temperature of sintered and un-sintered material heat
confidence interval
Figure 6-23 shows the average melting temperature for sintered and un-sintered
material for all heat treated specimens. The average melting peak temperature for
sintered material and un-sintered is not significantly different. The average melting
temperature for sintered specimens ranges from 186.69 oC to 193.24 oC while for un-
sintered is from 192.48oC to 194.61oC. The range different constitute less than 5%
difference which is not significant.
6.5.4 Morphology and Microstructure for Heat treated Specimens

To observe and analyse microstructure, the same equipment and procedure as
described in section 3.8 were used. Figure 6-24 shows the microstructure of some
selected specimens after heat treatment in the vacuum environment.
(a). Y3h5 (b). Y3h5-outer edge
(c). Y3h5 inner edge (d). Y3h5 middle

Figure 6-24 Microstructure specimens with heat treatment in vacuum, sample with
3mm section thickness (a) surface cutting; (b) outer edge (c) inner edge (d) in the
middle of specimen
Figure 6-24 (a) shows part of the surface of specimens made with default
setting on SLS machine, but after being heat treated in a vacuum environment. As it
can be seen in the photograph, the porous surface still appeared even though it was
heat treated. This probably occurred during the sintering process. From Figure (b) and
(c), it can also be seen in the outer and the inner edge of 3mm section thickness
specimens that the surface seems to be more solid even though uneven surface texture
was still present. Again, the presence of uneven surface texture may be due to
ineffective sintering or a low degree of powder particle melting during laser sintering.
Even though it was heat treated, there was probably insufficient amount of energy to
melt the powder particle and form a smoother surface. Moreover, it was also observed
that (d) had similar surface roughness in the middle area to (b) and (c).
6.6 Impact of Scanning Strategy (fill and Outline Scanning) on Mechanical

Properties
In this sub-section, the characteristics of mechanical properties of Nylon 12 laser

sintered material with different methods of scanning strategy are reported.
Six numbers of the closed hollow specimens with similar design as described in
section 6.2 produced in different section thickness (1mm, 2mm and 3mm) and solid
using the SLS Vanguard machine. However, in this study, the different setting of
outline scanning strategy was applied.
The default process setting for the study in this section is as shown in table 3-1,
except the setting for the laser power in outline scanning was altered from the default
of 6watt (default) to 41watt. This condition was selected to observe if any impact on
mechanical properties if outline laser power similar as fill laser power of 41watt. This
was intended to use this outline strategy in built the specimens rather than repeated fill
scanning motion again (for secondly or after). Therefore, this laser magnitude was the
same as the default setting for laser power of fill scanning. In addition, the fill and
outline scanning traced in one motion only following the design of the specimen. The
laser scan moved in the same direction as the fill scanning as discussed in the
previous chapters (4 and 5) and previous sections (6.4 and 6.5) of this thesis.
The tensile test was then carried out using the same universal testing machine
that was also used to examine the mechanical characteristics as reported in the
previous chapter 4 and 5, and previous section 6.4 and 6.5 . The mechanical properties
of specimens were then measured to analyse whether there was any changes and
impact on the different thickness of section and solid.
Figure 6-25 illustrates the scanning path of fill and outline scanning strategy for
solid and different section thickness specimens when it was produced.
X
0.242
“Outline””
0.248
0.268
“Fill” laser beam

0.262 movement
Figure 6-25 Scan path of fill and outline scanning strategy for (a) solid and (b) section
thickness specimens
The obtained results from this implementation of different scanning strategy

were then compared with the previous results as described in section 6.4 that was built
using the default scanning strategy (it was also mentioned as non-heat treated
specimens).
6.6.1 Tensile tests results

For specimens with different section thickness 1mm, 2mm, 3mm of closed hollow
structure and solid with fill and outline scanning strategy, tensile strengths were
determined and all the results are shown in Figure 6-27. Meanwhile Figure 6-26
shows the plot of stress-strain relationship. In addition, the average of tensile modulus
and elongation at break is presented in Figure 6-28 and Figure 6-29.
2mm Thickness
1mm Thickness
60 60
50 50
Stress (MPa)
Stress (MPa)
40 40
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0.12 0 0.02 0.04 0.06 0.08 0.1 0.12
Strain Strain
(a) (b)
3mm Thickness Solid
60 60
50 50
Stress (MPa)
Stress (MPa)
40 40
30 30
20 20
10 10
0 0
0 0.02 0.04 0.06 0.08 0.1 0.12 0 0.02 0.04 0.06 0.08 0.1 0.12
Strain Strain
(c) (d)
Figure 6-26 Stress-strain graphs for each of the tensile tests with fill and outline
scanning specimens for closed hollow with section thickness, (a) 1mm (b) 2mm, (c)
3mm and (d) solid
and solid specimens with fill and outline scanning. Error bars are at 95% confidence
interval
Figure 6-27 shows the average of tensile strength obtained from all specimens
with fill and outline strategy implemented. It can be seen that the distribution of
tensile strength on each individual different section thickness and solid specimens was
almost similar. In addition, it was clear that the average tensile strength tend to
increase from smallest section thickness to solid specimens. The ANOVA confirmed
this with the p value of 0.05.
interval
Figure 6-28 illustrates the average tensile modulus obtained from the specimens
with fill and outline implemented. In addition, the average modulus was different on
each different section thickness and solid specimens. This indicated an increasing
trend with the changes in section thickness. This observation confirmed that the
average tensile modulus of fill and outline scanning samples increases as the thickness
of section of the specimens increases up to solid. The ANOVA confirmed that the p
value was smaller than 0.05.
Figure 6-29 Elongation at break results of different section thickness of closed hollow
interval
Figure 6-29 demonstrates that the average elongation at break for 1mm, 2mm,
3mm section thickness and solid specimens with fill and outline scanning strategy..
This indicated that there was an increase in the average of elongation as the thickness
increased. This indication was also confirmed by The ANOVA in which the p value
was smaller than 0.05.
6.6.2 Energy Density

As described in section 2.2, energy density is one important and direct
processing parameter. This research investigates the effect of implementation of
different scanning strategy on their mechanical properties, on different section
thickness and solid specimens. The total energy density in each individual different
section thickness, with fill scanning can be determined by using the equation (6.1),
meanwhile for fill and outline scanning by using the equation (6.2):
ED
Total energy = xLxd ….……….. (6.1)
A
ED Po
Total energy = [ x L x d] + …………... (6.2)
A A
P
ED is Energy density as shown in equation 2.1 = (Joule/cm2), where P is fill
BS * SP
laser power (watt), BS is laser beam speed (cm/s), SP is scan spacing (cm), A is area
scanned, L is scan path length (cm), d is diameter of beam of laser (cm), Po is outline
laser power (watt).
The scan path length is the total number of length of laser scanning that traces
the material in the designated area in order to sinter it. This is then repeated until all
layers of the specimens have been completely sintered. Scanning pattern of scan path
as described in section 3.3, in which the laser moves parallel to X axis and
perpendicular of Y axis, was also applied in this study when producing the specimens.
When referred to the theory of scanning laser traces, as shown in Figure 2-11 and 2-
12 the scan path length can be determined. Using the pictures as shown in Figure 6-30
and Figure 6-31, the scan path length can be calculated due to the nature of the shape
of the specimens, which are round hollow bars and solid.
Figure 6-30 Scan path length of Fill (F) and Fill and outline (F/O) for different section
thickness and solid specimens
Figure 6-31 Pythagorean theory of each layer
To calculate the scan path length for each layer, from Figure 6-30, it can be
determined that the chord is 2rsinƟ. It is then calculated by multiplying this number
with the length of the specimen in Y direction (see Figure 6-3 specimens’ position and
orientation in the Vanguard SLS machine). In X direction, the number of scan path
length is the area of specimens divided to scan spacing for each layer then multiply
with the number of layers.
All the results obtained were presented in Figure 6-32 for specimens with fill
and fill and outline scanning strategy, with different section thickness and solid.
Figure 6-32 Total energy density of different section thickness and solid specimens
for non-heat treated with fill (non-HT with F scan) and non-heat treated with fill and
outline (non-HT with F/O scan) strategy.
Figure 6-32 shows the average total energy density obtained from all specimens
that was built with the fill scanning strategy (default) and with fill and outline
scanning strategy. It also can be observed that the average total energy density was
different on each different section thickness and solid specimens. The average total
energy density increased when there was a greater section thickness and it was the
highest with the solid specimens. This means that the sections with the smaller
thicknesses had a lower energy density compared to the solid specimen, which had the
highest energy density even though they were built using the fill scanning method.
This pattern was also seen when the specimens were produced by using fill and
outline strategy. The ANOVA confirmed this as both of the p values was smaller than
0.05.
6.7 Summary and Discussions
The tensile characteristics results for specimens with different section thickness of
closed and opened hollow structure and solid, for non-heat treated and heat treated,
for specimens with the implementation of different setting of fill and outline scanning
strategy as determined in section 6.4, 6.5 and 6.6 are summarised in Figure 6-33 to
Figure 6-35. Moreover, the summaries of their thermal characteristics are
demonstrated in Figure 6-36 to 6-39.
60
50
40
30
20
10
0
1mm 2mm 2mm (*) 3mm Solid
Section Thickness / Solid
non-HT with F scan HT with F scan non-HT with F/O scan
Figure 6-33 Comparison of average tensile strength of different section thickness and
solid specimens of non-heat treated (non-HT) with fill (F) scan, heat treated (HT) with
fill (F) scan and non-heat treated (non-HT) with fill and outline (F/O) scan. (*) open
hollow and others are closed hollow structure. Error bars are at 95% confidence
interval.
From Figure 6-33 it can be seen that the average tensile strength of heat treated
and non-heat treated specimens, in individual section thickness and solid specimens
were different. As it was described in section 6.4 and 6.5 , the smaller average value
of tensile strength can be seen on the smaller section thicknesses and the highest
average value of the tensile strength was for the solid specimens, for both heat treated
and non-heat treated specimens. This observation confirmed that the average tensile
strength was influenced by the different section thickness.
Furthermore, heat treated specimens have a higher value of tensile strength on
average compared to all non-heat treated specimens for individual different section
thickness and solid. From this Figure, it can be concluded that section thicknesses and
the heat treatment effects the average tensile strength and making them increase
significantly at a 95% confidence level. ANOVA also indicated that there was
significant increase on the average tensile strength on difference section thicknesses
with all p values smaller than 0.05. For all samples with a different section thickness
and solid, average tensile strength improved in the range of 2 – 6.34 MPa even when
samples that have been and have not been heat treated were used.
From Figure 6-33 it can be seen that there was significant effect of the
implementation of fill and outline scanning strategy on average tensile strength. The
average tensile strength increases as the section thickness increases up to solid
specimens compared to the specimens that were built using fill scanning strategy
only. This indication confirmed that there was an improvement in the average tensile
strength for 1mm section thickness by 12.50 MPa, while for 2mm it was by 4.20 MPa
and for 3mm and solid it was by around 2.99 MPa for both. ANOVA confirmed this
condition, in which the p value was smaller than 0.05.
2000
1500
1000
500
0
Figure 6-34 Comparison of average tensile modulus of different section thickness and
solid specimens of non-heat treated (non-HT) with fill (F) scan, heat treated (HT) with
fill (F) scan and non-heat treated (non-HT) with fill and outline (F/O) scan. (*) open
hollow and others are closed hollow structure. Error bars are at 95% confidence
interval.
From Figure 6-34 it can be observed that the average tensile modulus increased
with the increase of section thickness, for all specimens with and without heat
treatment. This indication has also been described in section 6.4 and 6.5. However,
the average value of tensile modulus of the heat treated specimens was higher
compared to the values of non-heated specimens, in each different individual section
thickness and solid. Again, this confirmed that average tensile modulus depended on
the section thickness and whether heat treatment was carried out on the specimen.
From this figure, it can be concluded that heat treatment influences the average tensile
modulus and the significant is at a 95% confidence level. The ANOVA confirmed this
significant increase with the p value being smaller than 0.005. Compared to the
specimens without heat treatment, the improvement of heat treated specimens
increased by 277.05MPa, 214.51MPa, 214.22MPa and 161.63MPa for 1mm, 2mm,
3mm and solid samples, respectively.
From Figure 6-34, it can be observed that the specimens built with fill and
outline scanning strategy has higher average tensile modulus compared to the
specimens that was built with fill scanning strategy only. It was clear that the average
tensile modulus increased significantly at a 95% confidence interval. ANOVA
confirms this because the p value was smaller than 0.05.
It can be summarised that there was a significant effect of fill and outline
scanning strategy on the average tensile modulus on each different section thickness
and solid specimens. This was indicated by the increase of the average tensile
modulus by 508.30 MPa, 258.08 MPa, 159.02 MPa and 162.05 MPa for 1mm, 2mm,
3mm section thickness and solid specimens, respectively.
16
Elongation at break (%)

14
12
10
8
6
4
2
0
Figure 6-35 Comparison of average elongation at break of different section thickness

and solid specimens of non-heat treated (non-HT) with fill (F) scan, heat treated (HT)
with fill (F) scan and non-heat treated (non-HT) with fill and outline (F/O) scan. (*)
open hollow and others are closed hollow structure. Error bars are at 95% confidence
interval.
As shown in Figure 6-35, the trend of elongation at break in all heat treated and
non-heat treated specimens were different to the trend of tensile strength and tensile
modulus. It can be seen that the average elongation at break for heat treated hollowed
specimens were similar. However, the elongation at break value decreased when heat
treated specimens were used as compared to non-heat treated specimens. Although the
value decreased, it is not significant enough to be at a 95% confidence level, except
for specimens with 3mm section thickness where there was a significant decrease in
the value. This indicated that there was no effect of heat treatment on elongation at
break.
Figure 6-35 demonstrates that the implementation of fill and outline scanning
strategy did not significantly affect the average elongation at break on each different
section thickness and solid specimens. It seems that there was no increase on the
average elongation at break for specimens with 1mm and 2mm. However, the average
elongation for 3mm section thickness and solid specimens decreased. Even though
there was a slight increase and decrease, they were not significantly at 95%
confidence level. ANOVA confirmed this condition, in which the p value was greater
than 0.05. It can be summarised that there was no impact on average elongation at
break by using fill and outline scanning strategy.
From Figure 6-33 to 6-35 in can be summarised that the section thickness of the
specimens influenced the average tensile strength and average tensile modulus, but
not fully impact the elongation at break. Furthermore, it also can be summarised from
the results obtained that the heat treatment at 180oC for 100 hours in vacuum
environment brought about an improvement, particularly for tensile strength and
tensile modulus for each different section thickness and solid specimens.
From Figure 6-33 to Figure 6-35 it can be summarised that compared to
specimens that built using fill scanning strategy, the average tensile strength and
average tensile modulus of specimens that built with fill and outline scanning strategy
were increased for each individual different section thickness and solid specimens.
However, this was not the case for elongation at break. This also indicated that the
implementation of different set of fill and outline scanning obtained a higher total
energy density as shown Figure 6-32. The total energy density increased with the
greater section thickness and the highest occurred with solid specimens.
Without HT With HT
35
30
Crystallinity (%)
25
20
15
10
5
0
1mm 2mm 3mm Solid
Figure 6-36 Summaries of test results: average crystallinity sintered material with and
without heat treatment of different section thickness and solid specimens. Error bars
are 95% confidence interval
Figure 6-36 shows that the average degree of crystallinity of sintered material
with heat treated were higher compared to average degree of crystallinity of the same
material but non-heat treated. It was clear that the increase of average degree of
crystallinity after heat treatment was significantly at a 95% confidence level. This was
also confirmed by ANOVA which showed that the p value was smaller than 0.05.
It can be summarised that the heat treatment affected the average degree of
crystallinity for sintered material. Compared with and without heat treatment, the
improvement of degree of crystallinity by 14.97%, 13.31%, 26.81% and 10.30% for

1mm, 2mm, 3mm and solid specimens, respectively.
Without HT With HT
45
40
35
Crystallinity (%)
30
25
20
15
10
5
0
1mm 2mm 3mm
Figure 6-37 Summaries of test results: average crystallinity un-sintered material with
and without heat treatment of different section thickness and solid specimens. Error
bars are 95% confidence interval
From Figure 6-37, although there was a slight improvement on the average
degree of crystallinity on thickness of section 1mm (2.34%), 2mm (1.82%) and 3mm
(1.28%), it can be observed that there were no significant differences for un-sintered
material, for each of the different section thickness, from heat treated and non-heat
treated specimens. This indicated that there was no significant effect of heat treatment
on the average degree of crystallinity for un-sintered material. ANOVA confirmed
this conclusion by showing a p value which was greater than 0.05.
Without HT With HT
200
Melting Temperature (C)

195
190
185
180
175
170
1mm 2mm 3mm Solid
Figure 6-38 Summaries of test results: average melting peak temperature sintered
material with and without heat treatment of different section thickness and solid
specimens. Error bars are 95% confidence interval
From Figure 6-38, it shows that for sintered material, the average melting peak
temperature increased significantly at a 95% confidence level by 8.73oC, 4.16oC,
8.89oC and 3.10oC for 1mm, 2mm, 3mm and solid specimens, respectively. This
indication confirmed that there was an increase of the average melting temperature
compared to non-heat treated specimens, even though in all specimens the
improvement was lower than 5%. This was clear that the heat treatment affected the
average melting peak temperature. The ANOVA confirmed this with a p value was
smaller than 0.05.
Without HT With HT
200
Melting Temperature (C)
195
190
185
180
175
170
1mm 2mm 3mm
Figure 6-39 Summaries of test results: average melting peak temperature un-sintered
material with and without heat treatment of different section thickness of specimens.
Error bars are 95% confidence interval
Figure 6-39 illustrates that heat treatment affected the average melting peak
temperature for un-sintered material significantly at a 95% confidence level. ANOVA
confirmed this condition, in which the p value was smaller than 0.05. Compared to the
specimens without heat treatment, the average melting peak temperature increased by
5.20oC, 2.91oC, 3.10oC for the specimens with 1mm, 2mm and 3mm section
thickness, respectively. However, the improvement was lower than 5% in all
specimens.
From Figure 6-36 to Figure 6-39, it can be summarised that the section
thickness had no effect on the average degree of crystallinity, either for heat treated or
non-heated specimens. However, the degree of crystallinity was higher after heat
treatment was carried out. This indicated that the heat treatment at 180oC for 100
hours in vacuum environment provoked an improvement in degree of crystallinity,
particularly for laser sintered material.
6.7.1 Effect of Section Thickness of Closed and Open Hollow Structure
Figures 6-33 to 6-39 show the comparison of mechanical properties and thermal
properties of different outline/cross-section thickness and solid specimens of Nylon
12 laser sintered material.
It can be seen that the outline/cross-section thickness influences the mechanical
properties. The figures demonstrated that the tensile properties increased when the
thickness of the outline/cross-sections were greater.
As shown in Figure 6.5(d) on page 134, the variation in the maximum tensile
strength of the solid sample is not huge. However, the variation in the elongation at
break of the same sample is huge for the solid sample.
The layer-to-layer bonding region in solid specimen is larger than for hollow
specimen. Thus greater energy is needed to break the layer-to-layer bonding. In
addition the laser scan lines on the solid samples are longer than the close-hollow
samples. Longer scan lines allows the previously scanned line time to cool and
solidify while shorter laser scan lines limits the cooling time between lines. This
resulted in stronger fusion of particle-to-particle bonds that increases ductility and are
more difficult to break [Nelson J.A., et al, 2014]. Furthermore in the region of effect
of end of vector (EoV) for hollow specimens are very together which exposure the
specimens to longer and higher intensity laser. This will causes greater sintering and
increased densification and affect the mechanical properties [Ajoku, U. et al, 2006].
This is illustrated by Figure 6-40 below.
EoV
EoV
(a) Hollow structure (b) Solid structure
Figure 6-40. End of vector effect (a). Hollow structure (b). Solid structure
Moreover, the variation in the tensile modulus changes with the different
section thickness. Laser sintered materials such as Nylon 12 exhibits anisotropy
behaviour (Faes, M. et al, 2015, Ajoku, U. et al, 2006b] due to the influence of
processing parameter. This is different from most polymer materials that are
manufacture from bulk material through moulding or extrusion processes that tend to
produce isotropy behaviour [Griskey, 1995]. Beside the orientation, secondary effect
of end-of-vector (EoV) during the sintering process will produce different tensile
modulus for varying cross section [Ajoku, U. et al, 2006b]. When the cross section of
sintering area (as view from above toward the build platform) is bigger, the tensile
strength and modulus will be different from smaller cross section as EoV causes
greater sintering and increased densification. And in more prevalent in small feature
or crossection [Ajoku, U. et al, 2006b]. This behaviour occurs due to the build
orientation of the components and different temperature changes which occur within
the parts during the LS process.
However, as shown in Figure 6-36 for sintered material, it was confirmed from
the crystallinity characterisation results there was no significant difference for each
section thickness and solid for either non-heat treated or heat treated specimens. This
signifies that the crystallinity was not affected by the different section thickness or
solid. In contrast, the crystallinity became increasingly higher when each of individual
specimens was heat treated compared to non-heat treated specimens.
In addition, the peak melting temperature, as revealed in Figure 6-37 for the
sintered material, illustrated that for each different section thickness and solid, non-
heat treated or heat treated specimens, the peak melting temperature was significantly
different and the peak temperature of the heat treated specimens were higher
compared to non-heat treated specimens. This is possible, because of the variation in
crystallinity of each different section thickness and solid. The peak melting
temperature of heat treated specimens was higher due to the increase in crystallinity
when the specimens were heat treated.
Regarding the closed and open hollow structures, there was no significant
difference because in the un-sintered area from both specimens there were no melted
particles due to the design of the specimens. This means the original (input) powders
were still in the same particle condition (see Figure 6-18, morphology of particle),
even though the specimens were heat treated. In the SLS process, the laser will scan
the design area only (the outline parameter/profile of the parts), as a sintering process
to melt a powder particle when components or parts are produced.
Figures 6-11 to 6-13 show the results of the 2mm section thickness; the closed
and open hollow structure confirm that there was no impact, as there was no
significant difference of their mechanical properties for both results. It can be
described that the mechanical properties occurs as an effect of force on the sintered
area. In the closed hollow structure, the original powder material is enclosed in the
hollow area; it is not sintered. This indicates, that this material can be observed as the
original powder. Meanwhile, for the open hollow structure, the original powder was
also not sintered during the SLS process. This trapped powder, which was the original
powder that was removed prior to conducting the tensile test. Therefore, it can be seen
that the force on the sintered area was similar. As a result, with respect to the un-
sintered area, this area has become un-melted (particle) material. In addition, this un-
melted material does not fuse with the sintered material area; therefore, when the
reaction force occurred during the tensile test, there was no distribution of force on
this area.
In this research, the specimens that were used were round bars with hollow
structures, which confirmed that the larger sintered area gave higher tensile properties.
This case is different with the results demonstrated by Majewski and Hopkinson
[2011] that used dog bone as the specimens in their investigation. Their results
demonstrated that there was no influence of section thickness on the mechanical
properties [Majewski and Hopkinson, 2011].
However, from Majewski and Hopkinson [2011] and this particular research,
the results confirm that there were no effects on the tensile properties, even though
there was a small increase in crystallinity at the larger section thickness. This is
because slight increase in crystallinity is not enough to have a significant effect on
them.
6.7.2 Effect of heat treatment on properties
Heat treatment conducted in this research has an impact on mechanical properties,

given that it has more (enough) energy to melt the particles. In processing the SLS
material, the amount of heat melting was lower, and not enough to melt the particles
completely. However, the bonding of the particles improved when the heat energy is
increased, as it is being heat treated, mainly when treated in vacuum condition. This
condition was confirmed by the DSC traces with heat treated materials, in the vacuum
environment. In the air heat treatment, the amount of heat melting energy was slightly
lower than the amount of energy that occurs in the vacuum condition, and this
resulted in the improvements of the mechanical properties and the thermal
characteristic constantly increases. Therefore, in order to change the semi-crystalline
materials that are of high ordered molecular structure with sharp melt points into a
viscous liquid, a given quantity of heat need to be absorbed. As semi-crystalline
materials do not gradually soften with a temperature increase but rather remain hard
until the molten polymer flows in between the powder particles, forming sintering
necks, enough heat must be provided and maintained for complete layer is fully
molten and overlaps to the previous layer. Since the rate of crystallization is minimum
near Tg and Tm, and reaches a maximum between those two temperatures. SLS
process is something between partial and full melting. The melt viscosity is linearly
related to the molecular weight (MW). For good interfusion of the polymer chains,
sufficient to provide particle necking and layer to layer adhesion, a low melt viscosity
is desirable. However, low viscosity results in high shrinkage and poor part accuracy.
Thus an optimum MW range exists but is not easily controllable, neither at the
polymer production nor during aging in process.
The thermal degradation over time when re-using the same powder repeatedly,
results in the decay of the MFI (Mold Flow Index) hence a drop in powder flowability
and a rise in melt viscosity preventing proper sintering and a drop in mechanical
properties can be observed as well.
It was believed that the porosity occurred as there was not enough energy to
melt the particle structure. It was expected that the heat treatment can be used to
reduce the porosity of the specimen, in which the amount of particle structure melting
would increase.
The use of heat treatment as a processing of laser sintered material can reduce
the porosity and shrinkage problem. This is a consequence of the greater energy of
melting after heat treatment. The greater energy of melting results in better fusion of
the particles and builds a more compact structure. Therefore, the tensile strength and
tensile modulus improved in the heat treated specimens rather than the non-heat
treated specimens.
Basically, in polymers science, molecular weight is an essential factor to form
microstructure in order to bond effectively. In this situation, the molecular weight is
increased relatively by increasing the heat melting temperature. This DSC indicated
that the increasing heat melting temperature was enough to melt the polymer
microstructure. Due to this effect, the increasing molecule weight would also increase
the melting viscosity and density of polymer structures. This significant trend was
observed by the degree of crystallinity.
The DSC traces of each specimen in different fabrication shows a different heat
melting energy to melt the polymer particle. From these conditions it can be described
that the heat treatment methods can give different amount of energy, which was
caused by the different settings of temperature and time of treatment for each heat
treatment in air and in vacuum environment.
Theoretically, the values of crystallinity are influenced by the shape of
specimens, the position in which the material was cut and then placed in pan or
crucible [Craig and White, 2005]. In this research study, the material cutting was
taken from the centre of the tensile specimens and used it as the tensile specimen test.
Therefore, the values of crystallinity in this study are different with the values of
crystallinity that was studied by Zarringhalam [2007]. He reported that he used the
specimens which were built in X orientation, with flat shape and different settings of
building parameters. However, the result values, when compared are relatively the
same.
The results that were obtained in chapter 6 have different impact and
relationship on the properties after heat treatment compared to the results obtained in
chapter 4 and 5. The results show that on different section thickness 1mm, 2mm, 3mm
and solid, either for heat treated or non-heat treated specimens, the mechanical
properties increase while section thickness increases up to solid. In addition, in all
individual different section thickness and solid specimens, it can be confirmed that the
mechanical properties were higher after heat treatment. Moreover, the crystallinity of
all heat treated were higher than non-heat treated specimens.
Figures 6-36 and 6-37 shows that the average degree of crystallinity and
melting temperature for sintered material of heat treated specimens was higher
compared to non-heat treated specimens. This condition confirms that heat treatment
has effect on crystallinity and melting temperature.
Table 6.1 show the relationship of changes in mechanical property and changes
in degree of crystallinity after heat treated.
Table 6.1. Relationship of changes in mechanical property and changes in degree of

crystallinity after heat treated
Percentage Changes in Mechanical
Degree of Crystallinity
Property after heat treated (%)
Specimen Without Percentage
With Heat
Heat Changes Strength Modulus Elongation
treatment
treatment (%)
Section 1mm 25.65 29.49 14.97 32.40 26.40 0.21
Section 2mm 25.24 28.60 13.31 16.42 14.35 -5.87
Section 3mm 23.65 29.99 26.81 5.69 12.46 -29.17
Solid 24.75 27.30 10.30 10.73 9.04 -8.55
From Table 6.1, it was clear that when crystallinity was higher, tensile strength
and tensile modulus would increase. This is due to enhanced secondary inter-chain
bonding which results from adjacent aligned chain segments as percent crystallinity
increases. This enhanced inter-chain bonding inhibits relative inter-chain motion.
However, the elongation at break was lower or similar when crystallinity is
higher. The reduction in elongation is due the changes in the molecular structure of
the materials where disordered quasi-amorphous interlamellar (IL) regions between
the crystal lamellae were reduced. This is as confirmed by research study by Bessel et
al [1975], Majewski et al [2008] and Hopkinson et al [2009].
Heat treatment of semi-crystalline polymers can lead to an increase in the

percent crystallinity and in crystallite size and perfection, as well as to modification of
the spherulite structure. For undrawn materials that are subjected to constant time heat
treatments, increasing the annealing (heat treatment) temperature lead to the
following: (1) an increase in tensile modulus, (2) an increase in yield strength and (3)
reduction in ductility [Callister and Rethwisch, 2012].
For a specific polymer, the degree of crystallinity can have a significant
influence on the mechanical properties because it affects the extent of the
intermolecular secondary bonding. For crystalline regions in which molecular chains
are closely packed in an ordered and parallel arrangement, extensive secondary
bonding typically exists between adjacent chain segments. As a consequence, for
semi-crystalline polymers, tensile modulus increases significant with degree of
crystallinity [Callister and Rethwisch, 2012].
In addition, these results confirm that the heat treatment has impact on thermal
properties of Nylon 12 laser sintered material and consequently influenced their
mechanical properties.
From Figures 6-38 and 6-39 for un-sintered material, where raw particle
powder of Nylon 12 is trapped in a hollow structure of each individual 1mm, 2mm
and 3mm section thickness; the degree of crystallinity of non-heat treated and heat
treated did not significantly change at 95% confidence level, even though there was a
small increase in the heated specimens. However, the melting temperature
significantly changes at 95% confidence level. This may be because of the size of the
grain particles of the original raw material used were not from the same batch and not
originally from fresh or re-used or might be mixture of fresh and re-used material.
Therefore, there was no effect of heat treatment on un-sintered material and in this
case on raw powder material. Due to this, there was insufficient heat supplied on the
powder particle to form as to a sintered area that fused with laser sintered area.
The degree of crystallinity and melting temperature of sintered material were
always lower compared to un-sintered material, either for non-heat treated and heat
treated specimens. Figures 6-38 and 6-39, demonstrated that the average degree of
crystallinity was between 27.30% and 29.99% for sintered material, while for un-
sintered material it was between 39.17% and 40.73%. This result confirms that the
degree of crystallinity of sintered material were lower than degree of crystallinity of
un-sintered material, for each individual 1mm, 2mm and 3mm section thickness. This
is because a melted solid material has lower particle crystallisation during SLS
process.
In addition for specimens with heat treatment as shown in Figures 6-37 and 6-
39, the average melting temperature was between 186.69 oC to 193.24oC for sintered
material and between 192.48oC to 194.61oC for un-sintered material. This result
confirms that the melting temperature of sintered material were lower than melting
temperature of un-sintered material, for each individual 1mm, 2mm and 3mm section
thickness.
Furthermore, since different in melting temperature between all the difference
section thickness is so small; it shows that there was no effect of crystallinity on
melting temperature with and without heat treatment on un-sintered material.
However, it is difference for sintered material as the result confirms that there was
effect of crystallinity on melting temperature.
For sintered and un-sintered material, melting temperature of heat treated
specimens was higher compared to non-heat treated specimens.
6.7.3 Effect of Scanning Strategy on Mechanical Properties
This experimental study on the effect of different setting of scanning strategy on

mechanical properties Nylon 12 laser sintered material has been implemented. As
mentioned previously, the setting of laser power of outline scanning was set up
bigger, to observe if there was an impact.
The implementation demonstrated that there was an impact on mechanical
properties. It is believed that this is because there are the relationship between the
energy density and mechanical properties. Figure 6-41 to 6-43 shows the relationship
between average total energy density and average tensile strength, average tensile
modulus and average elongation at break.
Figure 6-41 Average energy density and average tensile strength relationship for
specimens with default fill scanning and fill and outline scanning, with 1mm, 2mm,
3mm section thickness and solid
Figure 6-42 Average total energy density and average tensile modulus relationship for
specimens with default fill scanning and fill and outline scanning, with 1mm, 2mm,
3mm section thickness and solid
Figure 6-43 Average total energy density and average elongation at break relationship
for specimens with default fill scanning and filling and outline scanning, with 1mm,
2mm, 3mm section thickness and solid
This was also confirmed by the research of others, which shows that when
energy density is higher the mechanical properties of laser sintered material are also
higher [Tontowi and Childs, 2001; Gibson and Shi, 1997, Caulfield, et al, 2007].
It is clear that when the energy density increases, this means that the number of
particles melted is increasing and the particle bonding is stronger. This relation is
higher when it comes to mechanical properties.
The implementation of bigger laser power setting for outline scan made a
stronger bonding of the particle that melted. With higher laser energy, particles tend
to coalesce and had larger necking area between particles. Similar finding by
Salmoria [2014] was also reported on polymer sintering of PC using different laser
energy and particles size. Furthermore, when the particles were laser sintered once
and then it was laser scanned again resulting in an overlap of scanning, the heat
dispersed in the ‘−Z’ direction and the X-Y plane were almost uniform and
instantaneously heated up the successive slice layer thickness [Carter, et al, 2014].
That is why the specimens with 1mm section thickness make bigger improvements
compared to section 2mm and 3mm.
It is clear that when the length of scanning path increases, this means that the
number of particles melted is increasing and the particle bonding is stronger. This
relation is higher when it comes to mechanical properties.
CHAPTER 7. CONCLUSIONS AND FUTURE WORK
7.1 Summary of Findings and Conclusion on Practical implications of the

work
The summary and conclusion of the works that were performed in this research are as
outline in section 7.1.1 and section 7.1.2.
7.1.1 Summary of Findings
From the present research findings, the following conclusions can be drawn:
A. Material characterisation
 Nylon 12 laser sintered components can be considered as anisotropic material.
 The variation of the test results from those of other researchers indicates
inconsistent characteristics of materials, even though the same material and
same machine type was used.
 Tensile strength and tensile modulus are higher when it is heat treated
compared to non-heat treated for each individual different cross-section of
thickness and solid, while the average degree of crystallinity were
approximately similar for each individual section thickness and solid, heat
treated and non-heat treated specimens. The larger sintered area gives higher
mechanical properties, in which the solid specimens were the highest when it
comes to tensile strength and tensile modulus. Crystallinity is higher when
heat treated compared to non-heat treated specimens. However, when
compared to each individual section thickness and solid specimens, the
crystallinity is approximately the same for all non-heat treated and heat treated
specimens.
 Compared to the default setting strategy, better mechanical properties were
obtained when different (bigger) setting of laser power was applied to outline
the contour of scanning in the fill and outline scanning strategy implemented.
 The increase of tensile strength, tensile modulus and crystallinity of 1mm
section thickness was higher when heat treated compared to the other different
section thickness even when heat treated or non-heat treated. This is maybe
because more particles melt during sintering process due to the fact that it
Chapter 7. Conclusions and Future Work Page 173

occurs in a bigger sintered area. This condition is due to more availability of

energy to sinter the melting material rather than sintering a larger surface area.
This condition also appeared when fill and outline scanning strategy was
applied for 1mm section thickness, which also resulted in the in the increase of
their tensile strength and tensile modulus which were higher than the other
section thickness.
B. Heat treatment studies

 Heat treatment in air at short hour (2 hours) within the Tg and Tm temperature
range did not promote sufficient changes in crystallinity to significantly
change the mechanical properties. However, longer hours (100 hours) at
180OC of heat treatment in vacuum condition (as reported in chapter 6)
promote a slight difference in crystallinity, where the average degree of
crystallinity increased in the average of 11.63% significantly at a 95%
confidence level and melting peak for both different section thickness and
solid specimens.
 Crystallinity for heat treated specimens was higher compared to non-heat
treated specimens. However, compared to each individual different section
thickness and solid specimens, the crystallinity was approximately similar
within the non-heat treated specimens and the heat treated un-sintered
material.
 Increase in crystallinity from heat treatment influenced the mechanical
properties of tensile strength, tensile modulus and elongation at break.
C. Scanning strategy implementation

 Tensile properties increased when scanning strategy was implemented on
different section thickness and solid specimens (as reported in chapter 6). The
results showed that there was an improvement on their tensile strength and
tensile modulus. Compared to the initial specimens that was produced with
default fill scanning strategy and non-heat treated, at a 95% confidence level
tensile strength significantly increase in the average of 63.92%, 12.68%,
7.09% and 6.45% for specimens with 1mm, 2mm, 3mm section thickness and
solid respectively. In addition, average tensile modulus also increased

significantly at a 95% confidence level by 48.43%, 17.27%, 9.25% and 9.07%

for each individual specimen respectively.
7.1.2 Conclusion on Practical implications of the work
It might be considered that vacuum heat treatment could be adopted on an industrial

scale. The method is simple and only limited facilities are required. However, the
initial cost is relatively expensive, especially when a specific furnace or other
equipment is needed. In addition, the additional lead time might be considered too
long. As a result, such work requires more costly components and the time to market
can be longer. Even though the mechanical properties are better, the extra time
required does not justify the additional cost required.
Heat treatment in a vacuum environment can be a suitable option in order to
improve the mechanical properties of laser sintered Nylon 12 material. However,
when the component is fully sintered, the increase of properties is not sufficient
enough when heat treatment is used.
Another alternative proposed from this research work by alterations in different
setting of scanning strategy during building the specimens or the actual parts or
component to gain better mechanical properties. This approach is shorter and more
economical than using the heat treatment approach to gain better mechanical
properties. Furthermore, no additional equipment are required since in this setup the
existing SLS machine is used.
Table 7.1 summarises the results of tensile properties of Nylon 12 laser sintered
material, for closed hollow structure with different section thickness and solid, with
different build up scanning strategy, non-heat treated and heat treated specimens.

Table 7.1 Comparison of Mechanical Properties of Specimens with: default fill

scanning strategy and non-heat treated (A); default fill scanning strategy and heat
treated (B); fill and outline scanning strategy and non-heat treated (C)
Build up and Post processing Changes/Improvement
Treatment
Specimens A B C A to B A to C
(%) (%)
Section 1mm
 Strength (MPa) 19.57 25.91 32.08 32.40 63.92
 Modulus (MPa) 1049.62 1326.67 1557.92 26.40 48.43
 Elongation (%) 3.73 3.74 3.71 0.27 -0.54
Section 2mm
 Strength (MPa) 33.13 38.57 37.33 16.42 12.68
 Modulus (MPa) 1494.55 1709.06 1752.63 14.35 17.27
 Elongation (%) 4.57 4.30 4.77 -5.91 4.38
Section 3mm
 Strength (MPa) 41.63 44.00 44.58 5.69 7.09
 Modulus (MPa) 1718.88 1933.10 1877.90 12.46 9.25
 Elongation (%) 6.21 4.40 5.60 -29.15 -9.82
Solid
 Strength (MPa) 46.34 51.31 49.33 10.73 6.45
 Modulus (MPa) 1786.97 1948.60 1949.02 9.04 9.07
 Elongation (%) 12.77 11.68 9.70 -8.54 -24.04
Table 7.1 shows that the improvement and the changes of mechanical properties
by using outline scanning strategy was better than heat treatment and fill scanning
except for 1mm section thickness. However for both alternatives, when the
outline/cross-section thickness increases, the improvement and the changes tend to be
reduced. This is due to the fact that the amount of sintered material area had
increased. Hence, for solid part the improvement is less significant as compared to
thin hollowed cross-section.
But with scanning strategy there could be another problem that needs to be
anticipated regarding the nominal and dimensional accuracy. The surface roughness
can be impacted as suggested by other studies that scanning strategy can cause of
dimensional inaccuracy and surface roughness (Yusoff and Thomas, 2008). Due to the

limitation of time and budget, this was not investigated further nor observed in this
current study.
With heat treatment of 100 hours, the process is too long, defeating the concept
of rapid manufacturing (RM). As a result, the fill and outline scanning strategy
process can be used as an alternative to improve the mechanical properties. However,
further investigation is still needed to develop a more effective strategy in reality.
Another important issue to consider when the designer or user needs to confirm which
technology approach is more acceptable in order to obtain the right specimen and for
the process to be cost effective.
As a concluding remark, scanning strategy is more beneficial to be used on thin
outline/cross-section thickness and be taken into consideration during SLS process.
7.2 Achievement of objectives
The overall aim and objectives of this research is to achieve a greater understanding
of the factors that have effects on properties of Nylon 12 laser sintered material.
Moreover, from this understanding it can be used to control the behaviour of the parts
or component produced. This knowledge is useful for designer and other users in
order to achieve suitable performances of the parts and components.
This study was expected to achieve some objectives in particular to determine
the mechanical performance of one SLS material (Duraform PA), over a range of
stress states; to develop post processing methods for the material; to determine and
understand the impact that the post-processing methods have on the consistency of the
mechanical properties.
From chapter 4 it is clear that all the results obtained show that the mechanical
properties of Nylon 12 laser sintered material were varied and anisotropic. In addition,
as it was reported in chapter 6, their mechanical properties also changed when
different section thickness of specimens were used. Moreover, the implementation of
different setting of scanning strategy in manufacturing the parts shows different
mechanical properties compared to the default setting of scanning strategy. These
results confirm that first objective “to determine the mechanical performance of one
SLS material (Duraform PA), over a range of stress states” is achieved.
From chapter 5, it can be summarised that heat treatment did not show any
improvement on properties when the specimens were heat treated either in air or in a

vacuum environment, particularly when specimens were heat treated for 2 hours in air
and 16 hours in vacuum. However, as shown in chapter 6, there were some different
results obtained. For all specimens that were heat treated in vacuum for 100 hours, the
results confirm that there were effects of heat treatment on mechanical and thermal
properties on Nylon 12 laser sintered material. This means, the heat treatment can be
used as an alternative solution to improve their mechanical properties. All evidences
can be used as an approval for the second objective “to develop post processing
methods for the material”.
The third objective of this study is “to determine and understand the impact that
the post-processing methods have on the consistency of the mechanical properties”.
The impact of heat treatment shows that the crystallinity of specimens increased after
heat treated compared to non-heat treated specimens. During the sintering process in
SLS machine, the number of particle melted were still low, due to limited energy
density and sintering process occurred quickly, laser scanning or scanning process to
sinter the material done quickly. This mean that the amount of energy still insufficient
to complete to sinter the materials which better results. As a results the degree of
crystallinity were also low. However, after heat treated it was confirmed that the
degree of crystallinity increased. This mean that the melted particles after being heat
treated, increases compare to non-heat treated specimens. All of these conditions
confirm about the achievement of the third objective.
7.3 Research Contributions:
Very little information is available in published literature about mechanical properties

of Nylon 12 laser sintered material on hollow structure with different section
thickness and solid specimens. In addition, available information after different types
of heat treatment on this laser sintered material are also very limited, particularly heat
treatment in vacuum environment. Results from this study can fill some of this gap.
This experimental study clearly contributes to the novelty and originality.
Firstly, the mechanical properties of laser sintered product with different section
thickness of closed hollow structure and solid specimens create different result of
tensile strength and tensile modulus.

Secondly, identify the possibility of the use of heat treatment approach to

improve mechanical properties in vacuum environment. This study show that degree
of crystallinity can be increased by using heat treatment approach.
In addition, effects of scanning strategy in sample preparation were obtained
and showed different values of tensile strength and tensile modulus. Therefore,
another possibility to improve the mechanical properties is by using different fill and
scanning strategy when manufacturing the product. The results show that when the
product is manufactured using a combination of fill and outline scanning strategy, the
results were better.
From laser sintered product that was manufactured using a combination of fill
and outline scanning strategy, it is believed that the mechanical properties will be
higher and can be improved even more when also heat treated (with vacuum
environment).
7.4 Future work
From the research findings and conclusions described above, it is recommended that
further research work can be continued to provide more information and
understanding of the possible ways to improve and make more consistent
characteristics of laser sintered components. The following further research is
recommended:
1. From material characterisation (chapter 4), suggests that there is a need to

develop a standard method for measuring the mechanical properties of laser
sintered materials in order to provide greater consistency in methods between
researchers.
2. In the present research, heat treatment was conducted at 140oC for 120
minutes in the air environment and at 180oC for 960 minutes and 6000 minutes
in vacuum conditions at 1018mbar level. Future experimental work in
vacuum heat treatment should be conducted using the same temperature, but
for shorter periods. This future work can be used to observe or find the optimal
time to achieve the same degree of improvement in mechanical properties.

3. Another suggestion for heat treatment in vacuum would be to conduct heat

treatment with different cooling rates and different cooling media to observe
the effect on crystallinity and mechanical properties.
4. To investigate the effect of different setting of scanning strategy on degree of

crystallinity and how far their impact will be compared to when the default
setting of scanning strategy is used in producing the component or part.
Moreover, to observe the effect on surface roughness and accuracy of
dimensions.
5. Theoretically, in polymer science, crystallinity does not only depend on

crystallisation rate and molecular weight, but it also depends on the shape of
specimens parts [Craig and White, 2005] and also the cutting position of the
specimen. Therefore, further experiment using different shape of specimens
can be used to observe and to ensure if the outcome of SLS components is
different or same.
6. Another potential idea, the experiment can be conducted by using chemical

solution / chemical solvent to infiltrate the specimen. It would be useful to use
solvent based chemical, such as Buehler Epo-kwick fast cure epoxy (Epoxy
base solvent) or Polyurethane in which the specimens will soak in for limited
duration of time before being heat treated in vacuum condition.

LIST OF PUBLICATION RELATED THIS STUDY:
 Ahmad Kamil, K.W. Dalgarno, Material Model for Design Analysis of

Laser Sintered Products, Proceedings of the Eighth National Conference
on Rapid Design, Prototyping and Manufacture, Buckhinghamshire,
United Kingdom, 15 June 2007.
 Ahmad Kamil, K.W. Dalgarno, The Effects of Heat Treatment on the

Mechanical Properties of Laser Sintered Product Material, Proceeding of
The 32nd Japan Rapid Prototyping Symposium, Tokyo, Japan, 23-24 June
2008.

REFERENCES
3D Systems (2006a), Vanguard SLS systems, available online from:

http:/www.3dsystems.com/products/sls/vanguard/index.asp [Accessed April
2006]
3D Systems (2006b), SLS materials, available online from:

http:/www.3dsystems.com/products/solidimaging/lasersintering/datasheets.asp
[Accessed April 2006]
3D Systems (2010b), SLS materials, available online from:

http://www.3dsystems.com/products/materials/sls/index.asp [Accessed
February 2010]
Agarwala, Mukesh, Bourell, David, Beaman, Joseph, Marcus, Harris, Barlow, Joel.
(1995). "Direct selective laser sintering of metals." Rapid Prototyping Journal
1(1): 26-36.
Ajoku, Uzoma; Hopkinson, Neil; Caine, Mike. (2006a). “Experimental measurement

and finite element modelling of the compressive properties of laser sintered
Nylon-12”. Material Science and Engineering A 428: 211-216.
Ajoku, U., Saleh, N., Hopkinson, N., Hague, R., Erasenthiran, P. (2006b).
"Investigating mechanical anisotropy and end-of-vector effect in laser-sintered
nylon parts." Proc. IMechE, Part B: J. Engineering Manufacture 220: 1077-
1086.
Anderson, J. C., Leaver, K. D., Leevers, P., Rawlings, R.D. (2003). Materials Science
for Engineers, 5th edition. Cheltenham, Nelson Thornes Ltd.
ASTM Designation: E 9-89a: 2000 (2000). Standard Test Methods of Compression

Testing of Metallic Materials at Room Temperature. ASTM International,
West Conshohocken, USA.
ASTM Designation: E111-97 (2003). Standard Test Methods for Young’s Modulus,
Tangent Modulus, and Chord Modulus. ASTM International, West
Conshohocken, USA.
References Page 182

ASTM Designation D6272-02 (2004). Standard Test Methods for Flexural Properties
of Unreinforced Plastics and Electrical Insulating Materials by Four-Point
Bending. ASTM International, West Conshohocken, USA.
Baker, Anne Marie M., Mead, Joey (2002). "Thermoplastics" in Handbook of

Plastics, Elastomers and Composites, fourth edition, editor in chief: Harper, C.
A., New York, McGraw-Hill.
Beaman, Joseph J., Barlow, Joel W., Bourell, David L., Crawford, Richard H.,
Marcus, Harris, McAlea, Kevin P. (1997). Solid freeform fabrication: A new
direction in manufacturing. Dordrecht, London: Kluwer Academic Publisher.
Bessell, T. J., Hull, D., Shortall, J.B. (1975). "The effect of polymerization conditions
and crystallinity on the mechanical properties and fracture of spherulitic nylon
6." Journal of Materials Science 10: 1127-1136.
Bo, Qian, Yu-sheng, Shi, Qing-song, Wei, Hai-bo,Wang (2012). “The helix scan
strategy applied to the selective laser melting”. Int. Journal Advanced
Manufacturing Technology, 63: 631-640
British Standards Institution. (1996). BS EN ISO 527-1:1996. Plastics. Determination

of tensile properties. General Principles. London, BSI.
British Standards Institution. (1998). BS EN ISO 14125:1998. Fibre-reinforced plastic

composites. Determination of flexural properties. London, BSI.
British Standards Institution. (2005). BS EN ISO 15310:2005, amendment no.1 to BS

ISO 15310:1999. Reinforced plastic. Determination of the in-plane shear
modulus by the plate twist method. London, BSI.
Brydson, John A. (1999). Plastics Materials, seventh edition. Oxford, Auckland,

Boston, Johannesburg, Melbourne, New Delhi, Butterworth-Heinemann.
Bugeda, Gabriel; Cervera, Miguel; Lombera, Guillermo. (1999). “Numerical

prediction of temperature and density distributions in selective laser sintering
processing”. Rapid Prototyping Journal 5(1): 21-26
References Page 183

Callister, W.D. and Rethwisch, D.G. (2012). Fundamental of Materials Science and
Engineering: An Integrated Approach. 4th Ed., p290-293, John Wiley & Sons,
New Jersey.
Carter, Luke N.; Martin, Christopher; Withers, Philip J.; Attallah, Moataz M. (2014).
“The influence of the laser scan strategy on grain structure and cracking
behaviour in SLM powder-bed fabricated nickel superalloy”. Journal of Alloys
and Compounds 615: 338- 347.
Castle Island’s (2010), Worldwide Guide to Rapid Prototyping, available online from:
http://www.additive3d.com [Accessed February 2010]
Caulfield, B., McHugh, P.E., Lohfeld, S. (2007). "Dependence of mechanical

properties of polyamide components on build parameters in the SLS process."
Journal of Materials Processing Technology 182: 477-488.
Cee Kai, Chua, K. F. Leong, C.S., Lim. (2003). Rapid Prototyping: Principles and
Applications. New Jersey, World Scientific.
Chang, Jin Hae, Chen, Michael J., Farris, Richard J. (1998). "Effect of heat treatment
on the thermal and mechanical properties of a precursor polymer:
polyhydroxyamide." Polymer 39(23): 5649-5654.
Craig, I.H and White, J.R. (2005). “Crystallization and Chemi-Crystallization of

Recycled Photodegraded Polyethylenes”. Polymer Engineering and Science,
45 (4): 588-595.
Dalgarno, K. W., Stewart, T.D. (2001). "Manufacture of production injection mould

tooling incoporating conformal cooling channels via indirect selective laser
sintering." Proc. Instn Mech Engrs, Part B 215: 1323-1332.
Dalgarno, K. W., Wright, C.S. (2003). "Approaches to Processing Metals and

Ceramics Through the Laser Scanning of Powder Beds." Society of
Manufacturing Engineers (Technical Paper).
References Page 184

Dutta, Debasish, Prinz, Fritz B., Rosen, David, Weiss, Lee. (2001). "Layered
Manufacturing: Current Status and Future Trends." Journal of Computing and
Information Science in Engineering 1(March): 60-71.
Faes, Matthias, Wang, Ya, Lava, P., Moens, David (2015).”Variability in the
mechanical properties of laser sintered PA-12 components”. Proceedings of
the 26th Annual International Solid Freeform Fabrication Symposium, Austin,
Texas, USA, 10-12 August 2015: pp 847-856
Gibson, Ian (2002). Software solutions for rapid prototyping. London, Professional
Engineering Pub.
Gibson, Ian, Shi, Dongping (1997). "Material properties and fabrication parameters in
selective laser sintering process." Rapid Prototyping Journal 3(4): 129-136.
Gogolewski, S. (1979). "Effect of annealing on thermal properties and crystalline

structure of polyamides. Nylon 11(polyundecaneamide)." Colloid & Polymer
Science 257: 811-819.
Gogolewski, S., Czerniawska, K., Gasiorek, M. (1980). "Effect of annealing on

thermal properties and crystalline structure of polyamides. Nylon 12
(polylaurolactam)." Colloid & Polymer Science 258: 1130-1136.
Gornet, T.J, Davis, K.R, Starr, TL, Mulloy, K.M (2002).”Characterisation of selective
laser sintering materials to determine proses stability”. Solid Freeform
Fabrication Symposium Proceedings, pp 546-553.
Griskey, Richard G. (1995). Polymer Process Engineering. New York, USA,

Chapman and Hall.
Hague, Richard, Mansour, Saeed., Saleh, Naguib. (2003a). "Design opportunities with
rapid manufacturing" Assembly Automation 23(4): 346-356.
Hague, R., Campbell, I., Dickens, P. (2003b). "Implications on Design of rapid

manufacturing" Special issue paper, Proc. Instn. Mech.Engrs. Part C: Journal
Mechanical Engineer Science 217: 25-30.
References Page 185

Hague, R., Mansour, S., Saleh, N. (2004). "Material and design considerations for
Rapid Manufacturing." International Journal of Production Research 42(22):
4691-4708.
Hague, R., (2006). "Unlocking the Design Potential of Rapid Manufacturing" in

Rapid manufacturing : an industrial revolution for the digital age, editors:
Hopkinson, N., Hague R. J. M., Dickens, P.M., Chichester, England, John
Wiley.
Hitt, D.J., B. Haworth, and N. Hopkinson (2011)”Fracture mechanics approach to

compare laser sintered parts and Injection mouldings of Nylon -12. Proc.
IMechE Part B: J. Engineering Manufacture, 225: 1663-1672.
Ho, H. C. H., Gibson, I., Cheung, W.L. (1999). "Effects of energy density on
morphology and properties of selective laser sintered polycarbonate." Journal
of Materials Processing Technology 89-90: 204-210.
Hopkinson, N. and P. Dickens (2001). "Rapid prototyping for direct manufacture."

Rapid Prototyping Journal 7(4): 197-202.
Hopkinson, N. and P. Dickens (2003). "Analysis of rapid manufacturing-using layer

manufacturing process for production" Proc. IMechE, Part C, Journal of
Mechanical Engineering Science 217: 31-39.
Hopkinson, N., Hague, R., Dickens, P. (2006). "Introduction to Rapid Manufacturing"

in Rapid manufacturing : an industrial revolution for the digital age, editors:
Hopkinson, N., Hague R. J. M., Dickens, P.M., Chichester, England, John
Wiley.
Hopkinson, N., Majewski, C.E., Zarringhalam, H. (2009). "Quantifying the degree of

particle melt in Selective Laser Sintering" CIRP Annals-Manufacturing
Technology 58: 197-200
Hou, Xiaonan (2005). Material models for FEA of SLS component. MPhil Thesis,
University of Leeds.
References Page 186

Kalpakjian, Serope, Schmid, Steven R. (2006). Manufacturing Engineering and

technology, 5th edition. London, Pearson Prentice Hall
Kim, J., Creasy, T.S. (2004). "Selective laser sintering characteristics of nylon 6/clay
re-inforced nanocomposite" Polymer Testing 23: 629-636
Kohan, Melvin I. (1995) Nylon Plastics Handbook. Munich, Vienna, New York,
Cincinnati, Hanser Publishers Hanser/Gardner Publications, Inc.
Kolosov, S.; Vansteenkiste, G.; Boudeau, N.; Gelin, J.C.; Boillat, E,. (2006).
“Homogeneity aspects in selective laser sintering (SLS)”. Journal of Materials
Processing Technology 177: 348-351.
Kridli, Ghassan T. (2006). "Material properties and characterization" in Rapid

prototyping: Theory and practice, edited by Kamrani, Ali, Nasr, Emad
Abouel, New York, N.Y., Springer.
Kruth, J. P., Leu, M.C., Nakawa, T. (1998). "Progress in Additive Manufacturing and
Rapid Prototyping." CIRP Annals - Manufacturing Technology 47(2): 525-
540.
Kruth, J. P., Mercelis, P., Froyen, L., Rombouts, Marleen. (2005a). "Binding
Mechanisms in Selective Laser Sintering and Selective Laser Melting". Rapid
Prototyping Journal 11(1): 26-36.
Kruth, J. P., Wang, X., Laoui, T., Froyen, L. (2003). "Lasers and materials in selective
laser sintering." Assembly Automation 23(4): 357-371.
Kruth, J. P., Vandenbroucke, B., Vaerenbergh, J.Van, Mercelis, P. (2005b).

"Benchmarking of different SLS/SLM processes as rapid manufacturing
techniques". International Conference Polymers and Moulds Innovations
(PMI). Gent, Belgium.
Levy, G. N., Schindel, R. (2002). "Overview of layer manufacturing technologies,

opportunities, options and applications for rapid tooling". Proc. Instn. Mech.
Engrs Part B: Journal Engineering Manufacture, 216, 1621-1633.
References Page 187

Levy, G. N., Schindel, Ralf, Kruth, J.P. (2003). "Rapid manufacturing and rapid
tooling with layer manufacturing (LM) technologies, state of the art and future
perspectives." CIRP Annals 52(2): 589-609.
Ma, Liang; Bin, Hongzan. (2007). “Temperature and stress analysis and simulation in
fractal scanning-based laser sintering”. Int J Adv Manuf Technol. 34: 898-903.
Majewski, Candice; Hopkinson, Neil. (2011). “Effect of section thickness and build
orientation on tensile properties and material characteristics of laser sintered
nylon-12 parts”. Rapid Prototyping Journal 17(3): 176-180.
Majewski, C.E.; Zarringhalam, H.; Hopkinson, N. (2008). “Effects of the degree of

particle melt and crystallinity in SLS Nylon-12 parts”. Proceedings of the 19th
Annual International Solid Freeform Fabrication Symposium, Austin, Texas,
USA, 2008: pp 45-54.
Mansour, S., Hague, R. (2003). "Impact of rapid manufacturing on design for

manufacture for injection moulding" Proc. Instn. Mech.Engrs. Part B: Journal
Engineering Manufacture 217: 453-461.
Mateva, R., Delev, O., Rousseva, S. (1997). "Structure of poly-ω-dodecalactam

obtained in bulk by an anionic mechanism." European Polymer Journal 33(8):
1377-1382.
Narazaki, Michiharu and Totten, George E. (2007). Distortion of Heat-treated

Components in Steel Heat Treatment, Metallurgy and Technologies
(Handbook), Editor: Totten, George E, Second Edition, Boca Raton, USA
CRC Press, Taylor and Francis Group.
Nelson, J.A, Galloway, G, Rennie, A. E. W., Abram, T. N (2014). “Effect of scan

direction and orientation on mechanical properties of laser sintered polyamide-
12”. Int. Journal Advanced Design and Manufacturing Technology, Vol.
7/No.3: 19-25
Nelson, J.C. (1993). Selective laser sintering: A definition of the process and a
empirical sintering model. PhD Thesis, University of Texas at Austin, USA.
References Page 188

Noorani, Rafiq. (2006). Rapid Prototyping: Principles and Applications. Los Angeles,
CA, John Wiley & Sons, Inc.
Ostertagová, Eva, Ostertag, Oskar, (2015), Methodology and Application of Oneway

ANOVA, http://pubs.sciepub.com/ajme/1/7/21/ (Accessed at 17.00 on 29 Sept
2015)
Pham, D. T. and Dimov, S. S. (2001). Rapid manufacturing : the technologies and

applications of rapid prototyping and rapid tooling. London, Berlin,
Heidelberg, Springer.
Pham, D. T., Ji, C., Dimov, S.S., (2004). "Layer manufacturing technologies". Proc.of
the 1st International Conference on New Forming Technology (ICNFT),
Harbin, China, 317-324.
Pham, D. T., Dotchev, K.D., Yusoff, W.A., (2008a). "Deterioration of polyamide

powder properties in the laser sintering process." Proc. IMechE, Part C,
Journal of Mechanical Engineering Science 222: 2163-2176.
Pham, D. T.; Dotchev, K.D.; Yusoff, W.A. (2008b). "Improvement of part surface
finishing in laser sintering by experimental design optimisation (DOE)."
Conference.iproms.org/sites/conference.iproms/org/files/papers2007/112.pdf.
accessed: 9/9/2011, at 12.40
Ramazani SA, A. and Mousavi S, S.A. (2005). "Investigation of vacuum annealing

effects on physical-mechanical properties of thermoplastic parts" Materials
and Design 26: 89-93.
Rosenzweig, N. and Narkis, M. (1981). “Dimensional variations of two spherical

polymeric particles during sintering”. Polymer Engineering and Science, 21
(10): pp582-585
Rusenberg, S.; Schmid. H.J. (2014). “Advanced characterization method of nylon 12

materials for application in laser sinter processing” AIP conference
Proceedings 1593: 713-718. DOI: 10.1063/1.4873877
Rusenberg, S, Josupeit, S and Schmid, H. (2014), “A method to characterize the

Quality of a Polymer Laser Sintering Process”, in Advances in Mechanical
References Page 189
Engineering, p7-10, Hindawi Publishing Corporation,(http://

dx.doi.org/10.1155/2014/640496)
Salmoria, G.V, Hotza, D, Klauss, P, Kanis, L.A, and Roesler, C.R.M. (2014),
“Manufacturing of porous polycaprolactone prepared with different particle
sizes and infrared laser sintering conditions: Microstructure and Mechanical
Properties”, in Advances in Mechanical Engineering, pages 7, Hindawi
Publishing Corporation, 2014. (http:// dx.doi.org/10.1155/2014/640496)
Scobbo Jr, J. J. and Hwang, C.R. (1994). "Annealing Effects in Poly(Phenylene

Sulfide) as Observed by Dynamic Mechanical Analysis." Polymer Engineering
and Science 34(23).
Shi, Y., Chen, J., Wang, Y., Li, Z., Huang, S. (2007). "Study of selective laser
sintering of polycarbonate and postprocess for parts reinforcement." Proc.
IMechE, Part L, Journal of Materials: Design and Applications 22: 37-42.
Shi, Y., Li, Z., Sun, H., Huang, S., Zeng, F. (2004). "Effect of the properties of the
polymer materials on the quality of selective laser sintering parts." Proc. Instn.
Mech. Engnr Part L, Journal of Materials: Design and Applications 218(3):
247-252.
Stamp, R.; Fox, P.; O’neill, W.; Jones, E.; Sutcliffe, C. (2009). “The development of a
scanning strategy for the manufacture of porous biomaterials by selective laser
melting”. J Mater Sci: Mater Med 20: 1839- 1848.
Technopro, Energy (2009). Vacuum Furnace, available online from:

http:/www.energytechpro.com/Demo-IC/Gas_Technology/
Heat_Treat_Vacuum.htm [Accessed 24/08/2009]
Tolochko, N.K., Khlopkov, Y.V., Mozzharov, S.E., Ignatiev, M.B., Laoui, T., Titov,
V.I. (2000). "Absorptance of powder materials suitable for laser sintering."
Rapid Prototyping Journal 6(3): 155-161.
Tontowi, Alva. E., Childs, T. H. C. (2001). "Density prediction of crystalline polymer

sintered parts at various powder bed temperatures" Rapid Prototyping Journal
7(3): 180-184.
References Page 190

Totten, George E. (2006). Steel Heat Treatment, Equipment and Process Design. Boca
Raton, USA, FL: Taylor and Francis.
Tregub, Alexander, Harel, Hannah, Marom, Gad (1993). "The influence of thermal
history on the mechanical properties of poly(ether ether ketone) matrix
composite materials." Composites Science and Technology 48: 185-190.
Upcraft, S. and R. Fletcher (2003). "The rapid protoyping technologies." Assembly

Automation 23(4): 318-330.
Venuvinod, Patri K., Ma, Weiyin (2004). Rapid Prototyping: Laser-based and other
technologies. Norwell, Kluwer Academic Publishers.
Wikipedia, (2009a). Rapid manufacturing, definition, available online:

http://en.wikipedia.org/wiki/Rapid_manufacturing. [Accessed July, 2009].
Wikipedia. (2009b). Selective laser sintering, definition, available online:

http://en.wikipedia.org/wiki/Selective_laser_sintering [Accessed July, 2009].
Williams, A.N. (1994). Schaum’s outline of theory and problems of strength of

materials, Third Edition, McGraw-Hill, New York, USA.
Williams, John D.; Deckard, Carl R. (1998). “Advances in modelling the effects of
selected parameters on the SLS process”. Rapid Prototyping Journal 4(2): 90-
100.
Wohlers, T.T. (2006), Wohler Reports 2006: Rapid Prototyping and Manufacturing
State of the Industry, Fort Collins, Co, USA: Wohlers Associates.
Wunderlich, Bernhard. (1976). Macromolecular Physics: Crystal Nucleation, Growth,

Annealing. New York, San Fransisco, London, Academic Press.
Wunderlich, Bernhard (1980). Macromolecular Physics: Crystal Melting. New York,

London, Toronto, Sydney, San Fransisco, Academic Press.
Yadroitsev, I.; Bertrand, Ph.; Smurov, I. (2007). “Parametric analysis of the selective
laser melting process”. Applied Surface Science 253: 8064- 8069.
References Page 191

Yang, J., Bin, H., Zhang, X, Liu, Z., (2003). “Fractal scanning path generation and
control system for selective laser sintering (SLS)”. International Journal of
Machine Tools and Manufacture 43: 293-300.
Young, R. J. (1983). Introduction to polymers. London ; New York, Chapman and

Hall.
Yubero, Victor (2004). Development of material model for selective laser sintering
components. Final report: MECH 3810 Individual Project, University of
Leeds.
Yusoff, W.A.Y.; Thomas, A.J. (2008). “The effect of employing an effective laser
sintering scanning strategy and energy density value on eliminating “orange
peel” on a selective laser sintered part”. International Association for
Management of Technology. IAMOT 2008 Proceeding
Zarringhalam, H. and N. Hopkinson (2003). “Post-Processing of Duraform TM Parts

for Rapid Manufacture”. Solid Freeform Fabrication Symposium, Austin,
Texas, USA.
Zarringhalam, H., Hopkinson, N., Kamperman, N.F., Vlieger, J.J. de. (2006). "Effects
of processing on microstructure and properties of SLS Nylon 12." Materials
Science and Engineering A 435-436: 172-180.
Zarringhalam, H. (2007). “Investigation into Crystallinity and Degree of Particle Melt

in Selective Laser Sintering”. Doctoral Thesis, Loughborough University.
Zarringhalam, H., Majewski, C., Hopkinson, Neil. (2009). "Degree of particle melt in
Nylon-12 selective laser-sintered parts." Rapid Prototyping Journal 15(2):
126-132.
References Page 192

A: SHEAR AND BENDING TEST RESULTS

Three-point bending:
Four-point bending:
Appendix Page 193

Shear Test:
Appendix Page 194

B: ANOVA RESULTS
CHAPTER 4
A. Tensile Strength
One-way ANOVA: TStrength - X, TStrength - Y, TStrength - Z
Analysis of Variance
Source DF Adj SS Adj MS F-Value P-Value

Factor 2 446.12 223.058 48.14 0.000
Error 14 64.87 4.633
Total 16 510.98
B. Tensile Modulus
One-way ANOVA: T Mod - X, T Mod - Y, T Mod - Z

Factor 2 76580 38290 19.14 0.000
Error 14 28005 2000
Total 16 104585
C. Tensile Elongation at Break

One-way ANOVA: T Elong - X, T Elong - Y, T Elong - Z

Factor 2 38.29 19.146 12.05 0.001
Error 14 22.25 1.589
Total 16 60.54
D. Compression Modulus
One-way ANOVA: C Mod - X, C Mod - Y, C Mod - Z

Factor 2 189800 94900 4.18 0.036
Error 15 340281 22685
Total 17 530081
Appendix Page 195

E. Shear Modulus
One-way ANOVA: Shear Mod - , Shear Mod - , Shear Mod - , Shear Mod - , ...

Factor 5 0.13166 0.026332 9.97 0.000
Error 30 0.07920 0.002640
Total 35 0.21087
F. 3Point Flexural Modulus

One-way ANOVA: 3 P Flex Mod, 3 P Flex Mod, 3 P Flex Mod, 3 P Flex Mod,

Factor 5 1200768 240154 16.16 0.000
Error 29 431093 14865
Total 34 1631862
G. 4Point Flexural Modulus

One-way ANOVA: 4 P Flex Mod, 4 P Flex Mod, 4 P Flex Mod, 4 P Flex Mod, ...

Factor 5 667683 133537 24.98 0.000
Error 30 160387 5346
Total 35 828071
Appendix Page 196

CHAPTER 5
A. Tensile Strength:
A.1. Heat treatment in Air, 140oC, 2 hours
One-way ANOVA: TX2, TXht-2
Source DF SS MS F P
Factor 1 18.3 18.3 0.68 0.430
Error 10 269.6 27.0
Total 11 287.9
One-way ANOVA: TY2, TYht-2
Source DF SS MS F P
Factor 1 7.21 7.21 1.92 0.196
Error 10 37.59 3.76
Total 11 44.80
One-way ANOVA: TZ2, TZht-2
Source DF SS MS F P
Factor 1 46.53 46.53 4.82 0.053
Error 10 96.49 9.65
Total 11 143.02
A.2. Heat treatment in Vacuum, 180oC, 100 hours
Source DF SS MS F P
Factor 1 108.8 108.8 3.38 0.096
Error 10 321.5 32.2
Total 11 430.3
Appendix Page 197

A.3. Heat treatment in Vacuum, 180oC, 16 hours
Source DF SS MS F P
Factor 1 4.54 4.54 2.22 0.167
Error 10 20.45 2.05
Total 11 25.00
B. Tensile Modulus:
B.1. Heat treatment in Air, 140oC, 2 hours
One-way ANOVA: MoX2, MoXht-2
Source DF SS MS F P
Factor 1 3560 3560 0.87 0.374
Error 10 41092 4109
Total 11 44652
One-way ANOVA: MoY2, MoYht-2
Source DF SS MS F P
Factor 1 3496 3496 2.16 0.172
Error 10 16182 1618
Total 11 19678
One-way ANOVA: MoZ2, MoZht-2
Source DF SS MS F P
Factor 1 15827 15827 14.38 0.004
Error 10 11007 1101
Total 11 26834
B.2. Heat treatment in Vacuum, 180oC, 100 hours
Source DF SS MS F P
Factor 1 154576 154576 11.91 0.006
Error 10 129818 12982
Total 11 284394
Appendix Page 198

B.3 Heat treatment in Vacuum, 180oC, 16 hours
Source DF SS MS F P
Factor 1 675 675 0.59 0.460
Error 10 11440 1144
Total 11 12115
Elongation at Break
C.3 Elongation at break, 140C, 2 hours
One-way ANOVA: Elong. X2, Elong.Xht-2
Source DF Seq SS Contribution Adj SS Adj MS F-Value P-Value
Factor 1 0.8533 3.22% 0.8533 0.8533 0.33 0.577
Error 10 25.6333 96.78% 25.6333 2.5633
Total 11 26.4867 100.00%
One-way ANOVA: Elong. Y2, Elong.Yht-2

Factor 1 0.3008 13.26% 0.3008 0.3008 1.53 0.245
Error 10 1.9683 86.74% 1.9683 0.1968
Total 11 2.2692 100.00%
One-way ANOVA: Elong. Z2, Elong.Zht-2

Factor 1 0.7008 3.21% 0.7008 0.7008 0.33 0.577
Error 10 21.1217 96.79% 21.1217 2.1122
Total 11 21.8225 100.00%
Appendix Page 199

C.4 Elongation at break, 180C, 100 hours
One-way ANOVA: Elong. Y100, Elong.Yht-100
Factor 1 2.901 21.70% 2.901 2.901 2.77 0.127
Error 10 10.468 78.30% 10.468 1.047
Total 11 13.369 100.00%
C.5.Elongation at break, 180C, 16 hours
One-way ANOVA: Elong.Y16, Elong.Yht-16

Factor 1 6.163 16.31% 6.163 6.163 1.95 0.193
Error 10 31.637 83.69% 31.637 3.164
Total 11 37.800 100.00%
D. Crystallinity:
D.1. Heat treatment in Air, 140oC, 2 hours
One-way ANOVA: CX-08, CX140-120-08
Source DF SS MS F P
Factor 1 1.307 1.307 1.68 0.224
Error 10 7.763 0.776
Total 11 9.069
One-way ANOVA: CY-08, CY140-120-08
Source DF SS MS F P
Factor 1 2.55 2.55 1.37 0.269
Error 10 18.59 1.86
Total 11 21.14
One-way ANOVA: CZ-08, CZ140-120-08
Source DF SS MS F P
Factor 1 12.81 12.81 3.60 0.087
Error 10 35.61 3.56
Total 11 48.42
Appendix Page 200

D.2. Heat treatment in Vacuum, 180oC, 100 hours
One-way ANOVA: CY100, CYht-100
Source DF SS MS F P
Factor 1 7.21 7.21 2.17 0.172
Error 10 33.24 3.32
Total 11 40.45
D.3. Heat treatment in Vacuum, 180oC, 16 hours
One-way ANOVA: CY16, CYht-16
Source DF SS MS F P
Factor 1 14.02 14.02 13.80 0.004
Error 10 10.16 1.02
Total 11 24.17
E. Melting Peak Temperature:
E.1. Heat treatment in Air, 140oC, 2 hours
One-way ANOVA: MX-08, MX140-120-08
Source DF SS MS F P
Factor 1 1.300 1.300 1.55 0.241
Error 10 8.382 0.838
Total 11 9.682
One-way ANOVA: MY-08, MY140-120-08
Source DF SS MS F P
Factor 1 0.051 0.051 0.09 0.765
Error 10 5.379 0.538
Total 11 5.430
One-way ANOVA: MZ-08, MZ140-120-08
Source DF SS MS F P
Factor 1 1.96 1.96 1.69 0.223
Error 10 11.60 1.16
Total 11 13.56
Appendix Page 201

E.2. Heat treatment in Vacuum, 180oC, 100 hours
One-way ANOVA: MY100, MYht-100
Source DF SS MS F P
Factor 1 31.98 31.98 24.54 0.001
Error 10 13.03 1.30
Total 11 45.01
E.3. Heat treatment in Vacuum, 180oC, 16 hours
One-way ANOVA: MY16, MYht-16
Source DF SS MS F P
Factor 1 3.070 3.070 13.20 0.005
Error 10 2.327 0.233
Total 11 5.397
Appendix Page 202

CHAPTER 6
A. Effect of section thickness on Tensile Strength

One-way ANOVA: 1mm (w/o HT), 2mm (w/o HT), 3mm (w/o HT), Solid (w/o HT)
Source DF SS MS F P
Factor 3 2229.92 743.31 236.75 0.000
Error 19 59.65 3.14
Total 22 2289.57
B. Effect of heat treatment on Tensile Strength

One-way ANOVA: 1mm (w/o HT), 1mm (HT), 2mm (w/o HT), 2mm (HT), ...
Source DF SS MS F P
Factor 7 4493.59 641.94 308.39 0.000
Error 38 79.10 2.08
Total 45 4572.69
A. Effect of section thickness on Tensile Modulus

One-way ANOVA: 1mm (w/o HT)_1, 2mm (w/o HT)_1, 3mm (w/o HT)_1, Solid (w/o HT)_1
Source DF SS MS F P
Factor 3 1771458 590486 192.62 0.000
Error 19 58245 3066
Total 22 1829704
B. Effect of heat treatment on Tensile Modulus

One-way ANOVA: 1mm (w/o HT), 1mm (HT)_1, 2mm (w/o HT), 2mm (HT)_1, ...
Source DF SS MS F P
Factor 7 3732995 533285 150.87 0.000
Error 38 134320 3535
Total 45 3867316
A. Effect of section thickness on Crystallinity (sintered)

One-way ANOVA: 1mm (w/o HT), 2mm (w/o HT), 3mm (w/o HT), Solid (w/o HT)
Source DF SS MS F P
Factor 3 13.40 4.47 2.41 0.097
Error 20 37.09 1.85
Total 23 50.50
One-way ANOVA: 1mm (HT), 2mm (HT), 3mm (HT), Solid (HT)
Source DF SS MS F P
Factor 3 25.01 8.34 3.86 0.025
Error 20 43.18 2.16
Total 23 68.19
Appendix Page 203

B. Effect of heat treatment on Crystallinity (sintered)

One-way ANOVA: 1mm (w/o HT), 1mm (HT), 2mm (w/o HT), 2mm (HT), ...
Source DF SS MS F P
Factor 7 232.46 33.21 16.55 0.000
Error 40 80.27 2.01
Total 47 312.73
A. Effect of section thickness on Crystallinity (un-sintered)

One-way ANOVA: 1mm (w/o HT)_1, 2mm (w/o HT)_1, 3mm (w/o HT)_1
Source DF SS MS F P
Factor 2 20.69 10.34 6.41 0.010
Error 15 24.21 1.61
Total 17 44.90
B. Effect of heat treatment on Crystallinity (un-sintered)

Source DF SS MS F P
Factor 5 54.66 10.93 4.40 0.004
Error 29 72.07 2.49
Total 34 126.73
A. Effect of section thickness on Melting Peak Temperature (sintered)

One-way ANOVA: 1mm (w/o HT)_2, 2mm (w/o HT)_2, 3mm (w/o HT)_2, Solid (w/o HT)_1
Source DF SS MS F P
Factor 3 65.289 21.763 33.12 0.000
Error 20 13.140 0.657
Total 23 78.429
One-way ANOVA: 1mm (HT)_2, 2mm (HT)_2, 3mm (HT)_2, Solid (HT)_1
Source DF SS MS F P
Factor 3 154.34 51.45 21.60 0.000
Error 20 47.63 2.38
Total 23 201.97
B. Effect of heat treatment on Melting Peak Temperature (sintered)

Source DF SS MS F P
Factor 7 761.73 108.82 71.63 0.000
Error 40 60.77 1.52
Total 47 822.50
Appendix Page 204

A. Effect of section thickness on Melting Peak Temperature (un-sintered)

One-way ANOVA: 1mm (w/o HT)_3, 2mm (w/o HT)_3, 3mm (w/o HT)_3
Source DF SS MS F P
Factor 2 0.092 0.046 0.21 0.812
Error 15 3.267 0.218
Total 17 3.359
One-way ANOVA: 1mm (HT)_3, 2mm (HT)_3, 3mm (HT)_3
Source DF SS MS F P
Factor 2 16.39 8.19 6.59 0.010
Error 14 17.41 1.24
Total 16 33.80
B. Effect of heat treatment on Melting Peak Temperature (un-sintered)

Source DF SS MS F P
Factor 5 140.752 28.150 39.48 0.000
Error 29 20.677 0.713
Total 34 161.429
Effect of open and closed hollow structure (2 mm) on Tensile Strength
One-way ANOVA: 2mm (powder), 2mm (w/o powder)
Source DF SS MS F P
Factor 1 15.05 15.05 3.45 0.093
Error 10 43.63 4.36
Total 11 58.68
Effect of open and closed hollow structure (2 mm) on Tensile Modulus
One-way ANOVA: 2mm (powder)_1, 2mm (w/o powder)_1

Source DF SS MS F P
Factor 1 40102 40102 7.22 0.023
Error 10 55539 5554
Total 11 95641
Appendix Page 205

C: DSC TEST RESULTS
êxo Hollow 1mm sintered HT 26.08.2011 15:11:36
10
mW
145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 210.0 °C
16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 min
Lab: METTLER STAR e SW 9.01
Appendix Page 206

êxo Hollow 1mm unsintered non-HT 26.08.2011 15:48:02
10
mW
140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 210.0 °C
16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 min
Appendix Page 207

Appendix Page 208

êxo Hollow 2mm unsintered non-HT 26.08.2011 15:49:38
10
mW
145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 210.0 °C
16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 min
Appendix Page 209

êxo Hollow 3mm sintered non-HT 26.08.2011 12:24:31
5
mW
145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 210.0 °C
16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 min
Appendix Page 210

êxo Hollow 3mm unsintered HT 26.08.2011 14:56:16

Crystallinity 41.61 %
Integral -606.96 mJ
Onset 184.90 °C
Peak 191.81 °C
Endset 197.62 °C
10
mW
145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 210.0 °C
16.5
Lab: METTLER 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 STAR e
22.5 23.0
SW 9.01 min
Appendix Page 211

êxo Solid sintered non-HT 26.08.2011 15:51:56
Integral -358.71 mJ
Onset 172.10 °C
Peak 181.62 °C
Endset 189.38 °C
5
mW
Curve: YS6, 30.06.2011 11:53:21

Integral -378.21 mJ
Onset 173.98 °C
Peak 182.02 °C
Endset 190.64 °C
140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 210.0 °C
Lab:16.0METTLER
16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 STAR e22.5SW 9.01
22.0 23.0 min
Appendix Page 212

êxo X-08-Non HT-Air 03.06.2011 15:45:10
Crystallinity
Crystallinity 26.32
27.18 %
%
Integral
Integral -385.66
-392.50 mJ
mJ
Onset
Onset 174.43
174.44 °C
°C
Peak
Peak 183.06
184.27 °C
°C
Endset
Endset 193.12
193.01 °C
°C
5
mW
Integral -362.67 mJ
Onset 174.56 °C
Crystallinity
Peak 27.33 %°C
183.63
Integral
Endset -400.46
192.30 °CmJ
Onset 174.79 °C
Peak 183.95 °C
Endset 192.07 °C
Integral
Crystallinity -399.71
27.62 %mJ
Onset
Integral 174.58 °C
-410.42 mJ
Peak
Onset 183.66 °C
174.57 °C
Endset
Peak 192.86 °C
183.65 °C
Endset 192.74 °C
130.0 135.0 140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 °C
15.0 15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 min
êxo X-140-120-08-HT-Air 03.06.2011 16:01:35

Integral -385.04 mJ
Onset 175.15 °C
Crystallinity
Peak 25.60
184.01%°C
Integral
Endset -375.06 mJ
192.45 °C
Onset 174.77 °C
Peak 182.52 °C
Endset 191.82 °C
5
mW
Integral
25.84 %mJ
Onset
Integral 174.72 °C
-378.57 mJ
Peak
Onset 184.08 °C
176.09 °C
Endset
Peak 192.58 °C
184.70 °C
Endset 194.89 °C
Crystallinity
Integral 25.94 %mJ
-379.08
Integral
Onset -380.08
175.11 °CmJ
Onset
Peak 177.35
184.54 °C
Peak
Endset 186.32
193.35 °C
Endset 193.30 °C
130.0 135.0 140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 °C
15.0 15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 min
Appendix Page 213

êxo Y-08-Non HT-Air 03.06.2011 15:15:53
Integral -405.78 mJ
Integral -360.79 mJ
Onset 174.46 °C
Onset 174.96 °C
Peak 182.51 °C
Peak 183.73 °C
Endset 192.06 °C
Endset 194.13 °C
5
mW
Integral
27.14 %mJ
Onset
Integral 174.81
-397.64°C
mJ
Peak
Onset 183.93 °C
174.81 °C
Endset
Peak 192.85 °C
183.28 °C
Endset
Crystallinity 193.50
26.66 %°C
Integral -390.56 mJ
Onset
Crystallinity 174.67%°C
28.08
Peak
Integral 183.72 °C
-417.34 mJ
Endset
Onset 192.99 °C
175.47
Peak 184.72 °C
Endset 193.72 °C
130.0 135.0 140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 °C
15.0 15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 min
êxo Y-140-120-08-HT-Air 03.06.2011 15:32:27
Integral
28.17 %mJ
Onset
Integral 174.92 °C
-412.69 mJ
Peak
Onset 182.95 °C
175.22
Endset
Peak 194.08 °C
184.62
Endset 194.00 °C
5
mW
Integral
27.87 %mJ
Onset
Crystallinity
Integral 174.74%°C
28.09
-408.29 mJ
Peak
Integral
Onset 183.15 °C
-405.70
174.91 °C
mJ
Endset
Onset
Peak 193.25 °C
174.97
183.60
Peak
Endset 183.64
193.36 °C
Endset 193.41 °C
Integral -378.76 mJ
Onset 175.24 °C
Peak 184.71 °C
Endset 194.03 °C
135.0 140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 °C
15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 min
Appendix Page 214

êxo Z-08-Non HT-Air 03.06.2011 16:24:41
Crystallinity 26.25
26.89 %
Integral -384.53
-393.95 mJ
Onset 176.33
174.51 °C
Peak 186.06
182.91 °C
Endset 194.94
191.80 °C
5
mW
28.57
Integral -370.41 mJ
-418.60
Onset 175.16 °C
174.63
Crystallinity 27.42
184.39%°C
Peak 183.75
Integral -407.45 mJ
192.85 °C
Endset 193.02
Onset 176.44 °C
Peak
25.43 %°C
Endset
Integral 195.12
-377.95°C
mJ
Onset 174.81 °C
Peak 183.63 °C
Endset 192.16 °C
130.0 135.0 140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 °C
15.0
Lab: METTLER 15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5
STAR e
22.0 22.5
SW 9.01 min
êxo Z-140-120-08-HT-Air 03.06.2011 16:57:33
Crystallinity
Integral 25.94 %mJ
-377.04
Integral
Onset -374.68°C
173.52 mJ
Onset
Peak 175.51 °C
182.84 °C
Peak
Endset 183.63 °C
192.23 °C
Endset 192.54 °C
5
mW
Crystallinity
25.83 %
Crystallinity
Integral 22.51 %mJ
-304.50
Integral -378.47 mJ
Integral
Onset -329.86
175.79 °CmJ
°C
Onset 174.66
Onset
Crystallinity
Peak 173.87
26.65
183.91% °C
°C
Peak 185.18 °C
Peak
Integral
Endset 182.92
-390.41 °C
192.33 °CmJ
°C
Endset
Endset 192.14
190.20
Onset 174.68 °C
Peak 183.14 °C
Endset 191.16 °C
135.0 140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 °C
15.5 16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 e 22.5 min
Lab: METTLER STAR SW 9.01
Appendix Page 215

êxo COMPARE Y...VA 15.04.2011 16:38:33
20
mW
155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 °C
17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 min
êxo COMPARE NON HT - 16HRS 04.01.2012 17:16:39

Crystallinity
Crystallinity 25.42
25.51
26.06 %%
Integral
Integral -377.82
-373.71
-376.42 mJ
mJ
Crystallinity
Onset 25.41 %
174.61
174.40
Onset 174.73 °C
°C
Integral
Peak -377.58 mJ
183.52
182.91
Peak 183.50 °C
°C
Onset
Endset 174.97 °C
191.04
191.70 °C
Endset
Peak 191.79
182.60 °C
°C
Endset 191.59 °C
Curve:
Curve:Y2A,
Y3A,13.04.2011
13.04.201111:31:49
13:00:47
Curve: Y6A, 14.04.2011 11:28:23

Curve: Y1A, 12.04.2011 16:10:54
Integral -356.27 mJ 5
Onset 174.70 °C mW
Peak 183.70 °C
Endset 191.24 °C
Integral -397.44 mJ
Onset 174.74 °C
Peak 183.87 °C
Endset 192.09 °C
140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 210.0 °C
16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 min
Appendix Page 216

êxo COMPARE HT 100 Hrs (A) 06.05.2011 15:08:31

Integral
29.91 %mJ
Onset
Integral
-441.36
29.46 %°C
mJ
Peak
Onset
Integral 189.10
-432.78°C
176.58 °CmJ
Endset
Peak
Onset 196.89
176.07°C
186.26 °C
°C
Endset
Peak 196.08
186.91 °C
°C
Endset 197.25 °C
5
mW
26.99
Integral -416.49 mJ
-396.00
Onset
Crystallinity 175.60%°C
175.48
30.00
Peak
Integral 187.53 °C
185.10
-441.36 mJ
Endset
Onset 195.79 °C
195.74
176.22
Peak 186.41 °C
Endset 195.57 °C
140.0 145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 210.0 215.0 °C
16.0 16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 23.5 min
êxo Crystallinity Non HT 100 hrs 04.05.2011 14:27:36
Crystallinity
Integral 25.93
27.94 %mJ
-418.47
Integral
Onset -382.05
-409.37 mJ
174.57 °C
Onset
Peak 175.14
183.07 °C
175.11 °C
Peak
Endset 183.41
192.63 °C
184.47 °C
Endset 192.94
193.15 °C
Integral -411.20 mJ
Onset 175.32 °C
Peak 184.76 °C 5
Crystallinity
Endset 193.41%°C mW
25.45
Integral -369.60 mJ
Crystallinity
Onset 24.47
174.72%°C
Integral
Peak -361.59 mJ
183.63 °C
Onset
Endset 173.12
193.36 °C
°C
Peak 182.38 °C
Endset 190.70 °C
145.0 150.0 155.0 160.0 165.0 170.0 175.0 180.0 185.0 190.0 195.0 200.0 205.0 210.0 215.0 °C
16.5 17.0 17.5 18.0 18.5 19.0 19.5 20.0 20.5 21.0 21.5 22.0 22.5 23.0 23.5 min
Appendix Page 217

D: DATA RESULTS FOR CH. 6
Non-heat treated with filling scanning:
Table 6-1. Tensile Strength Results of all Non-heat treated Specimens (MPa)
(Noted: *closed hollow and ** open hollow Structure)
Sample 95 %
Specimen Average Confidence
1 2 3 4 5 6
Interval
Section 1mm* 19.21 21.18 18.65 17.40 - 21.43 19.57 ±1.50
*
Section 2mm 33.97 29.62 32.48 33.87 33.64 35.18 33.13 ±1.54
Section 3mm* 39.12 44.05 42.23 41.50 39.61 43.25 41.63 ±1.57
Solid 46.15 46.33 47.16 47.60 47.17 43.65 46.34 ±1.14
**
Section 2mm 28.40 34.13 28.20 31.57 31.46 31.56 30.89 ±1.66
Table 6-2. Tensile Modulus Results of All Non-heat treated Specimens (MPa)
(Noted: *Closed hollow and **Open hollow Structure)
Sample Averag 95 %
Specimen
1 2 3 4 5 6 e C.I.
±15.3
Section 1mm* 1061.20 1059.50 1065.20 1025.60 - 1036.60 1049.62
1
Section 2mm* ±69.0
1507.80 1327.50 1485.50 1561.10 1542.50 1542.90 1494.55
4
Section 3mm* 1666.00 1700.70 1736.60 1721.50 1692.70 1795.80 1718.88 ±35.8
7
Solid ±35.3
1770.60 1721.20 1783.70 1857.10 1801.90 1787.30 1786.97
3
Section 2mm
** ±44.8
1315.00 1489.50 1358.00 1380.80 1340.90 1389.40 1378.93
3
Table 6-3. Elongation at break of all Non-heat treated Specimens (%)

(Noted: *Closed hollow and **Open hollow Structure)
Sample 95 %
1 2 3 4 5 6
Interval
Section 1mm* 3.14 4.15 3.21 2.89 - 5.28 3.73 ±0.87
Section 2mm* 4.88 4.73 4.29 4.02 4.21 5.28 4.57 ±0.38
Section 3mm* 5.17 7.54 6.20 6.38 5.44 6.54 6.21 ±0.68
Solid 16.13 11.55 14.76 13.30 13.61 7.28 12.77 ±2.48
**
Section 2mm 3.84 5.48 3.67 4.79 5.79 4.78 4.73 ±0.68
Appendix Page 218
Table 6-6. Melting Peak Temperature Sintered Material of Non-heat treated Specimens (oC)
Sample 95 %
Avera Confiden
Specimen
1 2 3 4 5 6 ge -ce
Interval
Section 1mm 185.00 185.46 184.97 183.18 184.19 184.26 184.51 ±0.65
Section 2mm 185.42 185.83 186.52 185.11 187.92 186.06 186.14 ±0.80
Section 3mm 183.74 182.25 183.23 183.07 184.03 184.95 183.55 ±0.74
Solid 181.22 181.81 181.62 181.72 181.12 182.02 181.59 ±0.28
Table 6-7. Melting Peak Temperature Un-sintered Material of Non-heat treated Specimens
(oC)
Sample 95 %
Avera Confiden
Specimen
1 2 3 4 5 6 ge -ce
Interval
Section 1mm 189.82 189.41 189.41 189.46 189.34 189.03 189.41 ±0.20
Section 2mm 189.35 189.66 189.40 189.70 189.83 189.40 189.57 ±0.17
Section 3mm 189.58 189.25 190.99 189.35 188.90 189.25 189.55 ±0.59
Heat treated with filling scanning:
Table 6-8. Tensile Strength Results of Heat Treated Specimens (MPa)

Sample 95 %
1 2 3 4 5 6
Interval
Section 1mm 25.84 26.54 25.26 25.95 25.96 25.91 25.91 ±0.33
Section 2mm - 40.06 39.34 35.31 38.90 39.25 38.57 ±1.64
Section 3mm 45.21 42.90 43.40 44.74 44.21 43.55 44.00 ±0.70
Solid 51.50 51.24 51.75 51.22 51.52 50.61 51.31 ±0.32
Table 6-9. Tensile Modulus Results of Heat Treated Specimens (MPa)

Sample Averag 95 %
Specimen
1 2 3 4 5 6 e C.I.
Section 1mm 1197.60 1240.30 1432.80 1353.70 1323.40 1412.20 1326.67 ±74.58
Section 2mm - 1776.50 1690.40 1618.90 1707.30 1752.20 1709.06 ±53.48
Section 3mm 1983.30 1970.70 1898.20 1930.30 1917.40 1898.70 1933.10 ±29.05
Solid 1894.40 1945.30 2023.90 1902.40 1962.20 1963.40 1948.60 ±37.80
Appendix Page 219

Table 6-10. Elongation at break of Heat Treated Specimens (%)

Sample 95 %
1 2 3 4 5 6
Interval
Section 1mm 3.81 4.25 3.29 3.22 4.09 3.79 3.74 ±0.33
Section 2mm - 4.31 4.08 3.81 4.38 4.92 4.30 ±0.36
Section 3mm 4.04 3.62 4.55 4.50 4.87 4.82 4.40 ±0.39
Solid 10.58 11.15 9.79 13.05 13.02 12.19 11.68 ±1.10
Table 6-11. Crystallinity Results Sintered Material of Heat Treated Specimens (%)
Sample 95 %
1 2 3 4 5 6
Interval
Section 1mm 31.98 29.09 29.72 28.55 28.85 28.76 29.49 ±1.03
Section 2mm 30.07 30.31 28.19 27.49 28.24 27.30 28.60 ±0.96
Section 3mm 33.79 29.88 30.22 29.13 28.43 28.47 29.99 ±1.60
Solid 27.83 28.85 27.36 27.52 25.37 26.87 27.30 ±0.92
Table 6-12. Crystallinity Results Un-sintered Material of Heat Treated Specimens (%)
Sample 95 %
1 2 3 4 5 6
Interval
Section 1mm 37.99 40.02 39.51 37.99 38.97 40.56 39.17 ±0.85
Section 2mm 42.25 41.90 39.57 38.82 36.73 40.86 40.02 ±1.59
Section 3mm 42.69 41.61 - 41.86 40.60 36.90 40.73 ±1.99
Table 6-13. Melting Peak Temperature of Heat Treated Sintered Material (oC)
Sample 95 %
Avera Confi-
Specimen
1 2 3 4 5 6 -ge dence
Interval
Section 1mm 191.48 191.74 192.85 194.29 192.66 196.43 193.24 ±1.48
Section 2mm 189.57 187.39 189.51 191.26 191.61 192.47 190.30 ±1.47
Section 3mm 191.21 192.83 191.09 193.37 193.55 192.57 192.44 ±0.85
Solid 186.95 185.47 186.44 185.24 187.42 188.61 186.69 ±1.01
Appendix Page 220

Table 6-14. Melting Peak Temperature of Heat Treated Un-sintered Material (oC)
Sample 95 %
Avera Confi-
Specimen
1 2 3 4 5 6 -ge dence
Interval
Section 1mm 193.09 196.84 194.05 194.72 194.40 194.55 194.61 ±0.99
Section 2mm 191.42 191.94 192.37 193.49 192.66 192.99 192.48 ±0.59
Section 3mm 191.20 191.81 - 192.74 192.84 194.69 192.66 ±1.16
Non-heat treated with filling and outline scanning
Table 6-15. Tensile Strength Results (MPa) – With Fill and Outline Scanning
Sample 95 %
1 2 3 4 5 6
Interval
Section 1mm 31.68 30.08 32.79 32.70 33.21 32.01 32.08 ±0.90
Section 2mm 37.97 38.14 37.55 37.11 36.05 37.13 37.33 ±0.60
Section 3mm 42.75 45.68 44.86 43.39 44.95 45.86 44.58 ±1.00
Solid - 49.40 49.82 48.75 49.45 49.23 49.33 ±0.34
Table 6-16. Tensile Modulus Results (MPa) – With Fill and Outline Scanning
Sample Averag
Specimen 95 % C.I.
1 2 3 4 5 6 e
Section 1mm 1528.10 1554.90 1595.00 1529.80 1585.50 1554.20 1557.92 ±22.17
Section 2mm 1809.70 1853.90 1770.90 1731.90 1654.30 1695.10 1752.63 ±64.71
Section 3mm 1831.20 1834.10 1891.50 1830.20 1907.50 1972.90 1877.90 ±45.91
Solid - 1970.30 1923.20 1879.60 1946.30 2025.70 1949.02 ±47.66
Table 6-17. Elongation at break (%) – With Fill and Outline Scanning
Sample 95 %
1 2 3 4 5 6
Interval
Section 1mm 3.58 3.20 4.12 4.04 3.73 3.58 3.71 ±0.27
Section 2mm 5.28 4.98 4.37 4.67 4.43 4.87 4.77 ±0.28
Section 3mm 5.13 6.79 5.63 4.76 5.41 5.87 5.60 ±0.56
Solid - 8.73 10.21 8.82 10.20 10.55 9.70 ±0.75
Appendix Page 221

Table 6-18. Energy Density (J/mm2)– With Fill Scanning

Sample Averag 95 %
Specimen
1 2 3 4 5 6 e C.I.
Section 1mm 5004.00 5009.80 5010.77 5395.54 - 4988.92 5081.80 ±153.9
2
Section 2mm 10355.8 10383.3 10404.4 10389.9 10412.4 10416.9
76052.2 ±18.12
0 1 4 0 7 0
Section 3mm 14348.6 14323.6 14305.1 14285.8 14307.9 14302.4 104723.
±17.22
1 7 7 4 4 1 9
Solid 17645.9 17604.9 17693.3 17650.9 17686.4 17632.3 17652.3

±26.61
8 9 2 7 7 2 4
Table 6-19. Energy Density (J/mm2)– With Fill and Outline Scanning
Sample Averag 95 %
Specimen
1 2 3 4 5 6 e C.I.
±110.1
Section 1mm 8055.42 7881.59 8116.09 7826.61 7982.49 8184.64 8007.81
0
Section 2mm 13068.8 13160.8 13197.3 13179.4 13120.1 13161.6 13148.0
±37.20
8 8 0 3 7 9 6
Section 3mm 16332.6 16427.1 16401.4 16366.9 16395.8 16304.3 16371.4
±36.89
0 6 2 9 9 8 1
Solid 18913.1 18995.5 18916.7 18984.4 18964.4 18954.8
- ±33.40
1 0 9 1 0 4
Appendix Page 222

Post Processing For Nylon 12 Laser Sintered

Uploaded by

Copyright:

Available Formats

Post Processing For Nylon 12 Laser Sintered

Uploaded by

Document Information

Original Title

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Post Processing For Nylon 12 Laser Sintered

Uploaded by

Copyright:

Available Formats

Post Processing for Nylon 12 Laser Sintered

School of Mechanical and Systems Engineering

Thesis submitted for partial fulfilment for

I would like to express my appreciation and gratitude to my supervisor, Prof. Kenny

Newcastle upon Tyne, November 2016

This research investigates the effect of post-processing on the mechanical

Abstract Page iii

CHAPTER 2. LITERATURE REVIEW 7

2.2.5.1 Definition and basic process .................................... 17

2.5.2 Review of heat treatment of polymers and its impact .............. 49

CHAPTER 3. EXPERIMENTAL STUDIES AND PROCEDURE 60

CHAPTER 4. MECHANICAL PROPERTY CHARACTERISATION 78

Table of Contents Page vi

CHAPTER 5. HEAT TREATMENT STUDIES 97

CHAPTER 6. HEAT TREATMENT AND MECHANICAL PROPERTIES OF

Table of Contents Page vii

6.4.3 Melting Peak Temperature (Tpeak) Results for Non-heat treated

CHAPTER 7. CONCLUSIONS AND FUTURE WORK 173

A: SHEAR AND BENDING TEST RESULTS ............................ 193

B: ANOVA RESULTS .................................................................... 195

C: DSC TEST RESULTS ............................................................... 206

D: DATA RESULTS FOR CH. 6 .................................................. 218

Table of Contents Page viii

E: JIG AND FIXTURE FOR SHEARING TEST ....................... 223

Table of Contents Page ix

Figure 2-1 Principle process of AM .......................................................... 8

List of Figures Page x

polymerised at 453 K ............................................................... 44

Figure 3-1 VanguardTM SLS System ......................................................... 61

List of Figures Page xi

Figure 4-21 Comparison of results of test of flexural modulus (four-point) …… 93

List of Figures Page xiv

List of Figures Page xv

Figure 6-10 Elongation at break results of different section thickness of closed

List of Figures Page xvi

List of Figures Page xviii

Table 2-1 Mechanical Properties of Polymers ........................................................ 39

List of Tables Page xix

List of Tables Page xx

ABBREVIATIONS AND NOTATIONS

Abbreviation and Notations Page xxi

Cp : specific heat (J/kg K)

Abbreviation and Notations Page xxii

Manufacturing industry is seeking innovative methods to accelerate new product

Chapter 1. Introduction Page 1

1.2 Research Gaps and Problems

In conventional manufacturing processes, the repeatability of mechanical properties is

Chapter 1. Introduction Page 2

It has been recognised that material properties are influenced by microstructure,

Chapter 1. Introduction Page 3

1.3 Research Aim and Objectives

1.4 Research Scope and Limitation

Chapter 1. Introduction Page 4

3) Identifying current research on the tensile properties of laser-sintered products

In this research, was initially performed benchmarking for mechanical

1.5 Structure of the Thesis

Chapter 1. Introduction Page 5

mechanical properties. In addition, heat-treated and non-heat-treated specimens are

Chapter 1. Introduction Page 6