 

 

 

 

 

 

 

 

 

Fl. 3.553%

4V».—

'I

C.
a.
.

p i:
3.15:4.

_ .9. .
.Ww‘!

. .2:
xﬂww

J ‘. . ..
inﬁnfﬂhlu

"may

Jurindww
do‘: . ,
9

“mm;

nr .HE:

_ .0.
in?
W, In".
.. 5.31..
ﬁe... V;
I Inch“. oIKf.
:li-quh..."
a).
a!
\ﬂwd
.9.

ya» u!
:2». RE

3}. 6.1..1 L.

’71;

I.

.I on}:

J biolv...
.\

1
ll

{3 3... v0.5.1.3 l
5.15.! 9 0‘ .i
Eil- auD .

OIL \
I... I4“. . V
v’mﬁp1‘f‘

.10 rlbvtttxl .51.,
v05 2 . 1V. .-

0‘ rt.

.31?! a . I
, III...I .. 3-21.31
i: .1 - a. kw.) nu.
I‘Nvdlqcr" 1.. 'gltg 3 a
.....cu....u.....r‘ {3.1-6.31}... r13..,.u.ﬂu€u.le...6t 3 It
. v 1’ I it! I k. I ll-l \

 

 

THESiS

Date

' itiiﬂiiijﬂiiiiiiiiiiWilli

01050 0720

 

LIBRARY

Michigan State
University

 

 

 

This is to certify that the

dissertation entitled

LASER BEAM PROCESSING OF METAL
MATRIX COMPOSITES

presented by

Satyanarayana Kudapa

has been accepted towards fulﬁllment
of the requirements for

Ph. D. degree in Wience

 
 

Major profess r

  

November 51 1996

MSU is an Afﬁrmative Action/Equal Opportunity Institution 0-12771

 

 

 

 

PLACE iN RETURN BOX to remove this checkout from your record.
TO AVOID FINES return on or baton onto duo.

DATE DUE DATE DUE DATE DUE

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

LASER BEAM PROCESSING OF METAL MATRIX COMPOSITES

By

Satyanarayana Kudapa

A DISSERTATION

. Submitted to
Michigan State University
in partial fulﬁllment of the requirements
for the degree of

DOCTOR OF PHILOSOPHY
Department of Materials Science and Mechanics

Michigan State University
East Lansing, MI 48824

1996

ABSTRACT

LASER BEAM PROCESSING OF METAL MATRIX COMPOSITES
By

Satyanarayana Kudapa

Wide spread utilization of metal matrix composites (MMCs) for structural applications
can be fully realized when appropriate techniques for machining and joining are identi-
ﬁed and successfully implemented with minimal degradation to the engineering prOperties
of the NINICs. Laser beam machining (LBM), laser beam welding (LBW), and collinear
dual laser beam welding (CD-LBW) of 6061/A1203/10p (‘p’ stands for particulates) and
606l/A1203120p composites were recognized as viable processes for this study. Stationary
MIVIC samples were irradiated under a moving carbon—dioxide laser source with a TEMOI-
mode.

LBM of MMCs has been optimized in terms of the average surface roughness of the
cut surface and the width of the thermal damage, i. e., the heat affected zone (HAZ). The
inﬂuence of various machining and process parameters such as cutting speed, laser power,
focusing conditions, coaxial and off-axis cover gas type and pressure (ﬂow rate), and type
of laser optics were studied to minimize the average surface roughness and width of the
damage zone. '

Study of bead-on-plate conduction mode welds generated by LBW and CD-LBW tech-
niques were characterized in terms of porosity, micro-cracks, width of molten- and heat
affected zones, depth of penetration, and width of weld bead. Signiﬁcance of the vol-

ume percentage and distribution of the A1203(,,) on stability of hot plasma envelope were

studied. The stability and generation of plasma has a distinctive effect on the weldments

produced by LBW of MMCs. To reduce the wrappage of the MMC sample, to prevent mi-
crocracking, to avoid porosity, and to enhance the quality of the weld bead, a novel process
referred to as CD-LBW was successfully developed in this study.

To my parents

iv

ACKNOWLEDGEMENTS

I wish to thank Prof. K. Mukherjee, for his guidance, advice, and ﬁnancial support
throughout the duration of this work. I would also like to extend my sincere gratitude to
my dissertation committee members: Prof. K. N. Subramanian and Prof. Dahsin Liu of
Department of Materials and Mechanics, Michigan State University, and Prof. Otto Suchs-
land of Department of Forestry, Michigan State University, for their invaluable advice and
suggestions.

I would also like to express my sincere gratitude to Prof. Vladimir Bamekov for his help
in the design of dual beam laser heads. I would like to take this opportunity to thank Mr.
Charles T. Lane, Duralcan USA, San Diego for providing aluminum composite samples. I
would like to thank Dr. C. W. Chen for his constant encouragement and help, Dr. Moti Tayal
and Jim Howard for their friendship and many insightful discussions on various topics, long
Kook Park for his help in polishing the composite samples, Venkateshwar Rao Aitharaju
for helping me with FEM analysis, and Srinivas Gooty Tumbalam for being a considerate
room-mate. Iris A. Taylor, Jo Ann Peterson, Debbie M. Conway, Loma Coulter, and Leo
Szafranski deserve my sincere gratitude for all their help during my stay at MSU.

I will always cherish and value the warm relationship I had with my dear friends: Siva
Ramkrishna Nayudu, Sharathchandra Pankanti, Ravi Mandava, Kamalakar Gogineni, Pa-
van Chavva, Sai Krishna Yarlagadda, Vibhavasu Vuppala, and Ilsung Oh.

Finally, I reserve my deepest and most sincere gratitude to my family members: my
parents (Kameswara Rao and Sita Ratnam), my brothers (Brahma Rao and Ramesh), my
sisters (Lakshmi and Satyavathi), and my uncle and aunt (Suryapraksha Rao and Anusuya
Prabhavati Devi) for their endless sacriﬁces.

TABLE OF CONTENTS

LIST OF TABLES

LIST OF FIGURES

1 INTRODUCTION

1.1
1.2
1.3
1.4

2.1

2.2

2.3

2.4
2.5

2.6
2.7

Motivation ...................................
Machining ...................................
Joining .....................................
Objectives ...................................
1.4.1 Machining ..............................
1.4.2 Welding ................................

OVERVIEW OF LASERS

Basic Laser Operation .............................
2.1.1 Radiative Transitions .........................
2.1.2 Spontaneous and Stimulated Emission ................
Characteristics of the Laser Beam ......................
2.2.1 Spatial Pattern of the Laser Beam ..................

2.2.1.1 Gaussian Beam - TEMoo Mode ..............

2.2.1.2 Focal Spot Size ......................

2.2.1.3 Depth of Focus ......................
Measurement of Energy ............................
Types of Commercial Lasers .........................
Solid-State Lasers ..............................
2.5.1 Nd:YAG Laser ............................

2.5.1.1 Means of Excitation ....................
Excimer Lasers ................................
C02 Laser ...................................
2.7.1 Energy Levels .............................
2.7.2 Method of Excitation .........................

vi

10
10
11
12
14
14
l6
19
2O
21
23
25

26
29
3 1
3 1
34

2.7.3 Types of C02 lasers .........................
2.7.3.1 Slow Axial Flow C02 Laser ................
2.7.3.2 Fast Axial Flow C02 Laser ................
2.7.3.3 Helical Flow C02 Laser ..................
2.7.3.4 Cross (Transverse) Flow C02 Laser ............

2.8 Applications of C02 laser ...........................

2.8.1 Marking ................................

2.8.2 Scribing, Etching, and Drilling ....................

2.8.3 Fusing and Welding .........................

2.9 Summary ...................................

PROCESSING OF MMCs

3.1 Fabrication Techniques ............................
3.1.1 Matrix Materials ...........................
3.1.2 Reinforcements ............................

3.2 Physical Properties ..............................

MACHINING OF MMCs

4.1 Machining of Polymer Composites ......................

4.2 Machining of Metal Matrix and Ceramic Composites ............
4.2.1 Ceramics Processing .........................

4.3 Energy Transfer Mechanism in Laser Machining ..............

4.4 Other Non-traditional Machining Methods ..................
4.4.1 Water—jet and Abrasive Water-jet Machining ............
4.4.2 Electro—Chemical Machining (ECM) ................
4.4.3 Electron Beam Machining (EBM) ..................
4.4.4 Electrical Discharge Machining (EDM) ...............

JOINING OF MMCS
5.1 Joining Processes ...............................
5.2 Adhesive Bonding ...............................
5.3 Solid State Processes ..................... _ ........
5.3.1 Diffusion Bonding ..........................
5.3.1.1 Solid Phase Diffusion Bonding ..............
5.3.1.2 Transient Liquid Phase Bonding .............
5.3.2 Friction Welding ...........................
5.4 Fusion Processes ...............................
5.4.1 GTA and GMA Welding .......................

vii

35
35
37
37
38
39
39
39
4O

41
41
42
45
46

48
49
50
50
5 1
52
54
56
59
6O

63
64
65
67
67
68
69
70
7 l
75

5.4.2 Capacitor Discharge Welding .................... 77
5.4.3 Resistance Welding .......................... 78
5.4.4 Electron Beam Welding ....................... 79
5.4.5 Laser Beam Welding ......................... 79
6 EXPERIMENTAL PROCEDURE 87
6.1 Machining ................................... 96
6.2 Welding .................................... 100
6.2.1 Plasma Control ............................ 100
6.3 Microstructures and Hardness Testing .................... 101
6.4 Finite Element Analysis (PEA) ........................ 104
6.4.1 Finite Element Model ......... . ............... 107
7 RESULTS AND DISCUSSION 110
7.1 Energy Required for LBM and LBW ..................... 110
7.2 Laser Beam Machining ............................ 112
7.2.1 Inﬂuence of Cover Gas ........................ 113
7.2.2 Role of Cover Gas Pressure ..................... 115
7.2.2.1 Effect on Roughness .................... 115
7.2.2.2 Cutting Speed and Kerf Width .............. 117
7.2.3 Thickness Vs. Cutting Speed ..................... 119
7.2.3.1 Mathematical Model .................... 121
7.2.4 Position of Focal Plane ........................ 129
7.2.5 Roughness of the Cut Surface .................... 133
7.2.6 Mechanism of Cutting .................... - . . . . 142
7.2.7 Heat Affected Zone .......................... 144
7.2.8 Inﬂuence of Optics .......................... 146
7.3 Laser Beam Welding ............................. 149
7.3.1 Single Beam LBW .......................... 150
7.3.1.1 AbsorptiOn of Laser Power ................ 150
7.3.1.2 Characterization of Weld Bead .............. 152
7.3.1.3 Effect of Traverse Speed on Depth of Penetration . . . . 152
7.3.1.4 Inﬂuence of Power on Depth of Penetration ....... 154
7.3.2 Effect of Optics ............................ 166
7.3.3 Role of Particle Distribution ..................... 170
7.3.3.1 Effect of Hot-Rolled Composite .............. 172

7.3.3.2 Depth of Penetration in As—Received and Hot-Rolled
Composites ........................ 172

viii

7.3.3.3 Width of the Weld nugget in As- Received and Hot-

Rolled Composite ..................... 178

7.3.4 Dual Laser Beam Welding ...................... 180

7.3.4.1 Microhardness of the Weldment .............. 188

7.3.4.2 Size of Fusion Zone and HAZ ............... 192

7.3.5 Results of Finite Element Analysis .................. 194

8 CONCLUSIONS 205
8.1 Machining ................................... 205

8.2 Welding .................................... 206
8.2.1 Single Beam Welding ........................ 206

8.2.2 Dual Beam Welding ......................... 208

A Theoretical Aspects of Focal Plane Position 209
B Additional Finite Element Results 212
BIBLIOGRAPHY 217

2.1
2.2

4.1

5.1
5.2

6.1

6.2
6.3
6.4
6.5
6.6

LIST OF TABLES

Characterisitics of commerically available lasers. .............. 24
Some of the key features of carbon dioxide and Nd:YAG laser. ....... 28
Classiﬁcation of non-traditional processes ................... 53
Ratings for processes used for MMCs joining ................. 65
Advantages and disadvantages of the various joining processes used for

MMCs. .' ................... '. ............... 86
Nominal composition of the 6061 Al alloy, 6061/A1203/ 10p, and

6061/A1203/20p composite matrix in weight percent. ............ 88
Characteristics of the RF excited C02 laser beam ............... 92
CN C machine speciﬁcations .......................... 94
A list of experimental variables for machining. ............... 96
List of laser parameters and experimental variables for joining ........ 99
Average values of the thermo-physical properties of 6061/A1203/20p, 6061

A1, and A1203. ................................ 108

2.1
2.2

2.3

2.4

2.5
2.6
2.7
2.8
2.9
2.10
2.1 1

3.1

3.2

4.1
4.2

LIST OF FIGURES

Types of emission processes between two energy levels ............ 12
Transverse mode patterns of gas lasers. The integers indicate the number
of nulls as one passes through the beam pattern. (a) Rectangular symmetry
and (b) cylindrical symmetry .......................... 15
Variation of relative intensity and percentage total power with radius for a
Gaussian TEMoo beam. Note that for the vertical line representing e“, the
value should read 0.71 instead of 0.59 ..................... 17
Intensity distribution for TEMoo, TEMm- , and TEMIO modes. The radii
(wm, Wop, and Woo) are normalized to the beam radius Woo of the "FEM

mode. ..................................... 18
Inﬂuence of location of the focal spot with respect to the workpiece. . . . . 22 '
Energy levels of N d:YAG laser ......................... 27
Schematic sketch of N szAG laser machine .................. 27
Schematic of the energy-level curves for the KrF molecule .......... 29
The various energy levels for C02 laser. . . ' ................. 32
Schematic of the various types of lasers. ................... 36
Laser material interaction energy diagram ................... 38

Schematic representation of (a). powder metallurgy process for MMC fab-
rication and (b). spray deposition process. .................. 43
Schematic illustration of (I)(a). continuous compocaster, (b). monocharge
compocaster. (II) Illustration of squeeze casting process: (a). direct inﬁl-
tration of a preform, (b). squeeze casting of a compound, and (c). indirect

inﬁltration of several preforms ......................... 44
Schematic representation of water—jet cutting system ............. 55
Schematic sketch of cathode and anode conﬁgurations for ECM. (a) Initial

setup and (b) Final shape of the workpiece. ................. 57

4.3

5.1

5.2
5.3

5.4

5.5

6.1

6.2

6.3

6.4

6.5

6.6

6.7

6.8

6.9

7.1

7.2

Production of sparks in EDM process. (a) Avalanche of electrons, (b) posi-
tively ionized gas in gap, (c) secondary avalanches, and (d) streamer devel-
opment. .................................... 61

Microstructure of A14C3 reaction product in an electron beam melt run in

2024/SiC/20p composite ............................ 73
Laser welding process diagram to illustrate penetration weld geometry. . . . 81
Schematic to illustrate the two modes of laser beam welding: (a) Deep
penetration weld and (b) Thermal conduction limited weld .......... 82
(a). The length and width of A14C3 plates and (b). volume fraction of SiC
and A14C3 as a function of speciﬁc energy ................... 84
Schematic of a split beam laser welding system ................ 85
Optical micrograph of a polished as-received extruded 6061/A1203/10p-T6
composite .................................... 89
Optical micrograph of a polished as-received extruded 6061/A1203/20p-T6
composite .................................... 90
Optical micrographs of polished hot-rolled (a). 6061/A1203/10p-T6 and
(b). 6061/A1203/20p-T6 composite in the rolling direction. ......... 91
A schematic representation of laser setup for machining and single beam
LBW. ..................................... 93
Laser head nozzles used for (a) machinging experiments and (b) single
beam welding experiments ........................... 95
Schematic illustration of the differences between meniscus and aspheric
lenses ...................................... 98
Schematic illustration of the CD-LBW of 6061/A1203/10p and
6061/A1203/20p composite ........................... 102

(a) Schematic of the composite specimen displaying the cross—section used
for SEM studies and (b) schematic of the varying width composite specimen. 105
Finite Element mesh with a total of 880 “8-node” brick elements. ..... 109

Variation of rate of weight loss with cutting speed for 6061/A1203/20p
composite of 5.0 mm thick, with a power of 2400 W and cover gas pressure
of 415 kPa. .................................. 114
Photomicrograph displaying the kerf width and cut surface for a 5.0 mm
thick 6061/A1203/20p composite, with nitrogen cover gas at 415 kPa and
cut at 0.508 m / min (indicated by ‘X’). The exposed surface was cut at 300
kPa (indicated by ‘0’) ............................. 116

xii

7.3

7.4

7.5

7.6

7.7

7.8

7.9

7.10
7.11

7.12

7.13

7.14

7.15

7.16

The variation of kerf width with laser power and cover gas pressure for a
5.0 mm thick 6061/A1203/20p composite cut at a ﬁxed cutting speed of
0.508 m/mz'n .................................. 118 .
Dependency of thickness of 6061/A1203/20p composite on cutting speed. . 120
Schematic diagram showing cutting parameters and the coordinate system
used for the model .............................. 122
Comparison of theoretical and experimental data of depth of cut versus
cutting speed. The values indicated on the solids lines represent 6 values

and absorptivity is assumed to be 10% ..................... 126
Nonlinear relationship between B and the cutting speed. Absorptivity, a =
10% ....................................... 127
Predicted and experimental values of maximum thickness of composite that
can be cut through. Absorptivity, a = 10%. ................. 128
Variation of depth of cut with focal point position from the surface of
6061/A1203/20p composite ........................... 130
Formation of density gradient ﬁeld on the surface of the material ....... 131
Focal position relocation by density gradient ﬁeld and the original focal
plane location .................................. 133
Variation of roughness of the cut with cutting speed for 6061/A1203/20p
composite at a pressure of 415 kPa of N2 cover gas. ............ ‘134

Photomicrograph showing rough surface cut with nitrogen cover gas at 2.0
m / min and at 415 kPa pressure for an extruded 6061/A1203/20p composite. 136
Photomicrographs of as cut surface for a 6061/A1203/20p to show relative
roughness of nitrogen and oxygen cover gases. (a) Nitrogen cover gas at
415 kPa and cut at 0.508 m/mz'n and (b) oxygen cover gas at 415 kPa
and cut at 0.508 m/min. ........................... 138
SEM micrograph of as cut surface, with nitrogen as cover gas and cut at
0.508 m/min. Less recast layer and the white area is the oxidized layer
(shown by ‘X’): (a) at low magniﬁcation and (b) at higher magniﬁcation
for a 6061/A1203/20p composite ........................ 140
Photomicrograph of as cut 6061/A1203/20p surface, with oxygen as cover
gas and 0.508 m / min cut speed (a) showing recast layer (dark region indi-
cated by ‘0’) and vacancies left by ejected particles (indicated by ‘X’) and
(b) at higher magniﬁcation (vacancies indicated by arrow). ......... 141

7.17 Work material interaction with laser beam: (a) transverse section, (b) lon-

gitudinal section, and (c) material removal mechanism for through cutting. . 143

xiii

7.18

7.19

7.20

7.21

7.22

7.23

7.24

7.25

7.26

7.27
7.28

SEM micrographs of polished composite sample: (a) showing particulate
distribution in the as-received composite, (b) depicting Heat Affected Zone
(HAZ) at entrance side of the cut, (c) HAZ at mid—section of the cut, and
(d) HAZ at bottom-section of the cut. HAZ in (b) is large as compared to

(c) and (d) ...................................

SEM micrograph of the LBM surface of a 6061/A1203/20p composite (re-

fer text for processing conditions). ......................

Macrograph of bead-on-plate weld formed on a 3.2 mm thick
6061/A1203/10p particulate composite, at a traverse speed of 0.8 m/mz’n
and heat input of 1200 W. (a) The fusion zone comprised of thermal cracks
(indicated by arrow), (b) non-uniform melt ﬂow, and (c) numerous pores

(indicated by ‘x'). ..............................

Plot of variation of depth of penetration with traverse speed, for various
laser powers. The thickness of the 6061/A1203/ 10p specimen was 3.2 mm
(6 data points were used to generate the range indicated by the error bars).
The variation of depth of penetration with heat-input, for bead-on-plate
weld of as-received 6061/A1203/ 10p of ﬁxed thickness (3.2 mm) and con-

stant traverse speed ...............................

Scanning electron micrograph of the transverse section of the bead, indi-
cating the presence of both (a) spheroidal (arrow) and (b) angular pores
and thermal cracks (indicated by arrows), for power input of 1200 W and

a traverse speed of 0.8 m / min (30 z’pm) ....................

SEM micrograph of the cross-section of the weld bead, indicating the bot-
tom half. A traverse speed of 0.8 m/mz'n (30 11pm) and heat input of 1200

W were used. The recrystallized grains are indicated by arrows. ......

Weldment formed at 2.5 m/mz'n (100 rpm) and laser power of 1700 W:
(a)Macrograph of the transverse section indicating the shallow depth of
penetration and (b) SEM picture indicating micro-cracks on the top surface

of the weld bead. Thermal cracks are indicated by arrows. .........

Macrostructure of the 6061/A1203/ 10p composite, showing the top view
of the weld bead formed at 1700 W heat input and 1.3 m/mz'n (50 im)

traverse speed ..................................
. 164

SEM micrograph of the top surface of the weldment shown in Figure 7.26
A lamellar like structure observed in the transverse section of the weldment

shown in Figure 7.26 ................... ' ...........

xiv

145

148

151

. 153

155

156

159

161

163

165

7.29

7.30

7.31

7.32

7.33

7.34

7.35

7.36

7.37

7.38

7.39

7.40

Photomicrograph (at higher magniﬁcation) showing a region depleted of
alumina particles in the weld path. Off-axis argon gas at 345 kPa and
weld speed of 1.3 m/mz’n. .......................... 167
Photomicrograph of alumina composite showing presence of porosity at the
top side of the weld zone (off-axis argon gas at 345 kPa and weld speed of
1.3 m/mz’n). ................................. 169
Photomicrograph of SiC whisker reinforced composite showing presence
of porosity through out the weld zone (off-axis argon gas at 345 kPa and
weld speed of 1.3 m/mz’n) ........................... 171
The variation of depth of penetration with heat-input, for bead—on-plate
weld of a ﬁxed thickness sample (2.4 mm) of 6061/A1203/10p as received
and hot rolled composite ............................ 173
Comparison of depth of penetration with heat-input, for bead-on—plate
weld of a ﬁxed thickness sample (2.4 mm) of 6061/A1203/10p and
6061/A1203/20p hot rolled composites ..................... 175
Optical micrograph of the transverse-section of hot-rolled 6061/A1203/ 10p
composite, processed at traverse speed of 1.3 m / min. The fusion-, heat-
affected-, and base-matrix zones can be observed ............... 176
Optical micrograph of the transverse-section of a hot-rolled
6061/A1203/10p composite, processed at a traverse speed of 2.5 m/ min.
Note that, in comparison with Figure 7.34, the fusion zone was smaller in
size. ...................................... 17 7
The dependence of width of the weld bead and depth of penetration on
traverse speed for a 3.5 mm thick sample of 6061/A1203/20p composite. . . 179
Optical micrograph of the transverse section of the polished and unetched
weldment produced by CD-LBW at a traverse speed of 1.3 m / min in
6061/A1203lmp composite of thickness 3.5 mm (particle deﬁcit area is

indicated by ‘X’ and pores are indicated by arrow). ............. 182
Optical micrograph depicting HAZ (denoted by ‘0’) and interface between
the HAZ and the unaffected parent matrix (denoted by ‘X’). ........ 183

SEM micrograph of etched transverse weldment showing (a) HAZ, with
grain boundaries and precipitates of Mg28i (denoted by arrow) and (b) the
unaffected parent matrix, with very few visible precipitates. ‘X’ indicates
the alumina particles. ............................. 185
Optical micrographs of (a) fusion zone in a 6061/A1203/20p composite
with a process speed of 1.0 m / min, (b) fusion zone in the same composite
at a process speed of 0.8 m / min (Contd...) .................. 186

XV

7.40

7.41

7.41

7.42

7.43

7.44

7.45

7.46

7.47

7.48

7.49

7.50
7.51

A. 1
AZ

B.1

B.2

B3

B4

(Contd...) (c) fusion zone and HAZ on side B of the same composite pro-

cessed at 0.8 m/ min ..............................

Optical micrographs of the polished and etched transverse section of the
weld bead (6061/A1203/20p) (a) the cross-section of the weld bead at lower
magniﬁcation showing the proﬁle of both the side A and B weld beads

(Contd...) ..................................

(Contd...) (b) micrograph of the fusion zone of side A, and (c) HAZ of the

weld bead formed on side A. .........................

The variation of microhardness (Vickers’) with process speed in the fusion
zone of side A and B, HAZ, and the parent matrix (6061/A1203/20p).

The variation of microhardness with distance in a hot-rolled
6061/A1203/20p composite ...........................
Contour line temperature maps (a) for collinear conﬁguration at t = 0.1 s
(b) at t = 0.33. ................................
Contour line temperature maps (a) for collinear conﬁguration at t = 2.93
(b) at t = 3.63. ................................
Comparison of temperature variation with time for depth of the weldment. .

Comparison of temperature variation with time for width of the weld bead. .

Mathematical predictions of temperature proﬁles for a single beam laser

welded 2.0 mm thick Ti sample. .......................
The depth of the weld penetration for different time increments. ......
. 203
. 204

Temperature variation on the top surface along the width of the sample.
Symmetric distribution of temperatures along the width of the sample. . .

Actual and theoretical shape of the focussed laser beam. ..........
Geometrical model for the maximum position of the focal plane. ......

Variation of temperature with time for a: = 25, y = 12.5, and 0 5 z 5 6.0

for collinear mode. ..............................

Variation of temperature with time for x = 25, y = 12.5, and 0 g z s 6.0

for pro-heat mode ................................

Variation of temperature with time for a: = 25, 0 5 Ag 3 1.25, and z = 0

for collinear mode. ...............................

Variation of temperature with time for a: = 25, O 5 Ag 3 1.25, and z = O

for pre—heat mode ................................

xvi

187

189

190

. 191

193
195
197
199
200

201
202

209
210

213

214

215

216

CHAPTER 1

INTRODUCTION

The thrust of this research is to demonstrate the feasibility of laser processing (cutting
and joining) of metal—matrix composites as opposed to conventional processing methods.
Thus the objective of this chapter is to discuss the technological barriers associated with
conventional methods of machining and joining, which restrict the usage of composite
materials in many engineering applications (Section 1.1). The discussion in Section 1.2
and 1.3, further clariﬁes the necessity for cost-effective machining and joining methods for

these advanced materials.

1.1 Motivation

Composite materials have outstanding potential in aerospace, automobile, and Sports in-
dustry because of their higher strength to weight, and modulus to weight ratios, lower
coefﬁcient of thermal expansion, and excellent wear and abrasion resistance, when com-
pared to their monolithic counterparts [1-4]. Metal matrix composites (MMCs) represent
an important addition to the growing number of advanced composite materials. MMCs
are presenting new opportunities for design of materials in high temperature and structural
applications. Among the various MMCs (ﬁber (f) reinforced, whisker or short ﬁber (w) re-

inforced, and particulate (p) reinforced composites), particulate reinforced composites are

the least expensive, because they can be manufactured by using conventional techniques,
such as casting, extrusion, rolling, forging etc. They also exhibit on the average, isotropic
properties. Thus, particulate MMCs are showing great promise for stiff and light-weight
structural applications where cost is a major consideration.

The processing of composite materials can be grouped into two categories: primary
processing and secondary processing. Even if the primary processing (or manufacturing)
of a composite, with pre-speciﬁed properties, is established, it is the secondary processing,
such as machining, joining etc., that makes an otherwise unique composite unattractive to
end-users. This is partly due to the fact that the secondary processing of composites and
composite structures has been less studied in comparison to the primary processing. Two
of the key secondary processes are machining and joining, as has been stated earlier.

The high cost associated with machining, and damage generated during machining are
major impediment to implementation of advanced materials such as ceramics and compos-
ites [5]. In some cases, current machining methods cannot be used; innovative techniques 1
or modiﬁcations of existing methods are needed [6]. Also the potential for the utilization
of MMCs for engineering applications will be fully realized only when appropriate join-
ing techniques are identiﬁed and successfully implemented with minimal damage to the
engineering properties of the composite. I

Technological progress in many branches of engineering has resulted in the introduction
of materials specially developed to withstand increased service conditions such as higher
operating temperatures and for higher stresses. These enhanced mechanical and physical
characteristics have created problems in machining and joining of these materials using
the known machine tools. and welding processes; alternative cutting and joining techniques
have had to be developed.

Non-traditional machining processes, such as laser beam machining, electro-discharge
machining, electrochemical machining, water-jet machining etc., are being used to produce

complex parts of super-alloys, ceramics, plastics, ﬁber and particulate reinforced compos-

ites, wood, and textiles in a variety of applications throughout the aerospace, automobile,
furniture, and electronic industries [5, 7]. Non-traditional manufacturing operations could
be competitive because of the reduction in the number of rejects and the ease of imple-
menting computer control. Because of steadily improving cutting capabilities, computer
and adaptive controls, these non-traditional methods have potential for playing an increas-
ingly important role in manufacturing industries [8]. However, the chances of these non-
traditional processes to replace the conventional tools are unfavorable due to cost factor.
Presently conventional processes excel in rapid volumetric removal rates. Improvements
in non-traditional material removal rates will enhance their competitiveness and range of
applications in the near future.

Among the various non-traditional machining methods listed above, laser beam ma-
chining, water-jet machining, and electrical discharge machining methods are being rapidly
accepted as economical machining methods.

1.2 Machining

The function of all machine tools, regardless of the technique being used to remove mate-
rial, is to produce components of speciﬁed dimensional size, geometric shape, and desired
surface texture. For most MMCs, it is not economically feasible to achieve these features
by traditional processes.

The economic manufacture of high precision engineering components, with functional

surfaces, presupposes that the problems of machining are solved. The machining cost of-

‘ ten amounts to 10—100 times the raw material cost. Therefore, it is essential that a method

for the preparation of components (for e. g., sintered materials processed by powder metal-
lurgy route) having a near net-shape be adopted so that the machining cost is minimized.
Although the use of net shaping techniques can produce composite components with con-

sistent dimensional accuracy, for many applications the stringent restrictions necessitate

machining after sintering. Unfortunately, most methods for manufacturing MMCs (i. e. ,
pressing, extrusion, slip casting etc.) are limited to either simple shapes or low volumes.
They are subsequently trimmed to net size or shape (requiring machining) and attached to
an assembly by adhesive bonding, mechanical fasteners (requiring drilling), or welding.
Due to the presence of both hard and soft phases in the composite, machining of metal-
matrix composites is very expensive and difﬁcult. Tool wear, for example, is much more
severe than that encountered in machining the matrix metal alone. However, to fabricate a
functional component, the ﬁrst prerequisite is that cutting and grinding of materials can be
carried out easily.

Tool materials, cutter shapes, speeds, and feeds are largely determined by the machining
properties of the reinforcing material. In addition, precautions must be taken to avoid dam-
age to the workpiece during machining (i.e., delamination, ﬁber fraying, and drill break-
through) and premature dulling of tools [9]. Dust and excessive noise, caused mainly in
cutting reinforcing ﬁbers, are major problems in using traditional machining techniques.
Moreover, Machine tool contact can introduce stress into the workpiece. For these reasons,
alumina ﬁber and particulate reinforced MMCs are difﬁcult to machine and in one case,
conventional high-speed steel twist drills had a tool life of less than one hole for these
composites [10].

The accuracy of the work produced by any cutting technique is entirely dependent on
the relative movements of the work-piece and the tool. It requires the development of ma-
chines where the positional relationship and movements of the work and of the tool could
be controlled accurately for components to be produced to the desired size and shape. The
invention of the computer numerical control (CN C) machines solves the problem for accu-
rate control of the non-traditional machine tools like laser beam. Recent developments in
lasers have provided engineers with a new machining method. The ability of laser light to
heat objects to very high temperatures has been extensively used for manufacturing opera-

tions such as cutting, welding, scribing, drilling, cladding, and surface heat treatment.

Lasers are being employed in these various industrial processes for a large variety of
materials, which are hard, brittle, and difﬁcult to machine with traditional methods. The
abrasive and brittle characteristics of the particulate and ﬁber materials, makes processing
composites with mechanical methods difﬁcult, as explained elsewhere.

Aluminum based alloys and composite materials are characterized by a low absorption
factor for the C02 laser beam and high thermal conductivity [1 1]. When processed with
C02 laser these materials produce oxides with high melting points, liquid metal with low
ﬂuidity, and dross with poor peeling. As a result, these materials are very hard to machine
and to join using a laser beam that is not properly optimized in terms of the various process
variables.

Laser is fast replacing other conventional machining processes for secondary process-
ing of composites. The ability of the laser to cut intricate shapes and complex patterns
will make it a more competitive manufacturing tool. Other advantages of using lasers in

machining are outlined below:

0 Machining process can easily be made fully automatic by CNC.
o This automation improves precision, repeatability, ﬂexibility, and productivity (due
to high processing speeds) of the machining process.

o Non-contact nature of the process makes it a desirable process to use for components
that distort easily. Abrasive materials such as composites and ceramics can be cut

without wear of the cutting tool.
0 Large mechanical forces are not exerted upon the work piece.

0 Unlike other thermal machining devices, the laser can be used with materials having

low thermal shock resistance such as ceramics.
0 Its inherent ability to machine in locations which are otherwise difﬁcult to access.
0 LBM process produces uniform cut surface ﬁnish.

0 High processing speed can be obtained and minimum re—tooling time is required.

When compared to other thermal processes, narrow to negligible HAZ is produced.

Less moving parts are involved with consequent reduction in maintenance cost.

Narrow kerf-width of the cut leads to less material loss.

A wide range of materials can be processed without requiring a substantial change in
optical components of the system. Also, it can be used for applications like cladding,

heat treatment, welding, soldering, marking, etc.

LBM has a few disadvantages owing to its inherent thermal nature. The main drawbacks

in using laser as a machine tool are:

0 Production of toxic fumes in the cutting of polymer based composites.

0 Inability to cut thick sections (i. e., sections thicker than 1 inch), due to unavailability
of high power lasers 1and large focal length optics (large depth of focus enables

processing thicker sections).

1.3 Joining

The joining processes, which have potential for usage in MMC joining, can be classiﬁed
into three groups as listed below:

Fusion processes Tungsten inert gas (TIG), metal inert gas (MIG), plasma, laser,
electron beam, and resistance welding

Solid state processes diffusion bonding, friction welding, ﬂash welding, ultrasonic

welding, and explosive welding

Adhesion processes brazing and adhesive bonding

Often, any subsequent melting (secondary processing) of these newer advanced mate-

rials can reduce or seriously impair their desired properties. The need to fabricate complex

 

1Commercial lasers with an output of 20 kW are routinely available now.

shapes, or to combine the properties with those of other monolithic materials, necessitates
joints between similar as well as dissimilar material combinations. This can be achieved
either by conventional or advanced welding techniques. To produce a sound joint, all the
welding processes rely on the application of heat and/or pressure. Arc welding is one of
the most versatile and widely used process for the joining of monolithic systems. But the
potential for damage to engineering pr0perties prohibits the application of arc welding pro-
cess for the joining of MMCs. There are three primary considerations in joining MMCs:
1). Minimal disruption of the reinforcement, 2). negligible interaction between the rein-
forcement and matrix (i.e., to maintain interfacial integrity), 3). and the ability to achieve
mechanically and metallurgically sound weld joints. These considerations restrict the ap-
plication of conventional welding methods to MMCs. Despite their potential for minimal
damage, solid—state joining techniques, such as liquid phase diffusion bonding, capacitor
discharge welding, inertia welding, and diffusion welding techniques should be evaluated
with the above objectives. Alternative joining techniques are, therefore, required to mini-
mize or avoid extensive melting of the parent materials. High energy density welding pro-
cesses such as LBW and electron beam welding (EBW), have the potential for generating
sound joints in MMCs with properties approaching that of the as—fabricated materials [12].
Some of the advantages of LBW are [13, 14]:

0 High processing speeds with minimal HAZ and residual stresses. Low distortion of

the weld component, with little or no post-weld machining and straightening.

0 Lack of a need for ﬁller material resulting in minimal contamination. Because of
the simplistic geometry of the weld joints, little or no joint preparation is required
and tooling and ﬁxturing is typically very simple. For conventional joining (also
for machining) techniques, large amount of man hours are devoted in setting up and

ﬁxturing a workpiece.

e Time—sharing can be achieved. Can be easily automated for high volume applica-
tions. CNC positioning and programmable beam control assure accuracy and re-
peatability.

0 Work piece need not be electrically conductive and as a result is suitable for join-
ing dissimilar metals of varying thermal conductivity, coefﬁcients of expansion, and

melting points.

A brief review of the literature on laser physics, types of lasers and their applications
in material processing are presented in Chapter 2. Chapter 3 describes fabrication tech-
niques and physical & mechanical properties of metal matrix composites. In Chapter 4, a
review of laser beam-machining of composites, along with other non-traditional machining
techniques is outlined, with more emphasis on laser beam machining. In Chapter 5, liter-
ature review of solid-state and fusion joining processes of MMCs are presented. Chapter
6 describes the experimental work carried out on laser beam machining and welding of
6061/A1203/10 and 20 p metal-matrix composite. In Chapter 7, experimental results on
laser beam machining and welding of 6061 Al-A1203 particulate reinforced composite are
presented. In this chapter the discussion of the results is also presented. Finally, some of

the major conclusions of the present research are presented in Chapter 8.

1.4 Objectives

The main objectives of the present study are summarized below. Here we propose LBM
and LBW as feasible alternative techniques for processing of MMCs. with the following

objectives in perspective:

1.4.1 Machining

e Study the effect of various process parameters on LBM of MMCs

e Optimize the process parameters

0 type and nature of cover gas

0 roughness of the machined surface

0 the size of the heat affected zone

0 depth of cut and width of the kerf zone

0 inﬂuence of the type of optics (reﬂective vs. aspherical optics)

e Predict the amount of molten material in the laser interaction zone.

1.4.2 Welding

e Investigate the problems associated with autogenous laser beam welding of MMCs
(6061/A1203/10p and 6061/A1203/20p) and optimize the laser parameters involved

in processing of these metal matrix composites.

0 Investigate the nature of the laser beam welding of 6061/A1203/10p and '
6061/A1203/20p composite. A

0 study the nature of thermal damage

0 study the effect of cover gas on the weld quality in terms of porosity, HAZ,
depth of penetration, micro-cracking etc.

o quantify the porosity in the weld zone

0 effect of the gas-jet dynamics on the weld nugget

o microstructural analysis of the weld bead (Stereo-microscopy, Optical mi-
croscopy, SEM, etc.)

0 study the effect of heat-input and weld speed on the depth of penetration and
width of the weld

o role of plasma formation

0 effect of reinforcement content and distribution on welding parameters

9 Numerical modeling of dual laser beam processing

CHAPTER 2

OVERVIEW OF LASERS

Lasers are being used extensively in material processing. Understanding of laser principles
and their operation is essential in selection and proper utilization of various types of lasers.
Basic principles and related aspects of lasers are covered in Section 2.1 through Section 2.3.
The various types of industrial lasers and their distinguishing characteristics are presented

in Section 2.4 to 2.7, with more emphasis on the widely used C02 lasers.

2.1 Basic Laser Operation

Light Ampliﬁcation by Stimulated Electromagnetic (Emission) Radiation (LASER) re-
quires the combination of population inversion (a requirement for ampliﬁcation), stimu-
lated emission, and ampliﬁcation. A lasing medium/active medium, an excitation source
(e. g., optical pumping, electron excitation, and resonant transfer of energy), and an optical
resonator (a positive feedback system to support and couple the oscillation) are the essen-
tial components for lasing to take place. Because of maximum directionality (spatial and
temporal coherence) and monochromaticity, the discovery of laser phenomenon (in 1958
by Schawlow and Townes) has provoked immense interest from the outset. A number of
laser characteristics, like directionality, spatial uniformity etc., are largely determined by

the cavity properties. In contrast, attributes such as gain and output power are governed by

10

11

the amplifying media [15, 16]. To better understand laser principle, one needs to distinguish

between spontaneous and stimulated emission of radiation.

2.1.1 Radiative Transitions

An electron orbiting around the nucleus, either spontaneously emits or absorbs energy
when it undergoes transition in energy levels. These emission and absorption processes
are known as radiative transitions. Associated with each transition is a speciﬁc quantum of

energy and wavelength, given by the relation:

hc
E5 — EJ' = hl/ij = A— (2.1)

1.7

where E,- is the energy in the excited state, E,- is the energy available at the lower state, ‘h’
is the Planck’s constant (6.6 X 10'3“ J—sec), ‘c’ is the velocity. of light (3 x 108 m/s in free
space), ‘A’ is the wavelength in meters, and ‘u,,- ’ is the frequency of the emitted radiation.
The number of electrons/atoms/molecules in the lower energy state (ground state) normally
exceeds the number in the upper excited state, indicating that the coefﬁcient of absorption
‘a’ is positive or coeﬂicient of gain ‘9’ is negative (for a complete explanation of ‘a’ and
‘9’, refer to Section 2.1.2). It is possible by a process known as pumping to reverse the
situation and to create a stage where the population of the upper energy state exceeds that
of the lower energy state.

There are two primary methods of pumping or energizing electrons in a laser cavity: 1).
Photon energy transfer (e. g. , ﬂash tubes) and 2). transfer of energy in an electrical discharge
(collisions with electrons which have been accelerated by the electric ﬁeld leads to excited
electrons). The result of either type of excitation is to promote a ground state electron
to a higher energy level by absorbing a quantum of energy as prescribed by Equation 2.1
and, thus eventually creating an inverted state. The above discussion has to be qualiﬁed by

stating that, absorption depends on material properties and the intensity of the radiation,

12

while spontaneous emission is governed only by the internal properties of the material. In
order to reach equilibrium, Einstein had postulated that an additional type of emission that
is dependent on the intensity of the radiation must exist. This is the so-called simulated

emission that has made laser operation possible [15].

2.1.2 Spontaneous and Stimulated Emission

An excited atom (in higher energy level) when it decays to lower energy level, emits a
quantum of radiation of frequency, 14,-, as given by Equation 2.1. This release of photon
energy is known as ﬂuorescence or spontaneous emission. This emitted photon has ar-
bitrary phase and direction. A familiar example is the common neon sign, where excited
neon atoms “dc-excite” themselves by spontaneously emitting photons of visible light. This
phenomenon of de-excitation can also be made to occur by stimulation process, if the same
atom can be stimulated (instigated) to emit this radiation when it receives radiation of the

same frequency (see Figure 2.1). Both the stimulated and stimulating radiation have the

 

 

 

 

Spontaneous
:Q.T:N: MWR
Absorption ‘ Stimulated

9”"38100

Figure 2.1 - Types of emission processes between two energy levels [17].

same directional and polarization characteristics. Stimulated emission is the reverse pro—
cess of absorption, i.e., an induced downward transition occurs in which a photon (quantum
of energy) liberates a new photon rather than being absorbed. The rate at which new pho-

tons are generated is proportional to the energy density of the photon and the difference

13

in the population (i.e. the number per unit volume) of atoms between the upper and lower
states. This phenomenon is critical for the lasing process. These concepts are explained in
more detail with energy level diagrams in Section 2.5.1 and Section 2.7.

Absorption of light energy traveling in the z-direction through a medium with popula-

tion or densities of atoms ‘N,-’ and ‘N,’ in two energy levels 1is given by [18, 19]
I, = 10 exp (—a2) (2.2)

where ‘Io’ and ‘Iz’ are intensities before and after absorption, ‘a’ is the absorption coefﬁ-

cient and is given for a Gaussian (See Section 2.2) line shape by

01 _ c2 lln 2 . _1Y,_
_ 41r f 27 7r A f
where ‘c’ is velocity of light in the material, ‘7’ is the average ﬂuorescent lifetime (relax—

ation time) of the material, ‘f’ is the frequency, and ‘A f ’ is the half—width of the Gaussian
line. On a similar note, ampliﬁcation of light energy by stimulated emission is given by

I, = Io exp (+92) (2.3)
where ‘g’ is the coefﬁcient of gain and is given by (for a Gaussian line)
_ c2 /In 2 £
g - 47rf27" 7r A f '

The similarity in the expressions for ‘a’ and ‘9’ indicate that absorption and stimulated

emission are inverse processes. The net result of these two phenomena can be expressed

62 ll 2 1
I2 = Io exp [(47l'f27') 'n-7r— (E) (N; — Nj) Z] . (2.4)

1The index i represents the higher excited atomic state and ‘j’ indicates the lower atomic state.

as [19]

 

 

14

The condition N,- > Nj should be satisﬁed for lasing to occur. This condition is known as
population inversion. Spontaneous emission tends to prevent sufﬁcient change in energy
levels for a population inversion to occur. Hence population inversion can occur only if the
upper energy level is metastable when compared to the lower level (this indicates that the
probability of spontaneous transition to the lower level is very low). De-excitation of elec-
trons by spontaneous emission leads to photon yield which is random and approximately
equal in all directions. The process of stimulated emission, however can cause photon yield
ampliﬁcation (photon cascade) that travel in a particular direction. A preferential direction

is established by placing mirrors at the ends of an optical cavity (resonator).

2.2 Characteristics of the Laser Beam

2.2.1 Spatial Pattern of the Laser Beam

The output of the laser is often characterized by its spatial and temporal coherence. Spatial
coherence describes the phase relation in a plane perpendicular to the direction of beam
propagation and temporal coherence indicates the phase relation in the direction of the
beam.

The nature of application of the laser dictates the most desirable beam intensity proﬁle
(spatial distribution) mode. As described earlier the resonator cavity design is critical not
only in the generation of proper wavelength of laser light (lasing media plays a major role
in wavelength selection), but it also effects the phase of the electromagnetic wave, result-
ing in a variety of laser beam spatial proﬁles. In addition other factors like geometry of
the optics, gain of the resonator cavity, inhomogeneities in the laser medium (hence laser
gases of at least 99.995% purity should be used), and the pumping power affect the mode
structure of the beam. The beam proﬁle is normally characterized by its transverse elec-

tromagnetic mode (TEM) and represented in the form TEan, where the subscripts ‘m’

15

 

00 IO 20 30

 

 

"‘. 4h :17: :3
H 21 33 04

(a)
e e 0
00 or* no u" 20

- 1“ §'o ‘

Ol 02 03 04

(b)

Figure 2.2 - Transverse mode patterns of gas lasers. The integers indicate the number of

nulls as one passes through the beam pattern. (a) Rectangular symmetry and
(b) cylindrical symmetry [19].

l6

and ‘n’ (small integers) represent the number of nodes/nulls in directions orthogonal to the
beam propagation (see Figure 2.2(a) and (b)). Figure 2.2 depicts the intensity distribution
in the plane perpendicular to the plane in which the mode distribution lies.

The transverse modes can be represented in either Cartesian coordinates (Figure 2.2(a))
or cylindrical coordinates (Figure 2.2(b)). For rectangular symmetry, the notation TEan
is interpreted as the number of nulls that occur in the spatial pattern in each of the two
orthogonal directions, transverse to the direction of beam propagation [19]. For cylindrical
symmetry, the integer ‘m’ indicates the number of nulls in radial direction and ‘n’ indicates
half the number of nulls in an azimuthal direction. The modes marked by asterisk are linear

superposition of two modes operating simultaneously 90° out of phase [19—21].

2.2.1.1 Gaussian Beam - TEMoo Mode

It is often desirable to operate a laser in the lowest order mode known as TEMOO mode
(Gaussian mode), because of its uniform intensity and low beam divergence, as compared .
to other higher order transverse modes. TEMoo mode has maximum intensity at its center
with exponential fall-off (decay) proportional to the square of radius of the beam. A plot
of Gaussian intensity distribution (power per unit area vs. radial distance) is illustrated in
Figure 2.3.

The intensity (irradiance) ‘I’ of the Gaussian mode as a function of radius ‘r’ from the

center of the beam is represented mathematically by

 

_ 2
I(r)=Ioexp( :2" ) (2.5)
0

where ‘ro’ is the Gaussian beam radius, where the peak power has dropped to 31; or i of
the central value. It is more common to use width corresponding to :1; of the peak power
as the Gaussian beam radius. ‘Io’ is intensity of beam at its center. For the 31, case, the

spot radius ‘ro’, implies that at r = 7:0, I = ﬁg = 0.135 Io, i.e., more than 86% of the total

l7

 

 

 

 

 

 

 

 

 

 

1 o
08 - — 20
8
3
ea — .. 4o 8
‘1‘. E
z- 2
'3 o
S . a:
g 9:} - so ‘3
O o
0 2
g .2
O
c oz — - so
e'z'
1095
r
o 059 1 1-15 1-6 2 0°

Relative radius 5;

Figure 2.3 - Variation of relative intensity and percentage total power with radius for a
Gaussian TEMoo beam [22]. Note that for the vertical line representing 6",
the value should read 0.71 instead of 0.59.

power is considered and for i, I = 0.368 Io, which indicates that roughly over 63% of the
total power is accounted. The total power of the Gaussian beam is given by

2
P = 3131.9,(2g)

From diffraction theory of light, for a monochromatic, circular light beam of width ‘D’ ,

the approximate angle of divergence ‘6’ (half angle subtended by the diverging beam) is

given by

A
0 = 1.22 (I3) . (2.7)

Equation 2.7 applies equally well for TEMoo mode laser beam and it implies that the diver-
gence angle 6 is inversely proportional to the beam size. Of all the various TEan modes
of laser beam, only Gaussian beam is truly diffraction-limited [17]. Other TEan modes

W

18

exhibit larger beam divergence angles (compared with Gaussian beam) due to the fact that
they are composed of two or more individual beams of different phases. It is also important
to note that the Gaussian beam retains its Gaussian form as the beam transmits through
optical systems. In general, beam divergence can be reduced by expanding (using beam
expanders) and collimating the beam by a factor inversely proportional to the diameter of
the expanded beam [22]. The angle of divergence ‘6’ depends on the output optics; spher-
ical and window minors result in higher ‘0’ than plane mirrors. In Figure 2.2, the mode
denoted by TEMop is a linear superposition of two similar modes rotated by 90° about the
axis relative to each other. Thus, the TEMOI. mode, which is often referred as the doughnut
mode, is the combination of TEMm and IBM“). A plot of the intensity distributions of the
TEMoo, IBM“), and My modes are illustrated in Figure 2.4.

The above discussion on the mode pattern is relevant to gas lasers, such as the He-
Ne laser and the C02 laser. Solid-state lasers such as the Nd:YAG laser, Nd: glass laser
etc., exhibit more complicated spatial patterns which are not easily describable in simple

mathematical terms.

Power [mm and.)

 

 

Figure 2.4 - Intensity distribution for TEMOO, TEMOI. , and TEMlo modes. The radii (wm,
wm. , and woo) are normalized to the beam radius woo of the TEMOO mode [23].

19

2.2.1.2 Focal Spot Size

The spot size of a diffraction-limited laser beam at the focal point, assuming no aberrations
from the lens, is given by:

d = 2 (1.22%) fﬂm—Tﬁ = 2.44 (5%) (2M + 1) (2.8)

where d = focussed beam diameter, A = wavelength of the laser beam, f = focal length
of the lens, D = diameter of the unfocussed beam, and M = number of oscillating modes.
For Gaussian beam, M = 0, hence

_ "f
d _ 2.44 (7).) . (2.9)

The following relation is more commonly used to describe the focussed spot diameter of a

Gaussian beam

_ﬂ

d_ ”D. (2.10)

Factors that affect the focal spot size are directly related to the quality of the incoming raw
beam. An output beam with small divergence angle ‘6’ can be focussed to a smaller spot
than a beam with large ‘6’ [24]. Secondly, it is evident from Equation 2.8 that the smaller
the {,- value (F—number) the smaller the ﬁnal focussed spot size. A low ‘M’ value also
gives small spot size. Finally, the diameter of the incoming beam affects the focal spot
size (Equation 2.8). Mth a speciﬁc set of focussing optics, the only way of generating
small focal spot size beam is by increasing the incoming laser beam diameter. For efﬁcient
machining and welding process, small F—number and small ‘M’ (low order mode beam)
values are highly desirable. The Gaussian beam, with minimum divergence, uniform phase
front, and smooth dr0p off of intensity ﬁom the beam center is ideal for machining opera-
tions. While the TEMm- mode which has more or less ﬂat intensity distribution with rapid

fall-off at the edge of the beam is well suited for LBW and heat—treatment applications.

20

2.2.1.3 Depth of Focus

Depth of focus is deﬁned as the distance between two points on either side of the focal
plane at which the intensity of the laser beam is diminished by the same amount. For the
Gaussian beam, a useful deﬁnition for the depth of focus is given by [17, 25 , 26]

b=:l:7r p2—1(§:-) (2.11)

where ‘d’ is the focussed spot size, ‘p’ is a tolerance factor such that ‘pd’ gives reasonable
spot size, and ‘A’ is the wavelength of the laser beam. The depth of focus ‘b’, is an arbitrary
measurement that depends on the choice of the tolerance factor ‘p’: for example, if within
the distance ‘b’ the spot size is to increase by no more than 10% (0.10) from its minimum
value at the focal plane/waist, then p = 1.10.

The depth of focus (i.e., working distance) for a C02 laser at 10.6pm with a raw beam
diameter of D = 18.0 mm and a lens of focal length f = 254.0 mm (10”), can be computed

using Equation 2.9 and Equation 2.11. From Equation 2.9, the spot size is

(10.6 x 10‘3 mm)(254 mm)
18 mm

 

d = 2.44 ( ) = 0.365 mm (2.12)

and from Equation 2.11, the depth of focus for a 10% allowable spot size increase is

 

(0.365 mm)2
= 2 _ = , . .
b darn/(1.10) 1 (10.6 x 10_3 1181mm (213)

If the raw beam size (‘D’) is increased by a beam expander of 2:1 expansion ratio, then
for D = 36.0 mm, d = 0.183 mm, and, b = :l:4.5 m. If a short focal length lens is used,
say f = 127.0 mm, then for D = 18.0 mm, d = 0.183 mm, and, b = :l:4.5 mm and for
D = 36.0 mm, d = 0.09 mm, and, b = :l:1.1 m.

From these calculations, it is obvious that by increasing the focal length, the depth of

21

focus can be increased. However, the focussed spot diameter also increases, thereby reduc-
ing the energy density of the laser beam. The energy density can be increased by decreasing
the spot size, but this would also decrease the depth of focus. Hence, there is a trade—off
between beam spot size and depth of focus. For machining operations, maximum material
removal with narrow kerf width and less material loss can be affected by utilizing a small
focal length lens, which results in small spot size with large energy density. The spot diam-
eter can be decreased either by increasing the unfocussed raw beam size (by using a beam
expander) or by decreasing the focal length. The depth of focus is an important parameter,
since it determines the positioning of the focal plane with respect to the workpiece. The
distance between the lens and the workpiece (which affects the physical location of the
focal spot in the workpiece and hence the depth of focus) has a signiﬁcant inﬂuence on the
size of the weld, the depth of the weld, the width of weld nugget, and the metallurgical
quality of the weld structure, as illustrated in Figure 2.5. An improperly positioned laser
beam may result in the formation of a hole (Figure 2.5(A)), a concave weld with a hole in
the center (Figure 2.5(B)), a concave weld (Figure 2.5(C)), convex weld (Figure 2.5(D)),
a convex weld with a hump in its center (Figure 2.5(B)), or oxidation/burning effect in the
HAZ area (Figure 2.5(F)). Hence the relationship between the optical positioning of the
beam and the resultant thermal transition of the HAZ are very critical both for welding and
machining of composite materials [27]. .

2.3 Measurement of Energy

For the visible and infrared region, the energy-measuring devices are either the thermal
type or photoelectric or quantum detector type [28]. The choice of a detector in laser
processing experiments is inﬂuenced by the frequency of the radiation and the parameter
(e. g., intensity, total energy, spatial distribution, etc.) to be measured.

The construction of thermal detector is based on the principle of conversion of the

22

 

 

 

 

 

 

 

 

 

 

 

 

 

rby} x513, j
ESP—414E

 

 

 

 

(A) (B) (C)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

( ]\JF/ \
l/\Jr/ \m
T \ / \
’ (W
i if _l
(D (F)

 

Figure 2.5 - Inﬂuence of location of the focal spot with respect to the workpiece [27].

incident radiation by the receiving element into heat. This conversion reduces the detec-
tion of the incident radiation into. that of measurement of temperature rise of the receiving
element. For the conversion, any physical property that has a known variation with tem-
perature may serve as an indicator, and for calibration purpose, the absorbed energy can be
transferred into a known amount of electrical energy. Thermoelectric detectors, bolome-
ters, and calorimeters are based on the idea outlined above. Of these devices, the former
two act as intensity detectors and the latter as an indicator of energy received during a short
pulse. The response of thermal detectors is independent of the frequency range, except for
their ability to respond very slowly with frequency (hence more suitable for low frequency

radiation). With moderate care, such variations can be factored out of any measuring de-

23

vice. The ﬁnite mass of these devices limit the sensitivity and response time as opposed to
quantum detectors, which depend on the interaction of a quantum of energy with a single
electron, and not on the absorption of incident energy over an entire macroscopic body. For
these reasons thermal detectors are widely used in the measurement of total laser energy of
a short duration pulse.

Calorimeters (commonly referred as “power meters” in laser ﬁeld) fabricated for the
measurement of laser energy comprise of a small heat capacity absorber and embedded
temperature measuring elements. The absorber may be a light carbon cone or a parallel
stack of circular blades (for high energy measurement) welded to a base plate with em-
bedded matched pairs of thermistors, thermocouples, or resistance thermometers. These
electrical elements form a part of a balanced Wheatstone bridge. Water cooling is provided
to reduce any thermal damage to the absorbing elements. Moreover, for accurate measure-

ment the laser pulse should produce only small temperature changes (a few degrees).

2.4 Types of Commercial Lasers

Lasers are being used by many manufacturing engineers in industries ranging from automo-
biles to gas turbines. They can be coupled with CAD/CAM systems to produce prototypes
and production parts economically. The various types of lasers commercially available are
listed in Table 2.1. Of the various types of lasers systems available (see Table 2.1), CO;
lasers (Section 2.7) are widely used in fabrication of components from metals, alloys, and
composites [29—33].

The C02 laser, which radiates light at wavelengths of 10.6/rm (middle infrared range)
operates by the electrical excitation of C02 gas molecules. The majority of cutting lasers
are C02 because of their high processing speeds [34]. There are two common classes of
C02 laser cavity conﬁgurations: a) Axial ﬂow and b) Transverse ﬂow or gas transport type.

The output of most axial ﬂow types can be either continuous or pulsed, at rates up to several

 

SE: 335 55353 I e. *

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

a: :5 5:0 coma—
_ _ A "-2 Xm F _ 20c _ 83
ov 9 d ddN 72 8 mud“ o; o. Nod 2:620 8 wwd 320
Honda AouoNdd Honouunoowaom
_ _ _ _ 52: 3o _ conomzamaov
dd v ON 2 o; 73 XN d4 8 To” 0; 9 Nd a. BB :8: 8 dchd omn— oanaO
Honda Ugﬂbaq

o4 space—o med CoEonv

m 8 _ I 73 XON 8 Hd I Humanoid a. mend max

dmde Come, .825

I wd 9 Nd I I Nd o. mdd 32w 8 23d even

5: md €828:

dd V o; 8 mNd I I on 8 N 32» an dwwvd coma

dm 2 ON dd 8 mNd TS x w ddN 8 mud ddddN 8 m USCG mm v.2 NOU

dd V dd 8 Nd I mod 9 wood 8“ a. 32w wchd 023$

uuonoa new

wig x N

dd R m8 8 m 9 TS do“ 8 N - Rouge do.“ wma_Odz
”IS x m;

dam R 2. 9 m 9 ”IS OS 2 d 89 9 3 Hedge do.“ 03¢ch
”I3 x m;

do v m 9 TS Xm dd 8 d - Rouge «Jaded Bax

uHonUH ouuuu Umaom

9.8 .53 9.85:5 m .523 8:5 _ _. $925 83a: 833 £5885 A83 Home—

.55650 cocomBZe 8:5 _ o>§> macs—5:00 do Ego—963 do

56.8280 anm Ana—a a. 39.528 ”ode:— 38806 .838 8502 83m 09¢.

.SN.NN.3 .m— 6: £82 vim—HE 532.3888 we «agar—83.86 .u.N 035,—.

 

25

kHz [35]. Transverse ﬂow C02 lasers operate only in CW mode. Axial ﬂow lasers can
deliver an average of 5 kW power, and slow-ﬂow axial types provide less than 1.5 kW of
average output power. But pulsed and CW Nd:YAG lasers (Section 2.5.1) are also being
used for speciﬁc material applications. This class of lasers emit radiation at 1.06pm and
have an average output of 100 to 500 W. This shorter wavelength lasers make it practical
to direct high-power beams with ﬁber-optic wave-guides, which is an inexpensive way of
delivering power, to numerous workstations. However, due to the poor quality of ﬁbers, it
is not possible to get a good quality of focussed beam as compared to conventional optics.

Fiber-optic cables are too impractical to be used with the longer wavelength C02 lasers.

2.5 Solid-State Lasers

The ﬁrst solid-state laser was a ruby laser (a three—level system), with Cr3+ doped into
sapphire (A1203) crystal, and it was invented by T. H. Maiman in 1960. Shortly after the
introduction of ruby laser, Nd:YAG laser was invented. Among the various solid-state
lasers, Nd:YAG lasers and Nd: glass lasers are the most versatile and important lasers for

material processing. The main components of a solid-state laser are [19, 23, 26]:

Host material: The mechanical, thermal, optical, and lattice properties of the host ma-
terial are important. Desirable properties of the host material include
hardness, chemical inertness, absence of internal strain and refractive
index variation, and fabricability.

Active ion: The charge state and the free—ion electronic conﬁguration of the active
ion are very crucial. The host material is very important in determining
the absorption and emission characteristics of the active ion/dopant.

Optical pump: Geometry, spectral irradiance, and duration of pumping are factors to
be considered.

26

All the solid—state lasers are excited by optical pumping and are normally operated in
pulse mode. Commonly used optical pumping sources for solid—state lasers include ﬂash
lamps, continuous discharge tubes, incandescent lamps, and solid—state lights. To operate
a solid—state laser in CW mode, a high continuous output is required from the optical
pumping sources, in addition to the power required to overcome both the threshold level
and compensate the high cavity losses [22].

2.5.1 Nd:YAG Laser

The most commonly used solid-state laser is the 1.06 pm Nd: YAG (Yttrium aluminum
garnet, Y3A15012) laser. In the Nd:YAG laser, the active element is the rare-earth Nd3+
ion. A simpliﬁed energy level diagram of the Nd3+ ions in the YAG crystal is illustrated
in Figure 2.6, with conventional spectroscopic notations for the energy levels. All the
excited higher levels (see Figure 2.6) relax non-radiatively to 4F3/2 level. The 1.06 pm
laser transition is between the 4F3/2 and 4111/2 level. The lower laser transition level, 4F11/2
is 2000 cm'1 above the ground state (419 ,2) and will normally be unpopulated at room
temperature. This indicates that the Nd:YAG laser operates on a four-level system and that
the threshold for lasing is low. A schematic sketch of the Nd:YAG laser is illustrated in
Figure 2.7 and some of the salient differences between C02 and Nd:YAG laser are outlined
in Table 2.2 [36, 37].

2.5.1.1 Means of Excitation

The Nd:YAG laser is excited by optical radiation; either with xenon ﬂash lamps for CW
power levels of the order of 10.0 W, or with krypton (Kr) ﬁlled discharge lamps for power
levels of the order of 100.0 W or more. Kr lamps are more useful for CW operation (low
current density discharges) and Xe lamps are more efﬁciently used in pulsed operation
(high current density). The use of Kr lamps increases the operating cost of the Nd:YAG
laser'[19,39-41].

27

ll Higher Energy Levels

---I---~ll -----
Optical 4F”

Pum 1.06 um
ping Laser
'Ii'ansition

4‘ ~ ~ Non-radiative Decay

‘\
§

 

 

 

 

 

 

Figure 2.6 - Energy levels of Nd:YAG laser.

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

l ' ‘ I
“QM ﬂash lamps ' cooling water
2 \t ‘1 (\1 “Fm“
crossed double- l l /1 0W
larizer \ mirror
pocke harmonic
cell Q- generator
switch (optio
rear ‘

. l_(°l’“°ml) 1 f1 1064 nm /
use} output ' pump/ca ' Nd:YAG rod 532 or
power mointor laser ca{rty my 355 um

harmonic
output

 

Figure 2.7 - Schematic sketch of Nd:YAG laser machine [3 8].

 

28

 

00300 0.50 Hood oiioc .3

b0>=0e 800m

 

8009 do countenance Hod 033080 3:3: 30:85 8: - mound Sci
28 “Seaweed a 0.33 cm? 3:365:00
Rosco b0.» .0: 00:300com .3022 30 8d Eatonﬁ 30> 72502:.

 

83:80 55 000005 9 03603

30806 05 co mucomod .Eo d.m 9d: om use

0039.83 do 0005.25.

 

ch030d 32V 8.8 wEwmoooa 32 b0>

028 3508.5 :wE

BE wﬁmwoooi

 

mczdsono “0:09 cod menace BEA—Sm co

£< was :0 £23038

 

 

 

 

 

 

 

 

03.38% 088 wdoocéocdeana :mE b0> 1308 3.0:. 3 5:90me doom 3033—2 8588“?
93:0 36> - Adowmav 5.8.5 .083 can 5&8 @033 Eow - wedge 0>300c0c 8m

Ammo—m c380 0285098 3 mam—m Wv—m Guano 030.3555 032098 b0>-0w=N 8sz

A3 ddn 8 3; 32.528 can 83d C52 2 9 3 dd: 0:025:00 3:88 533 can ode:

SE mm A 083?: 3 dd: oev 9 m 9:53 3002: EW no 30$ oedN 9 oeﬂ 9:53 30:20am

III 80:35 .805 SE dd.“ 823%: 5.: EE dd— cameo—03$?

coma:— O<>ﬁ Z .6023 026% c2100 Sarcasm

 

BE 9362 o5 ooaoe 5&8 so 3233 so. 2: do 2:8 - «a 2.3.

 

29
2.6 Excimer Lasers

There are certain molecules (KrF, ArF, XeF, XeCl etc) that can exist only in excited state
and in the ground level they dissociate into their individual atoms. For such molecules, the
potential energy curve has no local minimum in the ground level and hence there is no stable
state for ground level [42]. Such molecules are called as excited-state dimers or excimers.
Figure 2.8 indicates a transition from the excited state to the ground state, where excimer
molecule dissociate very quickly into two unbound atoms. The most signiﬁcant feature
of excimer laser is the presence of favorable lifetime ratio of upper and lower level lasing
species. Excimer lasers emit light that extends from the visible to the ultraviolet portion
of the spectrum, depending on the particular excimer molecule. Because of the bound-free
nature of the laser transition and lack of well-deﬁned vibration-rotational transitions (as
compared to C02 laser), excimer lasers are highly tunable over a considerable range of

wavelengths.

Kr"+I-"

Kr'+F

b Excited states

 

A ‘g%

woe-103m

 

r groundsmemnsrable)
\————-Kr+F

 

‘
r

 

Inter-nuclear Separation

Figure 2.8 - Schematic of the energy-level curves for the KrF molecule [15].

30

Rare gas mono-halides, more speciﬁcally KrF excimer laser at around 248 nm, is one of
the most efﬁcient high-power ultraviolet laser that is most commonly used. For these rare
gas mono-halide excimer lasers, a large number of reactions are involved in the population
inverse process. In Figure 2.8, completely unbound ground states are marked as A and
X and the low-lying excited states are indicated by B and C. The B -) X is the most
commonly observed transition for the KrF laser. A typical mechanism for excitation is as
follows:

Kr+e’ =>Kr‘+e’.

This excited krypton atom has an electron in an outer shell that makes it behave like a

metal, which thereby reacts readily with a ﬂuorine molecule forming excited KrF (KrF‘).
Kr" +F2 => KrF‘ + F.

The excited dimer (KrF‘) can also be created by ionized atoms and molecules as indicated
below:

F2 + 6 => F’ + F dissociative electron attachment reaction

and Kr‘" ions are provided by:
Ar+ + 2Ar => Ar; + Ar

Ar; +Kr => Kr+ + 2Ar

molecular Kr; ions can also be formed by: '

Kr+ +2Kr => Kr; +Kr

31

which react with F ’ in an ion-ion recombination to form KrF'.
Kr’r + F" => KrF".

In a typical high-power KrF laser a gas mix of 90% Ar, 9.5% Kr and about 0.5% F2 is used.
The reaction kinetics for gas halides are very complicated due to hundreds of alternate
paths available for the formation of excited dimers. Detailed discussion about applications

and current trends of excimer lasers can be found in literature [36, 37,43—49].

2.7 C02 Laser

Four—level pumping schemes typical of many commercial lasers (C02 laser, Nd:YAG laser,
semiconductor lasers etc.), will be considered from the point of view of C02 laser. This will
also serve as an appropriate starting point for a general discussion of lasing phenomena. In
addition, C02 laser has great industrial signiﬁcance, since it is the only type of laser that is
being routinely manufactured upto power levels of 20 kW.

2.7 .1 Energy Levels

Population inversion can also occur if the lifetimes of the lower levels are very small com-
pared with those of the higher energy levels, as in C02 laser (Figure 2.9). A simpliﬁed
energy level diagram exhibiting the most important vibrational energy levels for C02 laser
operation is depicted in Figure 2.9. Each of these vibrational states (indicated by (001),
(100), (020), and (010) in Figure 2.9) is further spilt into a number of rotational states.
The notation (V1V2V3), where V1, V2, and V3 are integers that denote the quanta of energy
in each of the three vibrational modes (see lower half of Figure 2.9). The anti-symmetric
modes, V2 and V3, cause a change in the dipole moment of the C02 molecule, thereby

making it possible for these vibrational modes to be excited by direct absorption of an

32

 

 

   
  

 

 

 

 

 

 

 

 

 

 

   

[— (001) 18 crn'l ‘1 0'3
Laser transititfls (ll I (1)
2000 "" Vibrational energy
_ transfer by collision
'E _) 0.2
U
”I:
.3 (100) 9
g a (020)' 3
= .53. 9
g 1000 — a o
e 2 5 g 01 a
‘3 010 'a a “
E __ (ter(minal g g
B by collision level) 5 g;
g with He or by radiativ
state V 1 W2 V3
(000) coz Levels N2 Levels
V1 0 U 0 symmetric vibration
H H
V 2 I . I bending vibration

 

 

v. 0 H asymmetric vibrationI
e—v <——-> 4—0

 

Figure 2.9 - The various energy levels for C02 laser [18, 19, 22].

33

infrared photon (infrared—active fundamentals). On the contrary, the symmetric mode V1
involves no change in the dipole moment of the molecules and hence is infrared inactive
or Raman active. The vibration levels (shown in the upper half of Figure 2.9) derive from
the three fundamental modes shown in lower half of Figure 2.9. The energy levels for the
diatomic N2 molecule are also shown in Figure 2.9. This diatomic molecule has only one
vibrational degree of freedom and one vibrational quantum number. Hence N2 molecule
can be efﬁciently excited to its upper energy level, because of its simplicity. Because of
the negligible energy difference between this upper level and (001) level of C02, energy
is efficiently resonantly transferred between these two molecules via collisions. This is the
main excitation method for the (001) level of the C02 laser. The active media should also

exhibit the following characteristics, for lasing action to be practical [15].

o In a four-level laser (e. g. , C02) one must prevent the terminal level [(010)] from
being overpopulated by electrons raised from the ground level [(000)] by thermal
ﬂuctuations. Hence the separation between (000) and (010) level should exceed kT
(k is Boltzman’s constant and T is absolute temperature in degrees Kelvin), otherwise
the laser must be refrigerated?

o The terminal level, (001) should be prevented by being overpopulated by the de— .
excited (by spontaneous emission) molecules, since these molecules would not be
available for excitation from the ground level. In the case of C02 laser, He is added
to the discharge to encourage collisional de-excitation of the C02 molecules from
the terminal level. The He gas, because of its high thermal velocity provides many

opportunities for collisions and thus the increased deexcitation rate. He also aids

 

2The equilibrium population of the terminal level as given by Boltzman’s relation is
El _ —AE
N0 ‘ up kT
where AB is the energy gap between terminal and ground level. If AE >> kT, then N1 /N0 << 1, and the

terminal level will always be relatively empty. In a few laser materials, the energy gap AE is relatively small
and, therefore they must be operated only at cryogenic temperatures to function as a four level laser system.

34

in enhancing the rate of excitation of N 2 molecules to the upper vibrational energy
level. Gases such as N2 and He which aid the lasing process but do not emit photons

at the lasing Wavelength are known as buﬁ’er gases.

0 The top excited level should be a broad band [(001)]. This wide band is a practical
necessity since there is not enough energy available from ordinary excitation sources
of radiation in a narrow band. This condition need not be satisﬁed if another laser

source is used to excite a second laser.

The laser resonator not only maintains a large enough electromagnetic ﬁeld strength to
stimulate emissions from excited ions/molecules, but also operate as a feedback mechanism

and thus increase the coherence of the output beam. In fact, the resonator cavity serves to:

1. improve the gain of the lasing media (by multiple passes and by increasing the effec-

tive length of the cavity).
2. enhance monochromaticity of the output beam (by Fabry—Perot mode selectivity).

3. produce a coherent output beam, partly due to (1) and (2).

2.7 .2 Method of Excitation

The primary methods of exciting gas lasers are: DC, AC or AC with a DC bias, and RF
discharges. He-Ne lasers are DC—excited. Many C02 lasers are DC—excited, although RF
excitation is becoming popular for many high power industrial lasers. An RF-excited laser
resonator is relatively light and contains essentially no moving parts and therefore makes no
vibrational contribution to the local environment. The laser head can be easily positioned

by a variety of mechanical motion systems that are both very precise and very dependable.

35

2.7.3 Types of C02 lasers
2.7.3.1 Slow Axial Flow C02 Laser

The slow axial ﬂow C02 laser produces in excess of 100 W/ m of discharge. Because of
simplicity of design, this type of laser is easy to build. Basically the conventional CW
C0; laser is a water (or dielectric) cooled narrow bore tube with mirrors on either ends,
through which C02:N2:He gas mixture is circulated slowly at around 20 to 30 Torr and
axially excited by an electric discharge, as illustrated in Figure 2.10. Since cooling of the
resonator tube is by thermal conduction to the liquid ﬂowing through the outer jacket, the
tube bore size is small. This limits the amount of active material and thus the power output.
But the side effect of this is to produce a low—order beam of either TEMoo or TEMm.
(see Section 2.2). However, achieving stable, TEMoo Power levels of 100 W and higher
requires an extremely stable resonator structure. Slow ﬂow axial lasers up to 1500 W are
constructed by simply adding a number of discharge sections in series, creating a long

optical path.

2.7.3.2 Fast Axial Flow C02 Laser

As mentioned above, slow axial ﬂow C02 lasers can be constructed upto power levels of
about 1500 W. This power limit is imposed by subtle laser physics such as gain saturation,
gas disassociation, and thermal effects. These effects are bypassed by a fast axial ﬂow in
which the gas is cooled (with the help of heat exchanger) and reused. Since this convective
cooling is more effective in extracting heat than the conductively cooled slow—ﬂow design,
the cavity inner diameter can be much larger and hence more active medium is available
for a given tube length. Fast axial ﬂow lasers can be constructed to power levels of about
5000 W. Fast axial ﬂow lasers produce in excess of 1000 W/ m of discharge. Therefore,
commercially available 1500 W lasers are much compact than their slow axial ﬂow lasers.

The high mode quality inherent in axial ﬂow lasers is maintained in the fast axial ﬂow

36

 

0.05 lem

  

     

Discharge-tube 3

0.7 kW/m

>10.0 kW/m

     

Axial-ﬂow

    
  

fllﬂla
IOIIl-IIOUIII

   
 

Transverse-flow

  

Schematic of the various types of lasers [50].

Figure 2.10

37

design. Expensive roots type blowers are used to move the lasing gas at near-sonic speed,
thereby, increasing the operational cost. Fast ﬂow lasers are available both in pulsing and

continuous wave mode.

2.7.3.3 Helical Flow C02 Laser

The helical ﬂow laser is a hybrid between slow ﬂow axial lasers and transverse gas ﬂow
lasers. Helical ﬂow lasers produce in excess of 400 W/ m of discharge. Helical ﬂow lasers
can be constructed from 500 W up to 6000 W by adding discharge sections in series, sim-
ilar to axial ﬂow lasers. The helical ﬂow laser tube itself is of complicated construction,
utilizing hundreds of electrically— isolated sections that must withstand high plasma tem-
peratures and high thermal expansion. Since the discharge is aﬁ'ected by several variables,
such as gas pressure, gas mix, and discharge current, the discharge must be constrained
by a large magnetic ﬁeld. These lasers are usually of ﬁxed—power designs. Output beam

quality is a compromise between axial ﬂow and transverse ﬂow lasers.

2.7.3.4 Cross (Iransverse) Flow C02 Laser

A cross ﬂow laser is a laser in which transverse gas ﬂow and axial electrodes allow power
levels of 1.0 kW to 15.0 kW to be generated from a relatively small length by using large
mirrors to permit beam folding. The gas ﬂow rates can vary anywhere from 60 m / s to
supersonic speed in some of the newer designs. The lower half of Figure 2.10 depicts a
typical transverse ﬂow C0; laser. It is common to ﬁnd both, stable and unstable resonator
design in this type of lasers. Unstable resonator design is preferable for power levels above
6.0 kW, because of the absence of a transmissive output mirror. These type of lasers are
extremely compact with power output in excess of 3000 W per meter of laser cavity. Cross
ﬂow lasers typically have higher—order modes that yield beam outputs that are typically
non-Gaussian. This is due to the fact that the discharge is free to move around in the beam
path,'unlike axial ﬂow lasers in which the discharge is contained in plasma tubes. Cross

ﬂow lasers are also termed “Gas Transport Lasers” (GTL) by some manufactures.

38
2.8 Applications of C02 laser

The nature and application of the interaction between the C02 laser beam and the material
is controlled by the power density and interaction time, in addition to a number of other
processing variables. A single industrial C02 laser can be utilized for numerous applica-
tions by manipulating these two parameters alone [51]. The type of application as dictated
by the incident power density and interaction time is shown in Figure 2.11. A particular
set of power density and interaction time determines the nature of the laser process as indi-
cated in Figure 2.11. In the order of decreasing power density, the various laser processes
include shock hardening (high power density — low interaction time), drilling and metal
removal, laser-glazing, deep penetration welding, surface alloying, cladding, and transfor-
mation hardening (low power density -— large interaction time). The amount of energy that

is applied to the material is of less signiﬁcance as compared to the interaction time and

 

power density.
1010 \
\\\\\\ \
\
1091- §\\\\\\\ \\ \
\ \ SHOCK\\ \ \
‘§ HARDENING \ \\ \ \SPECIFIC ENERGY. J/cm2

:\\\\\\\\ \\ \\ \\\ \\\
\\ \102\\\\\\\\ \\ \\105

‘ DRILLING

\\ “i \\\\\\\\\ \\
\ useacuzmc\§ PﬁxéJfFEfEEON\ $\\ \\

\\\\\\\\\\\\\
104l- \\ \\\\\\;\\\\\\\

\\ T\RANSFORMATION‘
l l l HARDIENING \\
103

10‘8 10'6 10" 10'2 10°
INTERACTION TIME - sec

. i/

.5
O
‘1

POWER DENSITY — chm2

514

 

 

 

Figure 2.11 - Laser material interaction energy diagram [52].

39

2.8.1 Marking

A C02 laser beam produces a clean, smooth cut which usually requires no further ﬁnishing
steps. Sheet materials and ﬁlms may be cut without stretching and distortions that are
usually associated with mechanical holding and cutting pressures. Fabrics do not fray at
the cut edge as is characteristic with mechanical cutting methods; no troublesome machine-
jamming lint is produced. Cutting and perforating may be performed alternately without
changing tools. With many organic materials cutting and sealing of the material’s edge may
be performed in one smooth, continuous operation. Difﬁcult materials, such as adhesive-
backed materials, or two— and three-ply materials with an adhesive inter-layer are simple

to cut with C02 laser.

2.8.2 Scribing, Etching, and Drilling

C02 lasers are able to scribe and etch a variety of glass, ceramic, and other brittle materials.
Chipping and breakage are virtually eliminated, and no abrasives, slurries or hazardous
acids are used in the process. In one practical application, aluminum aircraft skin panels
(Boeing 777) are being scribed with a 140 W C02 laser to assist in chemical milling [53].

C02 laser beams can be focussed to only a few thousandths of an inch in diameter.
As a result, holes with very precise tolerances can be drilled. Because there is no tool
contact, these critical tolerances can be maintained indeﬁnitely and puncture distortion in

the material is eliminated.

2.8.3 Fusing and Welding

The intense localized heat that is produced by a C02 laser makes it ideally suited for fusing
and welding plastic materials as well as ceramics and glass. The cleanliness and localized
nature of the laser beam keeps the heat affected zone to an absolute minimum size and the

surrounding areas are not affected at all.

40

The technique of laser beam welding of materials involves the rapid interaction of ma-
terial with a focussed laser beam by traversing the material under a stationary beam or
vice versa. This procedure yields an elemental volume of melt at the surface, while the
substrate maintains its ambient temperature. Rapid melting occurs in a very short span of
time, during which a negligible amount of thermal energy is conducted to the substrate,
thus giving rise to steep temperature gradient between the solid substrate and the molten
phase [54]. Rapid solidiﬁcation occurs as a result of this steep temperature gradient. A The
solidiﬁcation and quench rates that are involved in LBW are mainly dependent on the pro-
cess parameters such as, the beam power and intensity, traverse speed, interaction time, and
pressure and type of cover gas. The microstructural changes which occur by these rapid
quench rates range from the reﬁnement of the solidiﬁcation microstructure to the enhance-
ment of the solid solubility and formation of non-equilibrium metastable phases, and in
more extreme cases, to the complete suppression of crystallization with the formation of
amorphous phase.

Carbon dioxide lasers are being used in a wide variety of other applications. Some of
the better known applications are: drilling baby bottle nipples, cutting contact lenses to
size, trepanning rubber seals, heat—treating metal components, stripping wire and coaxial

cables, and drilling polyethylene irrigation piping [32, 36, 37, 55—57].

2.9 Summary

The nature of application and choice of material dictates the selection of type of laser
required for material processing. Among the various types of lasers, only C02 lasers,
Nd:YAG lasers, and excimer lasers are being used for industrial scale applications. Despite
its large wavelength and multi-mode beam (for higher power industrial lasers), C02 lasers
are ﬁnding extensive utilization in material processing because of their cost effective power
levels (on per watt basis), ease of operation, and availability of maintenance free high power

optics.

CHAPTER 3

PROCESSING OF MMCs

MMCs are a class of materials designed to integrate the strength, ductility, and formability
of metallic alloys (e. g. aluminum, magnesium) 'with the speciﬁc modulus of non-metallic
materials such as SiC, A1203, B4C, Si3N4, AlN, carbon and graphite ﬁbers etc. [5,58—67].
Fundamental to the understanding of these MMCs is in knowing the characteristics that dif-
ferentiate MMCs from their monolithic alloys in terms of fabrication, physical and mechan-
ical properties, matrix/particle interfacial reactions, aging kinetics, segregation, strength-
ening mechanisms, and other metallurgical aspects. Among the various MMCs, SiC and
graphite ﬁber reinforced Al based composites have been widely studied. Developments
have centered around Al alloy matrices, since they have the ability to resist corrosion and
oxidation, are good conductors of heat and electricity, have low density, and would be-
come more attractive if stiffness, strength, fatigue, wear resistance, and high temperature

properties are increased by reinforcement [68-70].

3.1 Fabrication Techniques

MMCs are manufactured by a variety of techniques. These manufacturing methods can be
broadly classiﬁed into [71]: solid state, solid-liquid, liquid phase, and In Situ fabrication

techniques. A different approach classiﬁes the fabrication of MMCs as follows:

41

42

o Squeeze casting (also known as squeeze—inﬁltration, melt inﬁltration, pressure inﬁl-
tration, or liquid metal inﬁltration) [72—7 8]

o Diffusion bonding (a variation of this is roll diffusion bonding) [79]

o Superplastic forming

a Compocasting (semi—solid slurry or Rheocasting) [80]

0 Spray forming or spray deposition [79, 81]

o Electrodeposition and plasma spray deposition

0 Laser beam processing [82—89]

The reinforcement can be either continuous ﬁbers or discontinuous irregular particles
or whiskers. In the recent past the emphasis in this ﬁeld had been in the development of
MMCs with discontinuous reinforcements to lower the cost of manufacturing, by utilizing
conventional melting and casting techniques [2,68,90]. The major problem associated with
liquid processing techniques for discontinuously reinforced MNle is in ensuring sufﬁcient
wetting of the ceramic reinforcement without excessive chemical or mechanical degrada-
tion of the reinforcement properties. Moreover, majority of the fabrication techniques listed
above generate inhomogeneous distribution of the reinforcement in the MMC. Some of the
more popular fabrication techniques are illustrated in Figure 3.1 and Figure 3.2. A more

detailed review of the various fabrication techniques can be found in Girot et al [68].

3.1.1 Matrix Materials

In general pure metals are rarely used as matrix in MMCs. The most commonly used matrix
materials are aluminum [58,91-93] based alloys (2024, 2124, 5156, 6061, 7075, 7090, and
7091), magnesium alloys, titanium alloys [94, 95], copper, Fe, and Ni [96, 97]. Of these,
aluminum based alloys are the most commonly exploited matrices in MMCs, because of
their low density and reasonably high thermal conductivity. The main purpose and role of

the matrix material in the composite is:

43

      

     

7/2 74;
\\\ \\\\\ \
>\\§>\§§§§§\\

  

 

 

 

Figure 3.1 - Schematic representation of (a). powder metallurgy process for MMC fabrica-
tion [98] and (b). spray deposition process [98].

 

liquid "‘ F.doucpmulld

alloy

   

 

 

 

 

 

 

liquid alloy induction hunting

9 ﬂuor-

b‘l / continuou-
pawl-ions."
in alloy and "Don
C
e...
O
O

 

 

 

OO 00

 

cool. :-

 

 

O 0000

domi- col id alloy
0 "bar.

(I)

portions . libero

 

 

' I——0I....P and could!-

 

 

 

 

 

drawer

[:7——'-- .....
(5)

(II)

Figure 3.2 - Schematic illustration of (I)(a). continuous compocaster, (b). monocharge
compocaster [68]. (II) Illustration of squeeze casting process [68]: (a). direct
inﬁltration of a preform, (b). squeeze casting of a compound, and (c). indirect
inﬁltration of several preforms.

45

1. Transfer load between the reinforcements
2. Provide ductility

3. Provide transverse strength in ﬁber reinforced composites

Early in the development of discontinuously-reinforced aluminum matrix composites,
the choice of the matrix alloy was dictated more by the intrinsic characteristics and pro-
cessing ability (casting, powder metallurgy, forging, extrusion, rolling) of the matrix rather
than the type and properties of reinforcement utilized in the composite [68]. With increas-
ing body of knowledge in the area of reinforcement—matrix wetting behavior [99—101],
interfacial reactions [102—104], aging response [105,106], and strengthening effect of the
particulates [107-110] (for example, the overstraining of the matrix alloy lattice [62, 111]),
the focus of research has shifted to adjustment of the matrix composition with respect to

the reinforcement type.

3.1.2 Reinforcements

The most prominent reinforcements are SiC, graphite, carbon, A1203, boron, B4C, W ﬁl-
aments, Si3N4, Tle, and other materials. Reinforcements are used either in continuous
(ﬁbers) or in discontinuous (whiskers, particulates, chopped ﬁbers, and platelets) form
in the metal matrix. Traditionally, continuous reinforcements are expensive to produce,
leading to a very slow adaptation of MMCs in non-aerospace applications, despite their
numerous attractive mechanical and high temperature properties [112].

Before the advent of cheap pyrolizing rice hull process for producing ﬁ-SiC whiskers
(Sij), SiC ﬁbers were manufactured by a more expensive chemical vapor deposition
(CVD) [113—115]. In the CVD process, a pyrolitic graphite coated carbon monoﬁla-
ment substrate is exposed to silane and hydrogen gases, producing 140 pm diameter SiC
ﬁbers [58]. SiC,” produced by rice hull process have an average tensile strength of 1000
ksi, as opposed to CVD SiC ﬁbers with an average tensile strength of 550—600 ksi [116].

46

“0th the introduction of SiC,”, extensive research have been carried out to process discon-
tinuously reinforced MMCs, with particulate ceramic reinforcements, such as SiC, A1203,
and B4C in Al alloy matrix [104, 109, 117—119]. In addition to the inexpensive foundry
techniques used for fabrication of particulate MMCs. other conventional practices like ex-
trusion, forging, rolling etc. can be efﬁciently exploited to produce semi-ﬁnished particu-
late composite products. These conventional techniques can be applied to composites with
up to 40 vol% particulate reinforcements [120].

The characteristics of the reinforcement dictate the optimal properties of the composite.
In the case of particulate reinforced composite, particle morphology, particle size distribu-
tion, dispersion uniformity, surface chemistry, volume fraction, particle shape, particle wet-
tability are some of the factors that affect the properties of the composite [121]. Regardless
of the type of reinforcements, MMCs in general exhibit enhanced physical and mechanical
properties that are not achievable in monolithic alloys. Some of these improved properties

include:

0 Strength (both tensile and shear) and stiffness [59, 107, 108, 110, 122—125]
0 Wear [68,126] and fatigue resistance [127—129]
0 Corrosion resistance [69,130]

0 High temperature properties [68, 131, 132]

3.2 Physical Properties

Numerous mathematical models are available for predicting various physical and mechani-
cal properties of MMCs from the pr0perties of the individual components that comprise the
MMC. The most widely used model is the rule of mixture (ROM) approximation, by which
numerous parameters like coefﬁcient of thermal expansion (CTE), density, speciﬁc strength

and modulus, thermal conductivity etc. can be computed from the weighted average of the

47

individual components as shown below:
ac : (7me +0r‘fr (31)

where a is the property of interest, ‘V’ is the volume fraction, and the composite, matrix,
and reinforcement are denoted by subscripts ‘c’ , ‘m’ , and ‘r’ respectively. Limitations of
this simple weighted average relation (ROM) lead to numerous reﬁnements to account for
the non-isotropic nature of high aspect ratio reinforcements and the thermal barrier effects
at the matrix/reinforcement interfaces. Taking into consideration the effects of isostatic

stress ﬁelds, it was proposed that the CTE (a) of the composite can be computed by [120]:

_ (ameKm + arm-Kr)

C" " (Vme + V,K,) (3‘2)

 

where ‘K’ is the bulk modulus of the individual components. The a values computed by
Equation 3.2 are signiﬁcantly lower than those predicted by ROM approximation. A further
reﬁnement to Equation 3.2 was proposed to account for the effect of shear stresses between

the matrix and an approximately spherical and isotropic reinforcement:

ac = Crm — Vr(am —’ (151-)5 (33)

Q

where:

P = Km(3Kr + 41011? + (Kr — Km)(16u3n + 1210.10)
Q = (3K, + 4pm)(4V,pm(Kr — Km) + 3KmKr + 4umK,)

and ,u is the shear modulus.

CHAPTER 4

MACHINING OF MMCs

Laser systems intended for either stand alone use or for integration into existing production
cells, are becoming commonplace. C02 lasers provide numerous advantages over conven-
tional machine tools for the processing of a large variety of materials. Cutting, drilling,
welding, soldering, melting — all are accomplished quickly and precisely without tool con-
tact or tool wear. Laser cutting is a non-contact, thermal process. These two characteris-
tics are fundamental to its utility. Further, since the tool never comes in contact with the
work, all of the problems associated with tool loading are avoided. Chatter marks, material
galling, tool pulling - all of these problems associated with traditional machining methods,
are avoided by the lack of contact between the tool and the work. The absence of contact
also allows intricate cutting of fragile work-pieces and simpliﬁes work-holding devices.
Thermal cutting is independent of the strength or hardness of composite constituents. The
thermal nature of laser cutting, however, limits its use when charring or thermal degrada-
tion are unacceptable. In general the following materials have been found to be amenable
to processing with C02 lasers: ceramics, rubber, composites, stone glass, plastics, natural

ﬁbers, tissues, organic materials, and wood.

48

49

4.1 Machining of Polymer Composites

Polymer composites have attracted considerable interest because of their extensive use in
aircraft and automotive industry [133, 134]. Optimization of the following laser parameters

is essential to achieve good quality cut and surface ﬁnish [133, 135-137].

1. Input laser energy or power density.

2. Diameter of the focusing spot.

3. Scanning speed/velocity or interaction time of the laser with the material.

4. Degree of polarization of the laser beam which leads to curved grooves in the laser
machined parts [138, 139].

5. Pressure and ﬂow rate of the shroud gas or cover gas.

6. Clearance between the nozzle tip and surface of the workpiece.

In addition to the above beam characteristics, the physical properties of the workpiece
also dictate the nature of the laser interaction with the material. The following material

properties are worth mentioning [133, 136]:

1. Absorptivity of the material at the given wavelength of the laser.
2. Thermal diffusivity, heat capacity, and density of the material.
3. Latent heat of melting and vaporization.

4. Melting and vaporization temperatures.

Based on various logical assumptions, analytical models are developed to predict the ther-
mal damage zone [133] and optimal cutting speed [140]. The thermal anisotropy model ex-
plains the high ablation temperature of the carbon ﬁber. Mathematical modeling for predict-
ing a single beam pass groove depth in a homogeneous material is also reported [141, 142].-
Investigation on three dimensional machining of Carbon/Teﬂon and GFRP composites has
been studied in the literature [143]. An analytical model to predict the groove depth for a

50

composite material for multiple beam passes is also reported [143]. Laser drilling and cut-
ting of glass, carbon, and ararnid ﬁber reinforced epoxy composite materials, using excimer
and C02 laser, indicates that excimer laser is better suited for polymer-based composites,
in terms of quality of cut but the material removal rate is lower when compared to C02

laser beam machining technology [134].

4.2 Machining of Metal Matrix and Ceramic Composites

Materials that resist conventional means of machining can be laser cut with good results.
Using air-assisted ﬂow, good quality cut is obtained in the case of l—mm thick boron—
aluminum composite. It is also reported that a YAG laser provides a good out than a C02
laser'for B-Al composite [35]. SiC-Tl composite is also successfully cut, using argon as
the shroud gas to prevent oxidation of Ti matrix. C02 laser is used in this case [35]. One of
the biggest problems in the production of ceramics is the difﬁculty in machining materials
such as Si3N4, SiC, Zr02, and boron carbide.

Candidate materials like SiC, Si3N4, and Zr02 for use in adiabatic engine are machined
at high temperatures (hot machining) using C02 laser [144]. The preliminary investigation
for these materials show great potential for laser beam machining. Preliminary work on
silicon based ceramics indicates that these ceramics can not be machined without intro-
ducing major cracks at room temperature [144]. R. F. Firestone and E. J. Vesely Jr. [144]
have concluded that LBM is 10 to 30 times faster than diamond machining for this class of

ceramics.

4.2.1 Ceramics Processing

For ceramics substrate machining, high throughput, and minimal HAZ are important fac-
tors. Lasers can scribe, cut, and drill these substrates on a continuous basis irrespective of

their material hardness.

51

Holes drilled with a de-focussed beam in alumina, should be limited to approximately
0.015 inch (0.38 mm) to prevent fracturing from thermal shock. Larger holes from 0.020
inch (0.51 mm) to 0.20 inch (5.1 mm) in diameter can be cut using rotating lens accessory.

l-dimensional laser machining is performed by scanning the focussed laser beam over
the surface of a material inorder to melt or evaporate the region beneath the spot, forming a
“kerf”. Ideally the laser cutting process would be purely evaporative, continuously convert-
ing solid material to vapor by direct sublimation. The actual process involves a combination
of melting and vaporization and depends on the combined effect of beam intensity and in-
teraction time with the workpiece, as well as material properties and the characteristics of

cover gases.

4.3 . Energy Transfer Mechanism in Laser Machining

The energy of the laser beam is dissipated in four ways as described below [145]:

1. Part lost due to reﬂection (a major part of the focussed laser beam energy is lost by
highly reﬂective metallic surfaces). The laser absorptivity depends on wavelength of

laser radiation, temperature of the material etc.
2. A large portion of the remaining unreﬂected energy is utilized to melt the material.
3. Evaporation of the liquid material consumes a small part of the remaining energy.

4. A very insigniﬁcant part of the energy is conducted away into the unmelted base

material, provided the thermal conductivity of the base material is low.

The thermal and optical pmperties of the workpiece dictate the magnitude of each of the
above four phenomena. The evaporated liquid metal forms a plasma over the surface of the
workpiece and depending upon the power density in the beam, the laser power can initiate
an ionized metallic gas [146]. The importance of this plasma has not yet been quantitatively

assessed. According to one school of thought, it is necessary to have a stable plasma

52

located just beneath the surface in order to obtain deep penetration [146]. The surface
plasma increases the overall coupling between the beam and the workpiece. The surface
plasma increases both the convective and radiative heat transfer coefﬁcients. It is also
suggested that, the surface absorption coefﬁcient at 10.6,um (for C02 laser) may actually
decrease with temperature. But this decrease in absorption coemcient is compensated by
the conversion of infrared radiation to ultraviolet radiation (re-radiated) within the plasma
Hence, the heat transfer rates associated with the hot impinging shroud gas that has been

heated as it ﬂows through the plasma region will be increased [146].

4.4 Other N on-traditional Machining Methods

There is no clear-cut deﬁnition for non-traditional machining processes, because of the
diverse techniques included in this category. In general, they can be deﬁned as those pro-
cesses that came into existence in the last 50 years and which use common energy forms
in novel ways or those which use new forms of energy [147]. Depending on the way the
energy is employed, non-traditional machining process can be broadly subdivided into:
mechanical, electrical, and thermal methods. The list of the various processes that fall
into each of these methods are shown in Table 4.1 and the broad subdivisions are brieﬂy
discussed below [147].

Mechanical Methods: In this case, material removal is achieved by the abrasive action
of the mechanical forces developed in various ways. Water-jet and ultrasonic machining
are the main categories. In ultrasonic machining, material is removed by the erosive effects
of granular particles undergoing ultrasonic vibrations. High pressure water-jet provides the
mechanical energy for material removal in water-jet cutting, whereas incorporation of abra—
sive particles provides enhanced mechanical action in abrasive water-jet cutting. Materials
that can be damaged due to burning, charting, or cracking (typical of thermal processes)

and. those which are not electrically conductive (a requirement for electrical methods) can

53

Table 4.1 - Classiﬁcation of non-traditional processes [147].

 

 

Mechanical Thermal Electrical
Abrasive jet machining Electrical discharge machining Electrochemical
(AJM) (EDM) machining (ECM)
Abrasive ﬂow machining Electrical discharge wire Electrochemical
(AFM) cutting (EDWC) grinding (ECG)
Waterjet machining Electrical discharge grinding -

(WJM) (EDG) ‘

Abrasive waterjet machining Electron beam machining —

(AWJM) . (EBM) -

Ultrasonic machining Laser beam machining -

(U SM) (LBM) —

 

 

 

 

 

be successfully processed by this method. Materials which are difﬁcult to machine because
of high hardness, toughness, and brittleness are ideal candidate materials for this method
of machining [147].
Electrical Methods: These methods are for electrically conductive materials only. A large
percentage of materials that are difﬁcult to machine by conventional techniques are being
machined by electrical methods because of their ability to produce complex shapes in a
single pass of the tool. Electrical methods process parts without any tool wear [147].
Thermal Methods: Because of the diversity of energy sources grouped into this cate-
gory, it is difﬁcult to make any generalization of the applications of these processes. This
method is the fastest growing of all non-traditional methods, as witnessed by the rapid sales
of EDMs and laser equipment. The only drawback of these processes is the presence of
heat affected zone in the work piece [147].

Some of the more popular non-traditional machining methods are discussed in the fol-

lowing section.

54

4.4.1 Water-jet and Abrasive Water-jet Machining

Water-jets operate at high pressures and high velocities of about 50,000 psi and 900 rns‘1
respectively (approximately Mach 3). They are being used to cut titanium and other difﬁcult
to cut space-age materials. Water-jet cutting is not as productive as cutting with plasma or
laser beams, since water loses energy as it leaves the nozzle, thereby, slowing the process.
However, it does produce quality parts and provides ﬂexibility through the wide range of
materials it can shape. With pure water-jet, it is possible tocut only non-metallic materials.
Machining of hard and dense materials like metals, glass, and ceramics require the addition
of ﬁne sized abrasive particles to the water-jet stream.

The basic equipment utilized in water-jet machining is outlined in Figure 4.1. An elec-
trically driven hydraulic pump is used to supply oil at high pressures of 117 bar, to drive
a reciprocating plunger pump, known as an intensiﬁer. In the intensiﬁer, water let into the
low pressure chamber (at 4 bar) is expelled through the high pressure chamber at pressures
of 380 MPa (55 ksi). An accumulator is provided to smooth out any ﬂuctuations or spikes
in the high pressure water ﬂow. The accumulator can maintain uniform pressure and jet
velocity, because of the compressibility of the water, which is roughly about 12% at 3800
bar. The pressurized water is carried through transmission lines, which generally consist
of stainless steel tubing (diameter range from 6 to 14 mm) swivels and ﬂex joints. The
cutting action is controlled by the on/off valve. For abrasive water-jet, abrasive particles
are added to high pressure water-jet in the mixing chamber. The particles are sucked into
the chamber, because of the high pressures. In the mixing chamber, part of the momen-
tum of the water-jet is transferred to the abrasive particles, whose velocities increase to
300 - 600 m/s with ﬂow rates of about 10 g/s. This high pressure and velocity jet ladened
with abrasive particles (garnet is routinely used) exits through sapphire or diamond noz-
zle (oriﬁce size typically range from 0.075 to 0.75 mm). A focussed and coherent high

velocity water-jet exiting the accelerator nozzle performs the cutting action by a complex

 

 

I 1 =2 1 F11 I Went

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Todninorrecycie

Figure 4.1 - Schematic representation of water-jet cutting system [148].

erosion phenomena. The parameters that inﬂuence the cutting action are: water-jet pres-
sure, water-jet oriﬁce diameter, mixing tube length and diameter, type of abrasive material
and particle size, ﬂow rate, traverse speed, angle of cutting, and the type of material to be
cut [149]. Eventhough waterjet cutting has been accepted as a method to machine polymer
composites in aerospace industry, there a few inherent drawbacks in terms of quality of
cut. Waterjet cutting utilizes a shearing force to initiate and continue the cutting process.
Delamination is a major concern when machining polymer composite laminates. In water-
jet cutting, the excessive water pressure can penetrate between the ﬁbrous layers and cause
delamination [56,150]. Some of the other advantages and disadvantages of AWJ include
the following: ,

Advantages: l). Cuts without heat, which eliminates thermal distortion, localized struc-
tural change, and thermally induced oxidation in specialty metals like titanium, nickel, and

56

cobalt based alloys, 2). Cuts about 2 inch thick material with a clean and ﬁnished edge,
when compared to LBM. 3). Omni-directional cutting with a zero radius for outside cor-
ners and a radius equal to the radius of the jet for inside comers, and 4). Smaller kerf width
when compared to other non-traditional machining processes (except LBM).
Disadvantages: 1). High noise levels, in the range of 85-95 dBA. 2). Because of high
pressures involved, it leaves residual stresses in the workpiece and thin specimens deform
when out with such a high pressure. 3). Material removal rates are lower when compared to
LBM. 4). High initial cost of the equipment. It has a capital cost in the range of $117,000
to $300,000 depending on size, number of axes, and type of robotic control.

4.4.2 Electro-Chemical Machining (ECM)

Electro—chemical machining process relies on the controlled removal of atoms from the
surface of the workpiece, unlike Ion-Beam machining (IBM) where atoms are removed
by transferring momentum of the impinging ions to the workpiece surface. ECM is an
electrolytic process with removal of material being achieved by anodic dissolution in an
electrolytic cell.

Familiar applications of electrolysis are the electro—plating and electro-polishing pro-
cesses. In electro-plating, coatings of less than 10pm are achieved, with 10‘2Acm‘2
current density, upon the surface of a cathodically polarized metal. ECM is similar to
electro—polishing in that, the metal to be polished is made the anode in an electrolytic cell,
to achieve preferential dissolution of the anode. For both electro-polishing and electro-
plating, the electrolyte is either stationary or in motion at low velocities.

Anodic dissolution rate, according to Faraday’s laws of electrolysis, depends on the
atomic weight ‘A’, valency ‘z’ (of the anode), current ‘1’ which is passed and the duration
of time ‘t’, for which the current passes. ECM is independent of the hardness (similar to

LBM) or other characteristics of the anode metal. But the workpiece must be electrically

57

conductive for ECM process. The principles of ECM are outlined in Figure 4.2 [151].

 

Cathode- tool (brass)
feed-direction (rate 0.02 ms“)

 

>

'U

'2.

a mu.
<3 gap
8 O Electrolyte ﬂow (10 ms")

H 0 _—_—>

5 7

s /

 

Anode-workpiece (e.g., nickel alloy) (+ V")

(a)

 

 

 

 

 

i Cathode-feed
Cathode-tool

 

Anode-workpiece

 

Steady state gap 0.4 mm

 

 

 

(b)

Figure 4.2 - Schematic sketch of cathode and anode conﬁgurations for ECM. (a) Initial
setup and (b) Final shape of the workpiece [151].

A potential difference, usually ﬁxed at 10 V is applied across the anode workpiece
and the cathode (which acts as the tool) immersed in a suitable electrolyte (e.g., aqueous

NaCl solution). Retaining the cathode shape during electrolysis depends on the choice of

58

the electrolyte. The inter-electrode gap width should be about 0.4 mm and the average
current density should be of the order of 50 to 150 Acm’z. The rate at which metal is
removed from the anode is approximately in inverse proportion to the gap width between
the electrodes. It is essential to pump out the electrolyte (the conductivity of which is
about 0.2 Q'lcm‘l) at the rate of 3 to 30 rns‘1 to diminish unwanted effects that arise
due to cathodic gas generation, electrical heating and products of machining. With the
progression of machining, the inter-gap width will gradually tend to a steady-state value,
as the cathode is fed at the rate of 0.02 mms‘1 towards the anode. Under these conditions,
a shape roughly complementary to that of the cathode will be produced on the anode. The

governing equation for mass removal, from Faraday’s law is given by:

__An

_——=V, 41
"2 2F. p ( )

where ‘m’ is the mass of electrochemically machined material by a current ‘I’(amps) .
passed for a time, ‘t’(s). ‘F’ is Faraday’s constant (= 96500 C), ‘2’ is their valence, ‘V’ is
the volume of the machined material, ‘A’ is the atomic weight, and pa is the density of the
anode metal. The quantity ﬁ is called the electrochemical equivalent of the anode-metal.

From the above equation the volumetric removal rate of anodic metal V is given by:

_.M
_2Fm'

 

(4.2)

Detailed discussion of ECM can be found in [151—153]. Some of the advantages and dis-
advantages of ECM are brieﬂy outlined below:

Advantages: 1). To a certain extent the rate of material removal does not depend on the
hardness of the metal, 2). Complicated shapes can be produced, 3). No tool wear is present,
and 4). Smooth surfaces (about lam Ra) can be produced.

Disadvantages: 1). High equipment cost (may cost around $400,000 to $700,000 ex-

59

cluding the cost of peripheral equipment and facilities), 2). Mostly applicable to small-lot
productions, 3). Can be used only with electrically conductive materials, 4). Has to be
monitored continuously for short-circuit due to build up of sludge and machined prod-
ucts, 5). Low material removal rates, 6). Non-uniform surface ﬁnish, due to presence of
intermetallic compounds, non-uniform grain size, insoluble inclusions, and non-uniform
composition, 7). Cleaning of the workpiece is nearly always necessary after electro- chem-
ical machining, 8). Problem of hydrogen embrittlement with some electrolytes, 9). Local
chemical attack of the workpiece, 10). Periodic voltage polarity reversals to keep the cath-
ode surfaces clean, 11). Large set-up time and tool design cost, 12). Generation of vapors,
mist, and sludge, which pose environmental disposal problems, 13). Not highly accurate
process (sharp comers cannot be produced), 14). Poor reproducibility and ﬂexibility. Dif-
ﬁcult to machine internal radii smaller than 0.8 mm, and 15). Non-uniform surface ﬁnish

and accuracies for the frontal-cutting gap and side-gap areas.

4.4.3 Electron Beam Machining (EBM)

Focussed electron beam generated by an electron gun housed in a vacuum chamber, is
used as the source to remove material from the workpiece. A similar machining operation
involving ions as the source is used for machining and it is often referred as Ion Beam
Machining (IBM). Except for the mechanism of material removal, both the processes are
based on the same underlying principle.

EBM is primarily used for precision drilling of small holes (0.1 to 1.0 mm) in metals.
Moreover, industrial applications of EBM for non-metals and curvilinear cutting are very
limited. Problems like heat affected zone, rough surface of the cut, recast layers, etc., which
are persistent problems in other thermal machining methods, are also present in EBM. For
precision hole drilling, a backing material is required to eject the molten metal and vapors

from the site of the cut. Because of low volumetric removal rates and necessity of vacuum

60

severely limit the application of the process, this method would not be discussed further in
this report.

4.4.4 Electrical Discharge Machining (EDM)

This is a thermal machining process, applicable to electrically conductive workpieces.
Metal removal is achieved by generation of sparks between the tool-electrode (cathode,
-ve) and workpiece electrode (anode, +ve). The basic principle involved in EDM is same
as ECM, except that machining is done by erosion caused by the sparks. EDM is exten-
sively used to manufacture molds and dies (for example, plastic injection molds, extrusion
dies, forging dies, die casting dies, cutting of hardened steel etc.).

Two metal electrodes, one being the mirror image of the part to be machined (tool
electrode) and the other being the workpiece, are immersed in a dielectric liquid such as
light oil, paraﬂin wax, or de-ionized water. The electrode and the workpiece are separated
by a small gap, typically 0.01-0.4 mm. Direct current is used to generate a series of voltage
pulses of rectangular form, of magnitude 80 to 120 V and of frequency of the order of 5
kHz.

At these applied voltages, electrical breakdown of the dielectric will occur, due to the
close proximity of the electrodes. The process of breakdown or ionization of the dielectric
is a localized event, i.e., each spark occurs where the resistance is smallest, usually near
the last spark. The potential across the electrodes drops, when current passes through the
ionized dielectric. This process generates sparks, usually when the gap width is of the
order of 0.01 to 0.4 mm. When the breakdown occurs, electrons emitted by the cathode
and other stray electrons present in the gap accelerate towards the anode. On their way
to the anode, these electrons collide with the neutral atoms of the dielectric, generating
positive ions and electrons. The positive ions and electrons move towards the cathode and

anode respectively, giving up their kinetic energy in the form of heat, as soon as they strike

61

the electrodes. Figure 4.3 illustrates the mechanism involved in the EDM operation. When
the kinetic energy is converted into heat, temperatures of about 8000 to 12000 0C and

heat ﬂuxes of about 1017 Wrn‘2 can be obtained, even with sparks of very short duration

 

 

 

 

     
   

I I I I I . . .
.+++++++++
I I I I

+" Iii
0+++++++++
IIII

+" ili
I+I+++++++

   

++++++++++++++++++
I I I I I I I I I I I I I I I I I I
++++++++++++++++++

 

 

I I I I I I I I I I I I I I I I I I
++++++++++++++++++
I I I I I I I I I I I I I I I I I I

 

A
3

(a) (C)

 

 

 

 

Figure 4.3 - Production of sparks in EDM process. (a) Avalanche of electrons, (b) posi-
tively ionized gas in gap, (c) secondary avalanches, and (d) streamer develop-
ment.

(typically about 0.1 to 2000 as). Because of this large heat input, local temperatures of the
electrodes will exceed their normal boiling points. Existence of high pressures (of about
200 atm.), created in the plasma channel, due to the evaporation of dielectric liquid, will
prevent the evaporation of the super heated metal. But these high pressures will drop down
when the voltage is removed at the end of the “on” cycle. This sudden drop in pressure
will lead to explosive evaporation of the super heated metal, thus removing the metal in a
controlled way from the electrodes. Extensive literature on EDM and other recent advances
in EDM can be found in numerous journals and books [6, 8, 152, 154]. Modiﬁed EDM
processes like traveling-wire EDM are becoming popular for speciﬁc applications. In wire
EDM, the electrode is replaced by a continuously spooling conducting wire, which moves
with respect to the work by a CN C table.

Advantages: 1). Uniform smooth surface. Roughness can range from 0.2 to 12.5 pm Ra,

62

2). Materials like cemented WC and hardened tool steel can be readily machined, and 3).
Complex and ﬁne features are possible to machine.

Disadvantages: 1). Presence of heat affected zone and recast layer, 2). Low material
removal rates, 3). Inferior fatigue properties, due to surface cracks in the recast layer, 4).
Frequent replacement of electrodes, which are most often the most expensive part of an
EDM operation, 5). Cannot be used to machine non-conductive materials like ceramics,
glass, plastics etc., and 6). If the temperature in the dielectric exceeds the ﬂash point
temperature, then there is a danger of explosion. Also smoke can cause skin irritation and

other health problems.

CHAPTER 5

JOINING OF MMCs

Welding has progressed far less for aluminum based MMCs and other advanced materials.
Fusion welding, diffusion bonding, brazing, adhesive bonding, and spot welding, which are
commonly used for joining of aluminum alloys, can be used for joining aluminum based
MMCs, provided proper care is taken to preserve the unique properties of the composite.
Precautions, such as prior cleaning of the joint area and proper shielding of the weldment
(to avoid hydrogen contamination and subsequent formation of porosity), that are gener-
ally observed for monolithic aluminum joining are also necessary for joining of aluminum
based MMCs [155]. Fusion techniques like gas tungsten arc welding (GTAW) and gas
metal arc welding (GMAW) are economical and ﬂexible, and hence remain as basic meth-
ods of welding these advanced materials and other conventional alloys [156]. In addition,
for GMAW and GTAW, no special tooling is required (as is needed for diffusion, brazing,
and other such processes), there is no limitation on the thickness of the material to be joined
(as opposed to resistance welding or ultrasonic welding), and ﬁnally these processes lend
themselves very well to ﬁeld repair.

The major problem associated with GMAW and GTAW processes is the large heat-
affected-zone (HAZ) that cause the dissolution of hard phases and thereby reduce the
strength. Despite improvements in these conventional processes, problems are encountered

in the welding of advanced materials like aluminum metal matrix composites, rapidly solid-

63

64

iﬁed aluminum powder—metallurgy alloys and Al-Li alloys. Conventional fusion welding

of PM and intermetallic alloys and MMC present some serious problems:

P/M alloys Alloy unity in the heat—affected zone (HAZ) would be destroyed and
powder-formed alloys trap hydrogen, which forms gas porosity in the
weld area [157].

MMC Degradation of the reinforcement, formation of harmful interfacial reac-
tion products leading to inferior corrosion and mechanical properties, di-
lution of the reinforcement due to the use of ﬁller material, formation of

porosity, and large HAZ [71,99, 158-162].

Fusion welding techniques are quite adaptable to a range of discontinuously reinforced
aluminum composites. A few problems are associated with welding of these materials. For
the SiC-reinforced Al-MMCs, the high temperatures and relatively long times at tempera-
tures permit dissolution of the carbides and subsequent precipitation of detrimental phases
like Ath [156](refer Section 5.4). For the AlgOg-reinforced Al-MMCs, similar con-
ditions allow the individual reinforcing particles to agglomerate, which reduces material

performance.

5.1 Joining Processes

Published works concerning the various joining techniques used for MMC are summarized

as listed below [155, 156, 163-167]:

0 Adhesive bonding [49, 168,169]

0 Solid state processes (Solid phase diffusion bonding (SPDB) [167,170-173], Tran-
sient liquid—phase bonding (TLPB) [174-178], Friction welding (FRW) and Inertia
friction welding (IFRW) [179—183]

' o Arcwelding [184-189]

65

. Capacitor discharge welding [186, 190]
0 Electron beam welding [191]
0 Resistance welding [192-196]

0 Laser beam welding [49, 166, 187, 191, 197-210]

In Table 5.1 the various joining processes utilized for MMC joining are assessed in terms
of simplicity of joining technique, soundness of the weld, geometry of the MMC to be
joined, and mechanical & physical properties of the joint. The ratings given for laser beam

welding and electron beam welding are based on a very limited published research.

Table 5.1 - Ratings for processes used for MMCs joining [167].

 

 

 

 

Form of MM C

Joining Process Sheet Extrusion Casting
GTAW good good good
GMAW good good good
Resistance fair n/a n/a
Laser poor poor poor
Electron beam poor poor poor
Capacitor discharge good n/a n/a
Friction welding n/a good good
Diffusion bonding fair fair fair
Brazing fair fair fair
Adhesives good good good

 

 

 

 

 

5.2 Adhesive Bonding

Adhesive bonding has the advantage of not needing additional thermal cycle for bonding,
provided that the adhesive cures close to room temperature. A range of pretreatments have

also been applied, but the mechanical strength of the bonds are not as good as for unrein-

66

forced aluminum alloys [156, 166, 168, 169]. It has been found that the application of con-
ventional surface pretreatments to aluminum alloy reinforced with SiC,, produces surfaces
that are quite different ﬁ'om those produced on bulk matrix materials [168]. Grist blast-
ing smears the surfaces with matrix metal, leaving them aluminum-rich. Standard etching
treatments aggressively remove aluminum, produces surfaces enriched with loosely bound
SiC,, and change overall surface roughness. Proper surface treatment of an adherend sur-
face is one of the most decisive factor for achieving high quality joints. Various surface
treatments are practiced but most of them have their own limitations, i. e. solvent absorp-
tion by the matrix (in the case of polymer composites), mechanical damage to the ﬁber
or degradation of the matrix [49]. A recent study has reported that surface treatment by
KrF excimer laser is an effective means of improving adhesive bonding of polymer com-
posites [49]. Some of the problems associated with adhesive bonding are joint strength,
durability of the bond, the large number of fabrication and handling steps necessary for
production, and quality assurance [211]. The durability of an adhesive bond can be greatly
enhanced by proper preparation of the adherend surface [211]. Doesburg et al. have re-
ported that adhesive bonding of A6061/A1203/10p and A6061/A1203/20p composites re-
sulted in bonds that performed better than the base 6061 aluminum alloy in lap—shear tests.
Specimen surface preparation was carried out by etching in sulphuric acid/sodium dichro-
mate solution and then anodizing in either chromic or phosphoric acid bath. The improved
results in MMC’s may be due to the etching process, which preferentially attacks the inter-
face between the particles and the matrix, increasing the surface roughness to the beneﬁt
of the adhesive bonding process [169]. It was reported that laser cut specimens may be
better suited for adhesive bonding of MMCs, because of the inherent roughness of the cut

surface [211], which reduces the number of steps involved in surface pretreatment.

- 67
5.3 Solid State Processes

Solid state processes such as diffusion bonding and friction welding have shown great
potential for the joining of MC materials. Some of the advantages of solid state processes

are:

0 since temperatures lower than that in fusion welding are involved in these processes,
occluded gases do not appear to be a major problem. As a consequence of low
temperatures;

0 reactions that form deleterious products are eliminated

e segregation and viscosity effects in molten weld pools are not present

0 the reinforcement distribution and position can be maintained in a controlled manner

5.3.1 Diffusion Bonding

Though diffusion bonding is one of the most widely used process in the primary fabrication
of composite materials, it has been used to a very limited extent for secondary fabrication
of composites. It ispa joining process wherein coalescence between two clean and closely
ﬁtting parts is obtained under the application of heat and pressure [167,174]. There are two
variants in the diffusion bonding process i. e., deformation bonding and diffusion bonding,
with latter itself being divided into solid state and liquid phase diffusion bonding processes.
Deformation bonding relies on extensive plastic ﬂow to disrupt oxide ﬁlms and achieve
metallic bonding. In the solid state diffusion bonding process, all the reactions take place in
the solid state, while in the liquid phase diffusion bonding process, a liquid phase generated
either through a low melting eutectic interlayer or due to the eutectic reaction between two
components of a eutectic system helps the bonding process [212].

TLP bonding combines the advantages of solid state diffusion bonding with manufac-

turing ease of conventional furnace brazing. In TLP bonding, a eutectic system (comprised

68

of the base metal to be joined and a diffusion aid) is held isothermally under slight com-
pression at a temperature slightly higher than the eutectic temperature. The diffusion aid
can be a pure metal or an alloy system capable of forming a eutectic system with the
base material at the bonding temperature. The diffusion aid can be an interlayer foil or
an electro—deposited/plasma depositedfron deposited/vapor deposited coating [174]. Equi-
librium thermodynamics dictates that under such conditions a eutectic liquid is formed.
Under appropriate conditions the liquid spreads and wets the joint. With continuous hold-
ing at appropriate temperatures, the liquid solidiﬁes due to diffusional loss of solute to the
solid with which it is in contact. With continued depletion of solute and the consequent
solidiﬁcation of the solute-lean liquid, the solid-liquid interface moves into the liquid until
the liquid isothermally solidiﬁes. The last liquid to solidify will have attained a composi-
tion corresponding to the equilibrium maximum solid solubility at the temperature under
consideration. At this stage solidiﬁcation will be complete, and further continued holding
at the temperature results in homogenization of chemistry [177, 178]. For TLPB, there is
concern about damage to the properties of the composite base material during processing,
because the thermal cycles involved in TLPB hold the entire assembly at temperatures near

the solidus for a longer time than in the other fusion welding processes [156].

5.3.1.1 Solid Phase Diffusion Bonding

Trials on Al/SiC/20p and AA6061/SiC/20p produced successful bonds using Cu, Ag or no
interlayers [170]. Metallographic examination conﬁrmed the need for precise control of
the process variables, to ensure that particulate-rich zones or particulate-depleted zones,
were not formed at the interface. -

Metallic interlayers have also been used to bond AA8090 (Al-Li) MMC materials
[171]. VVrth no interlayer, there was high level of certainty that there will not only be
particle—matrix (P-M) bonds and matrix—matrix (M—M) bonds but also particle-particle

(P-P) bonds. Because of the inert nature of the particulate, the P—P bonds will be weak,

69

reducing the overall joint strength. It was reported that addition of an interlayer eliminated
the P—P bonds, and increased the proportion of P-M and M—M bonds [171]. The effect of
surface treatment, bonding parameters and insert (or interlayer) metal on the diffusion bond
strength has been investigated for AA6061/A1203/20w reinforced MMC [172, 173]. The
use of an interlayer increased the bond strength. The surface treatment was varied from
an electro—polished ﬁnish, to turning in a lathe and wire brushing, with the ﬁrst treatment

producing the highest bond strength and the later the lowest.

5.3.1.2 Transient Liquid Phase Bonding

Joints of good quality have been produced in AA6061/A1203/ 15p using TLPB [175]. Foils
of Ag, Cu, and BAlSi-4 were used as interlayers. The best consistency in terms of joint
shear strength (> 175 N /mm2) was achieved using the Ag at 575 °C for 100 min and
BAlSi-4 at 585 °C for 20 min. However, the Cu interlayer produced joints of higher shear
strength, but results appeared more scattered. “Without the use of metallic interlayers, high
levels of deformation are required to achieve a reasonable level of bonding [175]. More-
over, in TLPB, the lowest possible temperatures has to be used, to limit the damage to
the properties of the composite. Application of external pressures would aid in improVing
interlayer-substrate contact. But excess pressures can lead to expulsion of the liquid zone,
which will ultimately result in reinforcement—rich areas at the bond line [177].

TLP bonding with gold and Al—Si—Mg interlayers has been used to join
AA6061/SiC/25p composite sheets [174]. Based on thermomechanical simulation stud-
ies, a process window of time-temperature combinations was developed that would not
produce damage to the properties of the base material. The window ranged from 30 min
at 565 °C to 10 min at 580 °C. A 0.025 mm gold interlayer was shown to produce better
results than the Al-Si-Mg material. Three distinct zones were observed in the vicinity of
the bond interface: a particle-enriched zone, a diffuse ﬂow zone that showed a pattern of

material ﬂow, and the undisturbed base material. No mention was made of the formation of

70

A14C3 for these conditions. Low tensile joint efﬁciencies (< 30%) were attributed to voids
at the joint interface.

Sabathier et al. has used liquid gallium as the interlayer for bonding of
AA6061/SiC/20p composite sheets. Gallium was selected as interlayer material to take
advantage of its low melting point eutectic (it forms with Al) and the high degree of sol-
ubility of gallium in Al. This study concentrated more on the transport mechanism and
kinetics of the diffusion process. The joining mechanism occurs in three successive stages:
rapid inﬁltration of the liquid interlayer and homogenization of the solid matrix, liquid
phase depletion from the grain boundaries and interfaces (crucial in reducing the potential
for liquid metal embrittlement), and ﬁnally homogenization in the solid state. In a study on
TLP bonding of AA6061/A1203/20p composite, it was noted that particulate segregation
at the bond-line was dependent on the liquid ﬁlm thickness, bonding temperature, parti-
cle size, and particle spacing distributions in the MMC [177]. COpper foil was used as an
interlayer. The thickness of the copper foil was very critical (as it determines the liquid
width at the bonding temperature) and the thickness was related to the heating rate: 0.6 pm
for 1 K13 and 2.4 pm. for 0.01 K/s heating rates. The TLP bonding process was investi-
gated for continuous SiC ﬁber reinforced Tr-6Al-4V composites [17 8]. For this composite
Ti-Cu-Zr amorphous thin foil was used as interlayer. Extensive brittle reaction products,
such as (Ti,Zr)5 $13, TiC, and ZrC were formed at the interface between the ﬁbers and the
ﬁller metal. It was concluded that the dissolution of the base material and the isothermal
solidiﬁcation processes were hardly affected by the ﬁber volume fraction up to 45%, and
no marked differences due to the ﬁbers were observed.

5.3.2 Friction Welding

In friction welding, several variants (rotary, inertia, and linear) have been used [179—183].

In friction welding, the material is softened by the continuous relative rubbing motion be-

71

tween the contact surfaces, which results in a temperature rise. No melting is involved
in this process. The softened material in the interface will eventually ﬂow plastically,
with the formation of an up-set collar [166]. After a certain amount of up-setting has
occurred, the rotation is stopped and external compressive forces are increased to consol-
idate the weld. It has been noted that the high degree of movement of component parts
used to create the frictional heat causes break up of the reinforcement particles around
the interface, while the HAZ produced during friction welding can have a reduced hard-
ness [180, 183]. An integrated process model has been developed for the temperature
distribution and microstructural development during continuous drive friction welding of
A357/SiC/13p MMC [180, 181]. From this model, very good predictions of HAZ hardness
have been produced. ‘In most cases, postweld heat treatment was needed to regain full joint
strength [179]. In the case of chopped ﬁber reinforced MMCs, ﬁber break up and alignment
of ﬁbers occurs and this has been observed in rotary friction welding [170]. Extensive mi-
crocracking of the transition layers located at the joint interface of AA6061/A1203/10p and
A181 304 austenitic stainless steel welds was observed [182]. The average particle diame-
ter and inter-particle spacing decreased markedly in the region immediately adjacent to the
bond-line of dissimilar MMC/AISI 304 joints. Lienert et al. [183] studied inertia friction
welding of 8009/SiC/ 1 1p PM composite. No evidence of chemical reaction between the
SiC particles and the matrix was observed, but the reinforcing particles were reduced in

size, as conﬁrmed by a large number of ﬁne particles in the weld line.

5.4 Fusion Processes

Fusion processes include GTAW, GMAW, capacitor discharge, resistance, plasma, laser,
and electron beam welding techniques. Numerous problems are encountered in the fusion
welding of aluminum based MMC andthese include:

Melt viscosity: The presence of solid phase in the weld pool, because the reinforcement

72

does not necessarily melt at welding temperatures, result in a weld pool that is partially
molten [156, 166, 167, 188]. In other words, the very difference in the melting points of the
matrix and the reinforcement is responsible for poor ﬂuidity in the weld pool. The obvious
effect of this is in the increase of viscosity of the weld pool. The increased viscosity of the
weld pool as compared to the monolithic aluminum alloy, prevents proper manipulation of
the weld pool in terms of ﬂow and wettability. The viscous nature of the MMC weld pool
also makes adequate mixing of ﬁllers virtually impossible, because of the limited mass
transfer rates. The viscosity of the weld pool depends on temperature, volume fraction, and
type of reinforcement. It is commonly believed that conductive heat transfer plays a sig-
niﬁcant role in determining the temperature distributions and cooling rates in and around
the weld pool of Al-MMCs, as opposed to heat ﬂow by convection in monolithic aluminum
alloys [188]. This difference in type of heat transfer can affect the resulting microstructures
and stress distributions in the MMC weldment [159, 186, 190, 191, 197-199, 204—206].
Particle/matrix interaction: Fusion joining of SiC reinforced aluminum composite is '
plagued by loss of SiC and subsequent formation of large amounts of A14C3, because
of the change in mode of heat transfer (Figure 5.1). Large amounts of brittle A14C3
(in a very coarse and clustered form), which readily dissolves in aqueous environments
to form acetylene (thereby generating porosity), result in degraded mechanical proper-
ties [159,185-187, 203, 205]. The use of reinforcements, such as A1203, which is stable in
molten aluminum pool, will solve this problem. A14C3 formation is also known to form in
graphite reinforced Al-MMCs and it is generally believed to form in one of the following
ways:

3500(3) + 4Al(l) —> Al4C3(S) + 3Sz'(Al) (5.1)
30(3) + 4Al(l) —-> Al4C3(S) (5.2)

where (s) and (I) represent solid and liquid phases, respectively. The A14C3 phase pre-

cipitates out as platelets and needles, while the silicon tends to form “blocks” [155, 159,

73

191, 200, 210]. The needle shaped morphology of A14C3 can cause stress concentrations

‘ ‘ .- Pl} ‘ 42;“
J—:A\ F..."

1';
'. a": ”V"
. ,.
r '_ \ o
I. age? . x a l
'\ -" ‘ \n a. .
.r l- "e ‘5’

. .i .
, k, g. AA . , I 7,
“my .4 .' :~ ’~

I .' ’ x a .
‘33:“; :. 2.. 3’

. Ls. .

_, “Ts-"3i. .3qu .
' ELRFf: 5)" v—I. rcp'w"
~ \y

. l

~.5\'. "'1 g {

 

Figure 5.1 -Microstructure of Al4C3 reaction product in an electron beam melt run in
2024/SiC/20p composite [213].

during load transfer between matrix and reinforcement (see Figure 5.1). From a study of
thermodynamics of equation 5.1 in the temperature range of 700° C to 1200" C indicates
that standard free energy AGO of the above reaction becomes increasingly negative with
increasing temperature (i.e. the reaction proceeds further to the right) and that additions
of Si to the Al alloy matrix, which increases the activity of the Si, aids in limiting the for-
mation of A14C3 during joining [159]. Hence, use of Si-rich ﬁller materials, e.g. ER4043
(Al—4.5/6.0% Si) or ER4047 (Al-10/ 13% Si), should retard the reaction indicated in equa-
tion 5.1. Some success has also been reported with rapid thermal cycle process conditions
(pulse lasers) and with the use of Ti as an interfacial reactive ﬁller material [200, 208, 210].

In aluminum-magnesium composites (AA 6061 alloy) reinforced with A1203, magne-

sium spinel (MgA1204) is known to form by the following reaction:

3Mg + 4.41203 —> 3MgA1204 + 2A1 (5.3)

74

A similar reaction (copper spinel — CuAlgO4) is also known to occur in aluminum-copper
alloy composites reinforced with alumina [155]. The formation of magnesium spinel and
copper spinel is known to enhance the wettability of the A1203 reinforcement by the molten
aluminum, which limits the agglomeration or segregation of A1203 into large clusters. The
presence of spinels at the interface does not appear to markedly alter the mechanical prop-
erties of the composite [90] and the weldment [155].

Segregation: Segregation of the reinforcement can take place by rejection of the particu-
lates by the solidifying interface and this can occur both on a macro and micro—scale as it
is commonly observed in gas shielded welds [167, 188]. In general, this can be prevented
by the use of fusion welding processes such as laser and electron beam welding, where
rapid cooling and solidiﬁcation rates are achieved [191, 200]. In resistance spot welding,
massive segregation of the particulates result in inhomogeneous distribution of the rein—
forcement across the weld nugget [155, 156, 167] and it is not clear whether this phenom-
ena is caused by a hydrodynamic effect (which generates centrifugal force) or by surface
tension effect. In most fusion joining processes, the ceramic reinforcements segregate in
the interdendritic regions. The signiﬁcance of this is clearly inﬂuenced by particle size and
the cooling rate, which determines the dendritic cell size. In processes where large weld
pools are generated, gravitational effects may also lead to enhanced macro-segregation of
the reinforcement [156, 167].

Occluded Gas: MMCs produced by powder metallurgical routes are difﬁcult to weld, be-
cause of generation of pores in the weld pool during fusion joining. In addition, if the
occluded gas content is too high, gases will be liberated ﬂ'om the molten pool at high tem-

peratures, causing extensive cracking in the HAZ.

75

5.4.1 GTA and GMA Welding

GTA and GMA welding processes were used to join boron-aluminum [184], tita-
nium/graphite and titanium/tungsten [214], 6061/SiC/20w [185], 6061/SiC/40p [186],
2024/SiC/ 15p [187] composite materials. Although composite materials like Ti/W, Ti/Gr,
and Al/B, are not considered to be promising for structural applications, and the volume
fraction of the reinforcement was not sumcient for a high strength composite, these mate-
rials did serve well as models for the investigation of the fusion weldability of composites.
From the above studies, it was clear that these arc welding techniques pose numerous prob-

lems in joining composite materials, as listed below:

0 porosity; vacuum degassing prior to welding was required to reduce pore density
[185]

e delamination and splitting of ﬁbers [184, 214]

0 formation of undesirable chemical reaction products

0 degradation of mechanical properties and cracking in the HAZ.

The GTAW of 6061 Al—B MMC without ﬁller material, caused overheating and severe
B-ﬁlament fragmentation (cracking) and dissolution [184]. However, with ER4043 ﬁller,
this. problem was eliminated and the Si-rich ﬁller protected the outermost layers of the
B ﬁlaments, with metal ﬂow around the B ﬁlaments [184]. The welds deposited by either
GTAW or GMAW on 6061/SiC/20w with ER4043 ﬁller were characterized by gross poros-
ity and delaminations in both weldment and HAZ [185]. It was reported that the source of
the porosity was the liberation of hydrogen emanating from the aluminum powder used in
the fabrication of MMC. Vacuum degassing for long periods of time prior to welding did
result in reduced pore density in the weld [185]. However, A14C3 and Al-Si eutectic were
detected in the degassed composites. Autogenous GTAW of 6061/SiC/40p welds were
characterized by a total lack of ﬂuidity and gross porosity in the weldment [186]. The lack
of ﬂuidity is due to the large volume fraction of solid SiC particulates in a pool of molten

76

aluminum. Plate-like precipitates of A14C3 were also produced in the weldment as shown
in equation 5.1. It was concluded that joining of SiC/Al composite be performed under
conditions of minimal superheating and short contact time between the SiC particles and
molten Al [71,167,186].

Results similar to the above were also reported by Lundin et al. [187] for GTAW of
2024/SiC! 15p composite. For this 2024 aluminum alloy composite, maximum hardness
values were obtained for HAZ (85 DPH in the parent metal, 140 DPH in the fusion zone,
and 160 DPH in the HAZ), eventhough coarsened grain structure in the HAZ and predic-
tions for the appearance of the HAZ of unreinforced 2024 indicate otherwise. The HAZ
in precipitation hardenable 2XXX alloys is not dominated by recrystallization and grain
growth, but rather by dissolution of strengthening precipitates and thus a decrease in HAZ
hardness should result. It was speculated that residual stresses produced in the HAZ dur-
ing the joining process may have accelerated the natural aging of the matrix, promoting
the formation of metastable strengthening precipitates [187]. Diﬁ'erential thermal strains
due to different thermal expansion ratios of the SiC particulates and the 2024 alloy may
have also contributed to the increase in hardness in HAZ. A comparative study of LBW
and GTAW of 2024/SiC/ 15p by Lundin et al. [187], indicates that in the case of GTAW,
the penetration was minimal, the fusion zone was ‘ shallow and semicircular, and the size
(longer and thicker) and the density of the acicular A14C3 precipitates was larger than those
present in the LBW specimens.

Altshuller let al. [188], demonstrated the feasibility of GMA welding of
6061/A1203/20p composite. The volume percentage of the reinforcement was as low as
4.1 - 5.5% in the weld zone produced with Mg-rich ER5356 (Al-4.5/5.5% Mg-Mn-Cr-Tr)
as ﬁller alloy. As a result, these joints would exhibit lower stiffness, poor fatigue strength,
and inferior corrosion behavior than the virgin composite material. It was also reported that
post-weld heat treatment to T6 condition was necessary to achieve maximum mechanical

properties. No explanation was given for the negligible micro-ﬁssuring observed in the

77

HAZ, despite the tendency of heat-treatable aluminum alloys like AA2XXX, AA6XXX,
and AA7XXX to exhibit micro-ﬁssuring during fusion welding.

GMAW process of 6061/A1203 composite base plate with conventional ER4043 and
continuous composite electrode ﬁller of 4043 Al/A1203/10p resulted in welds with exten-
sive weld metal porosity, sizeable loss of A1203 particles in the weldment, and cracking of
the weldment [156]. The composite welding electrodes (1.14 mm in diameter) were fab—
ricated by extrusion/rolling/drawing (refer to Kivineva et al. [189] for extensive research
on composite electrodes for GMAW and GTAW). These unsatisfactory results are largely
due to poor wettability of the base plate and the dispersed A1203, by 4043 Al ﬁller metal.
Welds produced with magnesium rich ﬁller electrodes like ER5356 provided signiﬁcant
improvements in both weld metal soundness and A12ng retention. Other alloy composi-
tions that exhibit excellent wetting characteristics toward A1203 such as copper and silicon

containing A390 alloy might be a suitable weld material [156].

5.4.2 Capacitor Discharge Welding

Capacitor discharge welding process utilizes electrical energy (arc) discharged from a bank
of capacitors to rapidly heat the weld zone via 12R heating. The are is initiated between two
parts, one of which is stationary and the other moving under a force towards the former.
This are heats both faying surfaces to the melting temperature [186]. As the parts meet,
solidiﬁcation occurs and excess molten material is expelled. Compared to conventional
resistance welding techniques, capacitor discharge resistance welding process is character-
ized by short thermal cycles (typically less than 10 to 20 ms in duration) [186, 190]. These
short duration time cycles promote concentration of the heat in a small zone near the weld
interface, rather than allowing the heat to be conducted away from the weld line. As a con-
sequence, capacitor discharge welding technique generates rapid cooling rates in the weld

pool, which further minimizes HAZ [190].

78

Devletian [186] has reported that autogenous capacitor discharge welding of
6061/SiC/40p produced weld fusion zones that were structurally sound, with no porosity
and no interfacial reaction products (A14C3). Absence of any chemical reaction between
molten Al and SiC particles was attributed to low superheating temperatures experienced
during the welding cycles. For high volume SiC particulate reinforced composites, ﬁller
metal in the form of thin foil (20 to 40 pm thick) can be placed between the two faying
surfaces, to wet the entire surface area of solid SiC particles. The short exposure times
and rapid cooling rates (approximately 106 00/3) were responsible for the retention of
the SiC particles and inhibition of Ath formation in the weld fusion zone and HAZ.
This technique in its present form is not suitable for producing continuous welds and is
in some respects similar to resistance spot welding. By controlling the energy levels and
the electrode force, controlled fusion welds and solid state welds (high temperatures can
be induced at the faying surfaces without melting) can be generated by capacitor discharge
welding technique [190].

5.4.3 Resistance Welding

Earlier research on resistance welding was restricted to boron-aluminum composites
[192-195]. Very poor joint efﬁciencies ranging from 39.2 to 45.8% were obtained for
B-Al composites. Metzger [158] observed that the average joint efﬁciency in Al-B resis-
tance spot welded composites ranged from 14 to 38%. Spot welds on A359/SiC/ 10p and
6061/A1203/10p composites, exhibited some redistribution of the reinforcement in the fu-
sion zone [196]. Typically, the fusion zone was relatively free of the reinforcement near
the fusion line, with an increased density of the reinforcement near the center of the weld.
The redistribution of the reinforcement was due to macroscopic liquid ﬂow during solid—
iﬁcation. In SiC,/Al composite, weld nugget microcracking was also observed [71]. The

weld nugget microstructure exhibited some presence of A14C3 phase in areas where the

79

reinforcement had been dissolved. These areas were at the center of the weld nugget and
were isolated from the weld/HAZ interface [71,196]. The effect of particulate depletion on

mechanical performance of the weld is not known.

5.4.4 Electron Beam Welding

It is clear from published literature that welding techniques such as EBW, which employ
a rapid thermal cycle and low heat input, should restrict the formation of A14C3 [197-
200, 204, 205]. However, the physics of beam-material interaction for EBW and LBW is
different. Heating during LBW results from the absorption of photons by the substrate,
whereas heating during EBW occurs by the transfer of kinetic energy to the atoms of the
substrate via collisions with the high—energy electrons of the beam [156]. The physics of
an electron beam are such that there is no preferential absorption of the energy by the SiC
particles (in A356/SiC/ 15p composite) which could produce excessive heat and hence an .
Al/SiC reaction [191]. Certainly, the electron beam has been found to be less deleterious
(lower levels of Al4C3) than when using laser beam welding at the same power levels and
speeds. A high travel speed and sharp focus (i.e. an estimated beam diameter of 0.4 mm)
of the electron beam was recommended to prevent the formation of Al4C3 [191]. The
different mechanisms of energy transfer in EBW and LBW appear to affect the ﬁnal weld
microstructures of SiC-reinforced Al-MMCs and, thus, the weld properties. This occurs
despite the fact that both welding processes allow for rapid thermal cycles with low overall
heat input [156,191].

5.4.5 Laser Beam Welding

The ideal alternative for welding MMC systems would be solid-state welding processes
and advanced beam technologies that allow fusion welding with a very narrow HAZ.

The other non-conventional welding techniques have their own disadvantages. For ex-

80

ample, inertia ﬁiction welding used to weld advanced materials restrict the shapes that
can be welded. SOme of the difﬁculties with conventional welding of composites like, arc

(GMAW, GTAW), resistance, or capacitor discharge welding etc. are [158]:

e Distortion of the weld part, due to excessive input of heat.

0 Disruption of the reinforcement in the matrix and considerable interaction between
the reinforcement and the matrix.

0 Porosity, delamination, or ﬁber pull-out.

e Undesirable chemical reactions leading to products which could degrade the materi-
als corrosion resistance.

0 Involve use of a non-composite ﬁller/interlayer metal/alloy which results in weld
beads that have a low particulate volume fraction and therefore exhibit lower stiffness

than the base composite material.
0 Large heat-affected zone.

0 Problems with composites which have refractory or highly reactive matrices.

Two mechanisms are involved in laser welding depending on the laser parameters. A
deep penetration weld is characterized by an hour-glass shape with a high depth-to-width
ratio and is accompanied by excessive melting, loss of material through vaporization, low
cooling rate, shrinkage cracks, and gas porosity. These effects are the consequences of
the formation of a “key-hole” and energy transfer via the key-hole. Figure 5.2 depicts the
penetration weld geometry, where multiple reﬂections of the laser beam in the long, nar-
row keyhole results in much higher absorption as well as spreading of the absorbed energy
along the whole length of the keyhole [215]. In contrast, a conduction weld exhibits a
small depth-to-width ratio and minimal defects. The geometry of conduction weld is hemi-
spherical and is similar to that of conventional arc welds. These two modes of laser beam
welding are depicted in Figure 5.3. These two modes differ in the amount of heat input

”into the sample and the rate of heat conduction from the interaction zone. In the case of

81

conduction mode, for a normal reﬂection the laser beam interacts at a single surface as op-
posed to multiple reﬂections of the beam from the walls of the keyhole in deep penetration
mode.

High
Intenstty 1U

Beam N
Power PTOT . Melted Zone

  
  
 

Molten
Pool

 

 

Penetration ..
Depth 0 l

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

lll

Figure 5.2 - Laser welding process diagram to illustrate penetration weld geometry [215].

 

1-11M.

 

 

 

The welding of aluminum and its alloys with CO; lasers is not yet state of the art
[201, 202, 217]. The reasons for this are: the high reﬂectivity, high thermal conductivity,
the low ionization energy, chemical reactions of the alloyed elements, poor ﬂuidity of the
molten material in the weld pool, plasma suppression etc. These same reasons are also
responsible for the amount of skill required in the welding of aluminum-based composite
systems. Consequently, conventional and advanced welding processes are presented with
these same problems in the joining of aluminum-based composite materials. Moreover, the
lower coefﬁcient of thermal expansion and the abrasive nature of the MMCs “reduce the

tendency for the joint to close up in front of the arc” or key-hole and make it undesirable

82

 

 

 

 

    
 

 

 

 

 

 

 

Deep penetration weld Thermal conduction limited weld
Laser Beam Laser Beam
dE/dt 4 ‘Key hole’ dE/dt4
guy U
dH/dt dI'I/dt
Material Material
dE/dt >> dH/dt dEdt < dH/dt

dE/dt = Rate of laser energy input
dH/dt = Rate of heat conduction

Figure 5.3 - Schematic to illustrate the two modes of laser beam welding: (a) Deep pene-
tration weld and (b) Thermal conduction limited weld [216].

“to back gouge for two-sided welds” [185, 188].

As with EBW, the use of fast thermal cycles (low heat input) for welding of MMCs
will be beneﬁcial. The majority of the work reported on laser welding has involved the
Al-SiC MMC system [187, 191, 197-200,203, 204, 208, 210]. The effects of heat input and
duration of thermal cycle have been studied using a continuous wave (CW) C02 laser and
a pulsed Nd:YAG laser on an A356/SiC/15p MMC [198]. The high Si content (around 6.5
to 7.5 wt%) of the A356 matrix material helps in preventing A14C3 formation by reducing
the extent of the reaction in Equation 5.1. The microstructures of the welds produced from
the two lasers were similar, but that from the Nd:YAG laser was somewhat ﬁner. The Al—
SiC reaction (Equation 5.1), which produced A14C3, primary Si, and Al-Si eutectic, was
apparent as a small region in the center of the fusion zone. It was observed that the laser

energy was not uniformly absorbed by the composite, and that the nonuniform absorption

83

of energy appeared to accelerate the formation of A14C3 along the weld centerline. Pref-
erential absorption of laser energy by SiC particles, caused it to rapidly heat and dissolve.
Hence, it was suggested that A14C3 did not form by reaction indicated in Equation 5.1, but
by precipitation from the liquid solution on cooling [198]. Based on a phase transformation
model, it was argued that SiC particles dissolve in liquid Al to form a homologous liquid at
predicted welding temperatures of 2500 - 3000 °C in the central fusion zone. On cooling
(approximately 1950 0C), the microstructure evolves as a succession of phases precipitat-
ing from the homologous liquid. Three phases form in the central fusion zone; the ﬁrst
phase is A14C3, followed by primary Si and ﬁnally an Al-Si eutectic structure [198].

Laser processing of Alcan 2014 Al alloy - 20 Vol% SiC“), produced by spray casting,
exhibited considerable amount of porosity in the HAZ [203]. There was complete loss of
the original SiC particles in the fusion zone, as the carbide particles melted and formed
A14C3. The excessive porosity in the HAZ was mainly from the solidiﬁcation shrinkage of
the molten matrix dispersed with mostly unmelted carbide particles. Lundin et al., [187] l
have investigated LBW and GTAW of 15 Vol% SiC(,,)/2024 plates of 1 cm thick, at a
traverse speed of 1 and 2 m/min., with and without argon shroud gas. For both types of
welding methods, the formation of A14C3 was observed in the fusion zone. It wastsuggested
that, low heat input (by controlling the traverse speed) would prohibit the formation of
Ath and also prevent inhomogeneous distribution of the SiC particles in the weld zone.
Welds made with a constant power level at various travel speeds were found to alter the
size and distribution of A14C3 plates in the fusion zone. The presence of these A14C3 plates
were also found to impart high hardness to the weldment.

Dahotre et al. [197,199] used C02 laser (pulse mode) at varying power levels to produce
bead—on-plate welds on A356/SiC! 10p composite. It was found that as the energy density
(J / cm2) increased the SiC particle dissolution became greater and the A14C3 plates in-
creased in both size and quantity (Figure 5.4). A drop in the volume of SiC associated with
an increase in the volume of Ath as shown in Figure 5.4(b). In a recent study Dahotre et

al. [208] reported that by using a Ti reactive ﬁller for C02 laser welding of A356/SiC/10p

84

 

 

   
  

 

 

 

 

 

 

 

3 1953 1c
2‘ E
i3 175... . v
g. 5 E N403
‘ ”‘0': i 3L-
J
3:2- 3. ~ 8
a «35:. 2
(a) 3 8 (b) E 4..
8
i 4115.3. “2‘
314 5 :1
3 ~ 6' 2
g -955 >
U U
i I 3 o
20’ 7 , 'l‘ '13 ”75 s 7 9 ll 13 15
encirlc eucaev.x10‘(.ue.1). . SPECIFICENERGYX 10 4(.l/crrlz)

Figure 5.4 - (a). The length and width of Ath plates and (b). volume fraction of SiC and
A14C3 as a function of speciﬁc energy [197].

and A356/SiC/ZOp composite, A14C3 formation (as indicated by Equation 5.1) was avoided
and instead TiC was formed. Formation of TiC was expected to enhance the joint proper-
ties, since TiC has excellent thermal stability, melts without decomposing, has high hard-
ness, and is more denser than SiC and Ath. Stable TiC in the fusion zone forms according
to the reaction: Ti + SiC —-> TiC + 52'. The Gibbs free energy change for TiC formation
is substantially lower than that of A14C3 (Equation 5.1).

Studies carried out on Duralcan W6A.20A (6061/A1203/20p), using 5.0 kW C02 laser
had indicated that plasma breakdown was major concern for this composite, because of the
presence of alumina particles [201,202]. It was also reported that porosity can be controlled
by using powers lower than 3.5 kW for samples of thickness 1.6 mm or lower. Extruded
tubes of wall thickness 2.3 mm were used in this study to carry out both bead-on-plate and
butt welds.

High intensity joining processes, such as laser beam welding, generate very high cool-

ing rates that are detrimental for materials like high hardenability steels. In Figure 5.5,

85

a split beam laser processing concept is depicted, which in theory should result in re-
duced cooling rates [218]. The cooling rate achieved depends, to a large extent, on the
preheat temperature. Split dual beam technique proposed here would eliminate the extra
step required for preheating many poor thermal shock resistant materials. Furthermore, the
traditional preheating methods might affect other portions of the workpiece, unlike laser
beam that preheats only the required width of the weld sample. The collinear dual laser
beam welding process proposed in this research is similar to the technique presented by
Kannatey-Asibu, Jr. [218], in some aspects. The joining techniques discussed above are

 

 

 

 

Figure 5.5 - Schematic of a split beam laser welding system [218].

summarized in Table 5.2. In Table 5.2, some of the advantages and disadvantages of the
various joining techniques are given. Again, LBW technique provides the greatest potential
for these materials both for like joints and for joining MMCs to other materials. But LBW

technique have their limitations too.

86

808308-80: E03 60m “E3.—

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

608308 >33er 8 PE?— eea €08.68 88.8088
09 .38 88.03.» 8 3.0.8883 :08 88838“. 88. 3 800.: 0c :30 9800.5
8883.. 8.52. 8.5
638038 88838.0 .8.“ 6000 3MB
880030.. 088388883 02 A535” 80:3 88
62680.. on 9 008.50.. :08"— 033>0an 8080 89.0.50 econ :3 3mm
$888808; Bo..— .e0888 w83 803—8008
3.88: .8 00580 .80er 2880me $883 8898:. 88 8 08am mas;
6:3 deem 0085.8 .38 £9
tonnes: $38 080m0oxm .5305 259 w8e=em
88880 new @0862 08 .38 803—8085 888838 8808 0.883 02 880.55
808095 38m 83m
3888 88882 he 8839.88 2820A 8.00% meme—03 :8: 3.888%
60832 8 3w $88.50 8309 89m 800%
982085 .522 0%-? 8 8800308 ”UV? 9:203 :8: 608.82 88.08, 02 3m:
.8==0«> 88.60% .622
0%-? 8 805030.. ”Us? .«e .0888?” I1 _ 3mm
802212 “002882 DB ecu ”on?
.000: 8 083 80:: 388808: 8 00—888 he w8303 030808 8:83 3<20
80:3 mﬁmeobm 308 203 33 .0022 .8883.— »Uvz 0260.. .w.» .80er 83
08-2 8 88830.. ”Us? we 8:888 meme—03 80 9 800080 on 88 083 0:5 Bea—.0
800009:— 5.05%
unmeaegeena a0m820>e< A 8.008%

.8: .52 do: .0022 88 800: 880008 w88e.- 888$ 05 .8 mow—3.56388 88 m0w85>e< .. ﬂm 030a.

 

CHAPTER 6

EXPERIMENTAL PROCEDURE

Samples of commercial aluminum alloy 6061 reinforced with 10% and 20% vol A1203
particles (W 6A. 10A and W6A.20A respectively) of thickness varying from 2.5 mm to 11.5
mm cut from an extruded rod of 38.0 mm, were used for the present study. The length and
width of the specimens were 50 mm by 12 mm. The materials supplied by Duralcan Inc.,
San Diego, CA, were fabricated by using a proprietary casting technique. After casting,
the composite materials were hot-extruded to homogenize the particle distribution. The
composite materials will be designated as 6061/A1203l10p and 6061/A1203/20p following
the Aluminum Association standards, where “p” stands for particles [219]. The as—received

extruded samples were subjected to the following heat treatment [220]:

Solution treatment 540°C for 40 to 65 min (depends on the thickness of the test sam-
ple) in air furnace

Artiﬁcial Aging 175°C for 14 h in air furnace.

The chemical compositions of the monolithic 6061 Al alloy [221] and the composite
materials [222] are listed in Table 6.1. Figures 6.1 and 6.2 are optical micrographs of the
as-received extruded 6061/A1203/10p and 6061/A1203/20p composites heat treated to T6
temper. The microstructures of both these composites exhibit a light texture of the matrix

parallel to the extrusion direction. The irregularly shaped A1203 particle size was about 3 to

87

88

25 pm with an aspect ratio of #2. Despite extrusion of the composite samples, Figures 6.1
and 6.2 display a clear clustering (non-uniform distribution) of A1203 particles through out

the microstructures, forming stringer like arrangements parallel to the extrusion direction.

Table 6.1 - Nominal composition of the 6061 Al alloy, 6061/A1203/10p, and
6061/A1203/20p composite matrix in weight percent.

 

 

Element 6061 Al 6061/A1203/10p 6061/A1203/20p
Si 0.4 to 0.8 0.59 0.58
Fe 0.7 max. 0.11 0.08
Cu 0.15 to 0.40 0.27 0.25
Mn 0.15 max. 0.04 0.03
Mg 0.8 to 1.2 1.12 1.11
Cr 0.15 to 0.35 0.10 0.10
Zn 0.25 max. 0.02 0.02
Ti 0.15 max. ' 0.004 0.005
Ni - 0.001 -
Others 0.15 max. total - -

 

 

 

 

 

 

To compare the effect of the inhomogeneous distribution of the A1203 (9) , a different set
of samples that were hot-rolled and heat treated to T6 condition were also prepared. The
thickness of the hot-rolled samples varied from 0.8 to 3.5 mm. Figures 6.3(a) and (b) show
typical optical micrographs of hot-rolled 6061/A1203/10p and 6061/A1203/20p composite
samples respectively. Figures 6.3(a) and (b) exhibit a more homogeneous distribution of the
A1203 particles both in the rolling and transverse directions. For hot—rolled samples, the
bonding between particles and matrix was enhanced as evidenced during metallographic
surface preparation; for extruded specimens most A1203 particles were pulled out, but for
hot-rolled samples the particles remained intact. Prior to the start of the experiment, the test
samples were cleaned in an ultrasonically stirred bath of acetone, to remove dirt and any
oily or greasy layers from the surface of the samples. The hot-rolled composite samples

were used only for laser beam welding.

89

 

Figure 6.1 -Optical micrograph of a polished as-received extruded 6061/A1203/10p-T6
composite.

90

 

 

Figure 6.2-Optical micrograph of a polished as-received extruded 6061/AlgOg/20p-T6
composite.

91

 

Figure 6.3-Optical micrographs of polished hot-rolled (a). 6061/Al203/10p-T6 and (b).
6061/Al-303/20p-T6 composite in the rolling direction.

92

A continuous wave C02 laser operating in TEMm- mode at a power levelof 200 to 2500
W was used for machining and joining experiments. A 2500 L Model ~Trumpf (T rumpf In-
dustrial Lasers Inc., MA) laser with characteristics given in Table 6.2 was used for the

experiments. A schematic representation of the laser set-up used for the composite ma-

Table 6.2 - Characteristics of the RF excited C02 laser beam.

 

 

 

 

 

Parameter ' Value/Type

Type of Laser C02 laser

Mode of the beam TEMOI-

Output stability :1:2%

Beam dia. (raw) z 18 mm

Beam dia. at focussing optics z 36 mm (with beam expander)
Excitation RF excited

Type of laser Fast ﬂow coaxial laser

Temporal mode Continuous and pulsed (1-10 kHz)
Max. output power 2800 W (continuous)

Spot dia. 250 pm (for 25.4 cm FL reﬂective optics)
Gas consumption He -—> 641/};

 

chining experiments is illustrated in Figure 6.4. Eventhough the schematic illustrated in
Figure 6.4 shows a transmissive optics (19.0 cm focal length (FL)) in place, experiments
were also can'ied out using a 25.4 cm and 19.0 cm FL reﬂective optics and 12.7 cm FL as-
pherical lens . At the present time it is difﬁcult to ﬁnd commercially available transmissive
optics, which can withstand high power C02 laser for long and continuous exposure times.

It was reported in literature that in the absence of a polarizing element in the path of the
raw beam, the radiation polarization reﬂected from the specimen being processed back to
the resonator cavity, affects the degree of laser radiation polarization [223]. Hence for all

the experiments, a circular polarizer was installed as outlined in Figure 6.4.

93

t beam of the laser, Circular Polarizer/Beam bender

(dia~ 18 m)
I \

Beam Expander (1:2)
(dia of beam ~ 36 m)
t

 

 

V V

 

 

r Water cooled
lens holder

n n

 

Water in

 

Water out 4—___-: mt...— <-—
LL *1 - I 1 . .
~ \ - - ., Memscus Focusmg Lens (ZnSe)

 

 

 

 

 

 

 

 

 

 

 

 

 

 

Protective housing
M, (anodized Al tube)
Cover G
R l l 11 / D
Jet Standoffl 141 III
......... (I: Focused laser beam
1-----.tv ‘- . .
Di Kerf Workpieoe
the focal point
from the surface {,1 \‘83
~ Dross

Figure 6.4 - A schematic representation of laser setup for machining and single beam LBW.

94

Environmental temperature changes that are large enough to affect the temperature
equilibrium of the laser enclosure can result in loss of power, change of mode, and change
in beam direction. A change of temperature equilibrium in the laser head can cause ther—
mal distortion of the resonator components, altering resonator alignment, hence affecting
the laser beam. The prime cause of these problems is the external environment. In almost
every case the cure is to restore the temperature equilibrium in the interior of the laser
and to maintain the ambient room temperature at around 25 to 28 "C. The laser machine
was allowed to warm up for half an hour, to take care of the initial instability in the laser
pumping process and the room temperature was maintained around 28 °C.

A high precision X-Y table with a computer controller (General Numeric Inc.) and
with the following speciﬁcations, indicated in Table 6.3 was used for laser machining and
joining experiments. The specimen mounted in a ﬁxture (that was secured on the X-Y

table) was translated under a stationary laser beam.

Table 6.3 - CN C machine speciﬁcations.

 

Description Value

Range of motion 1.27 m x 1.27 m
Maximum traverse speed 0.254 m / sec
Maximum contour speed 0.127 m / sec

 

 

 

 

Accuracy 0.0166 cm / m
Resolution 0.0127 cm
Repeatability 0.00254 cm

 

Nozzles with different conﬁguration were used for LBM and LBW experiments. The
nozzle exit shape can have a substantial inﬂuence on gas jet effectiveness in machining.
Convergent nozzles of 2.5, 3.0, 3.8 mm were initially tried for the machining of compos-
ites. The diameter of the oriﬁce affects the cover gas jet velocity and hence the cutting

' speed and quality of the cut. After numerous trials it was decided to use a convergent noz-

 

 

 

 

 

(a)

 

 

 

 

 

 

 

 

 

 

(b)

.25

 

 

 

 

 

 

 

 

D025

 

 

 

Figure 6.5 - Laser head nozzles used for (a) machining experiments and (b) single beam
welding experiments.

96

zle with 3.8 mm oriﬁce. The cover gas ﬂow rate varies by changing the pressure of the
gas exiting the oriﬁce. Initially machining experiments were carried out using cover gas
pressures of 300 to 450 kPa. It was found that the maximum cutting speed was achieved
at 415 kPa. The convergent nozzle indicated in Figure 6.5(a) was used for all machining
experiments and Figure 6.5(b) depicts the nozzle tip used for single beam laser welding
experiments. The welding nozzle depicted in Figure 6.5(b) was designed to be held in
a supporting nozzle holder. To protect the molten weld pool, the gas jet exiting the ori-
ﬁce was diverted into a medium pressure central region and a very low pressure concentric

region around the periphery of the nozzle. The nozzle had an oriﬁce of 3.5 mm.

6.1 Machining

All the 6061/A1203/20p specimens were cut using the experimental parameters given in
Table 6.4. Coaxially ﬂowing gas at a pressure of 415 kPa was used to minimize absorption ‘
of laser energy by plasma, to remove any debris (dross, molten products etc) from the

interaction zone, and to prevent oxidation (when He and N 2 are used as the cover gases).

Thble 6.4 - A list of experimental variables for machining.

 

 

Parameters Value/Type

Power 2500 W

Feed rates 0.25 to 2.8 m/mz’n (10 to 110 ipm)
Gap between the nozzle and sample 500 um

Optics 25.4 cm FL, reﬂective optics

19.0 cm FL, reﬂective optics
19.0 cm FL, transmissive optics
12.7 cm FL, aspherical optics
Cover gas pressure 300 to 450 kPa

Cover gases (co-axial) Helium, Nitrogen

Oxygen, and Compressed air

 

 

 

 

97

To obtain clean and quality cuts a small diameter spot size, and thus a high power
density is required This can be achieved either with a short FL lens, which gives a short
depth of ﬁeld or a a beam expander (an optical device) that increases the beam diameter
and reduces divergence of beam. The beam expander provides a smaller focussed spot for
a larger distance between lens and workpiece. Since a large depth of ﬁeld (which is the
working range of the beam) is needed to cut thicker sections with straight edges, a 25.4 cm
FL lens was used to machine thick composites. It is important to note that the depth of ﬁeld
depends not only on FL, but also on diameter of the unfocused beam and wavelength of the
laser (in the case of C02 it is 10.6 pm). A 25.4 cm FL lens has an approximate depth of
ﬁeld of 1.0 to 1.8 cm (see Section 2.2.1.3).

The beam spot size can also be controlled by using different optics, for example aspher-
ical lens and also by switching to short FL lenses. The effect of the type of optics on the
spot size is clearly portrayed in Figure 6.6. Aspheric optics provides optimal collimation
and focusing spot size. Figure 6.6 portrays the inﬂuence of the raw beam diameter on the
ﬁnal focal spot size, for a given class of optics. Hence, proper selection of optical system
plays an important role in laser processing of composites. ‘

Since the reﬂectivity of pure aluminum is very high, the laser energy gets reﬂected
away from the surface of the specimen, and transfering z20% of its energy to the metal.
The reﬂectivity for 6061/A1203l10p and 6061/A1203/20p composites is less than that of
pure Al. To accommodate this observed phenomenon, trial cuts were made with both
modiﬁed surface layer and as received specimen surfaces. From preliminary experimental
observations, it was concluded that modiﬁed surface layers did not help in increasing the
energy coupling required to process the composite material of a ﬁxed thickness. Hence all
the machining and joining experiments were carried out without any surface modiﬁcation.
The presence of A1203 particulates in the Al composite enhances the beam coupling, as
compared to the base 6061 aluminum alloy matrix [155]. Only results for 6061/A1203/20p
composite are discussed in this report, as 10% particulate reinforced exhibited identical
behavior to the 20% composite in all respects.

98

 

10000
10' FL
1.5" FL
(I)
C
9
25' FL m
C
F 5" FL g
0 O
C a)
A 10' FL :5:
L
s
P 10' FL
0 0)
T 5" FL 5
D 2.5" FL -‘=’
' 2
A 1.5" FL 3
1.0" FL <
(u 0’3
C
N
' 0.0 10.0 20.0 30.0 40.0 50.0 60.0

BEAM DIA (mm)

Figure 6.6 - Schematic illustration of the differences between a meniscus and aspheric lens
in terms of their spot diameter and output laser beam diameter [224].

99

 

 

 

 

 

 

some new .850 E233. 305.33 .31 nﬁ d6 8am
83—0: new 3.60 33860 examines: .31 com d6 comm
an: 3: Bzwmoa 3w mum—33‘ 268%.. .31 ena .Eu 8%
53 v2 953er mam 3:860 $355253 a coma 59:0 Ens—:82
‘E :8 5.2 330 Row—onset com—am a. 3055280 088 1:3th
1E Eu no.2 8qu 033888“. 3:88 Bo: “mam .83— .«o 09¢.
Am :8 v.3 mango 02.85% 3:88 "a cougoxm
E: 08 new cos—moan a. oﬁnoz EE 3 R 96.: .56 83m
SE}: ON I md mega poem swan.” ban—Sm 59:0
\S 8mmu¢ou .826.”— .82m.—. Ewen 05 .8 ope—2
cages—‘3 Sue—SEE omens—Sr egos—Sum
neBetg ~S§§Le§£ Euaeﬁegem .83

 

 

 

.mﬁﬁﬁ not. avian? 3:08:03». 28 E80883 .852 .«o as .. m6 0358

 

100
6.2 Welding

LBW process for the fusion joining of the 6061/A1203/10p and 6061/A1203/20p compos-
ite material without using a ﬁller metal (autogenous) was investigated. Autogenous single
pass bead-on-plate laser weldments were used for weld bead characterization. Two sets of
composite samples with different initial microstructures were used in this study as expla-
nined elsewhere. The other experimental parameters used for the present investigation are
given in Table 6.5.

The laser set-up used for the SB-LBW experiments is identical to the one illustrated
in Figure 6.4. All the welds were made parallel to the extruded and rolled direction. For
the SB-LBW process, the stationary laser beam was traversed over a horizontally held
composite sample. The laser beam spot size at the surface of the sample was manipulated
to achieve power intensities of z 105 to 107 W/cm2 that is required for full penetration
welds. Just before the start of the experiment, the ultrasonically cleaned test samples were
slightly polished with 600 grit emery paper to remove any nascent oxide layer from the
surface. For each experimental conditon, a minimum of ﬁve bead-on-plate welds were

produced.

6.2.1 Plasma Control

An auxiliary gas jet was aimed precisely at the weld pool at a pressure of 138 kPa (20
psi). Argon was chosen as the off-axis gas because of its excellent covering capabilities
(due to its higher molecular weight than any other gas, except C02). The auxiliary gas was
introduced in the weld zone through a nozzle set at an angle to the laser beam, to break
the plasma plume that forms above the interaction zone. In the case of C02 laser welding,
strong plasma absorption can lead to blocking of the beam and result in a weld section
having a characteristic “nail-head” shape. To mitigate this effect a high ionization potential

. gas, such as helium, has to be blown across the workpiece to “shield” the surface. A coaxial

101

gas jet of He at a very low pressure of 104 kPa (15psz') and at a controlled ﬂow rate was
used to protect the optical assembly from the debris.

To suppress plasma shielding of the weldment (especially for SB-LBW), the sample
was held in a ﬁxture at an angle of 85° to the beam. Figure 6.7 illustrates the experimental
set-up used for the CD-LBW of the composite samples. In CD-LBW the test sample is
held vertically between two horizontally focussed laser beams. This modiﬁed process uses
laser heads equipped with unequal FL lenses. The single source laser beam was split by a
beam splitter into two beams of 60% and 40% power levels. A various set of experiments

were carried out with the following conﬁgurations for the two laser heads:

1. Both the collinear (inline) laser focus spots were adjusted to maintain the 60 : 40
ratio in energy densities of both the beams.

2. The 40% strength beam was offset by 0.5, 1.0, and 1.5 mm, ahead of the 60% beam
and the spot size was controlled to accommodate the 60 : 40 ratio.

3. Same conditions as in item (2), except for the spot size. For this case, the 40% beam
was located at various distances from the test sample as to have a beam spot size of

1.0, 2.0, and 3.0 mm on the sample surface.

For case ( 1), it is essential to have a perfect alignement of the two laser beams. A transpar-
ent polycarbonate block was used to check for collinearity of the beams, by taking a burn
pattern on either side of the block. The 60% laser head was equipped with a 25.4 cm FL
reﬂective optics and the 40% laser head was equipped with a 19.0 cm FL reﬂective optics.

6.3 Microstructures and Hardness Testing

The laser processed specimens were sectioned, mounted, polished, and etched using sligtly
modiﬁed standard metallographic techniques. When polishing the cross section of an al—
tered surface, special techniques for mounting and polishing are used to reduce edge round-

ing. Edge retention is important because surface alterations are only about 0.025 to 0.075

102

63858 gonzoessoo a; aoznosssoe co 33-8 2: co SE82: 8.2.23. .3 2:5

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

.vmmmﬂcemmq

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

 

//I.csmmm heme;

103

mm (0.001 to 0.003 in) deep. One technique is to vacuum cast the metallurgical mount
and to use special procedures for polishing. Polishing units with a horizontal specimen
holder, which is rotated with a preselected force against various grinding and polishing sur-
faces, have also been successfully used for edge retention. In this investigation, the samples
were mounted in Bakelite with a backup plate to ensure edge retention, during polishing.
Polishing was done on a Buehler Vibromet 2 vibratory polisher, using the following rec-

ommended procedure:

0 Grind the mounted samples ﬂat with 60 grit SiC paper. Set the polisher for 150 RPM
with water lubrication. Grind for 45 sec each with 180, 240, and 320 grit paper.
Finally rinse with hot ﬂowing tap water, immerse in ultrasonic bath for 15 sec, rinse
with hot water again, and dry.

0 For the second stage set the polisher at .300 RPM with oil lubrication. Polish for 10
min on a cloth prepared with 15 pm diamond. Rinse with hot tap water, immerse in .
ultrasonic bath for 15 sec, rinse with hot water again, and dry. Repeat this stage with
6 pm and then with 1 pm diamond paste.

0 Finally, set the polisher at 300 RPM with no lubrication. Polish for 10 min on a cloth
wet with colloidal silica. Additional silica must be added every one or 2 min to keep
the colloidal silica from drying up. Turn on the water lubrication for the last 30 sec
of polishing to remove the colloidal silica from the specimen and the cloth. Avoid
colloidal silica from getting dry since it is extremely difﬁcult to remove when it is
dry. Rinse specimen in hot tap water, clean in ultrasonic bath for 15 sec, rinse with

distilled water, and dry.

Microstructural investigations of the as-cut surfaces and transverse sections of the as-
cut and weld surfaces were carried out by optical microsc0pel and scanning electron micro-

scope”. The transverse cross—section used for thermal damage characterization is illustrated

 

1 Olympus.
2 Hitachi S-2500.

104

in Figure 6.8(a). Preliminary 'weld experiments were conducted to establish optimal pen-
etration depth and weld width, at a constant laser power and weld speed on samples with
varying thickness as shown in Figure 6.8 (b). Metallography was used to evaluate weld
bead integrity, reveal hot tears, cracks, porosity, reaction products etc. The polished sam-
ples were etched with modiﬁed Keller’s reagent (0.5 9 NaCl, 1 mL HNO3, 2 mL HCl
and 97 mL water) to detect precipitation in the molten and heat-affected zones. The grain
structure in the weldment was revealed by using an etchant consisting of (a) 12 g NaOH,
100 mL water + (b) 2 to 3 drops of HF, 100 mL water. These are standard reagents used
for aluminum metallography.

It is also possible to study thermal damage by measuring micro-hardness values in the
immediate vicinity of the cut zone. Microhardness measurments were obtained using a
Leco Hardness Tester (M-400-G1) equipped with a Vickers diamond pyramid indenter. An
average of 10 hardness readings were taken for each test sample, with care being taken
to avoid any contact between the indenter and the reinforement particles. The Vickers
hardness readings were taken at a depth of 0.5 mm below the surface of the weld head,
with 100 9 weight.

Roughness measurements were done by using surface proﬁlometer. A cone-shaped
diamond stylus was run over the as-cut surfaces, generating the proﬁle of the surface pro-
jections. These proﬁles are produced using a x-y plotter, which were then digitized and
transferred to a PC for further analysis.

6.4 Finite Element Analysis (FEA)

To study the inﬂuence of the various variables mentioned in Table 6.5, on the CD—LBW
process, a 3-D ﬁnite element transient heat transfer analysis was carried out. A commerical

ﬁnite element program, MARC was used for the analysis. The governing equation for the

105

 

AsCutSurfaee

 

 

   

vmammz C «the

mum -/

fraudulent

(a)

 

 

 

Wel
bead

 

 

 

(b)

Figure 6.8 - (a) Schematic of the composite specimen displaying the cross—section used for
SEM studies and (b) schematic of the varying width composite specimen.

106

3-D heat transfer analysis is:

6 6T 6 6T 6 6T 6T
5; (Kl-5;) + '63 (KyB-g) + 5; (K25?) + Q - pCpux‘a'? (61)

This non-linear partial differential equation (PDE) (see Equation 6.1) for the transient prob-
lem has to be linearized to make the ﬁnite element analysis less computer intensive. In
Equation 6.1, Q is the internal heat generation term or heat input to the workpiece, K is
the thermal conductivity, C, is the speciﬁc heat, T is the temperature, p is the mass density,
and it; is the welding speed. Transient heat-transfer problems involving melting or so—
lidiﬁcation are generally referred to as “phase-change” or “moving-boundary” problems.
In generally, the boundary conditions associated with change of phase (i. e., melting, so-
lidiﬁcation, and ablation) are also non-linear boundary conditions. The above non-linear
PDE with non-linear boundary conditions is very complex to compute using numerical

techniques. Hence the following assumptions are made to simplify the above problem:

0 The composite material was assumed to be isotropic, since the Al-MMC used in the

present study was a particulate reinforced composite.

0 Thermal conductivity (K) is assumed to be constant (i. (3., independent of position and
temperature), speciﬁc heat (C1,), and mass density (p) are assumed to be independent

of temperature.
0 Initial temperature of the body was at room temperature (25°C).

0 The heat ﬂux (Q(x,z)) was assumed to be constant with time and position, eventhough

it was known to have Gaussian distribution.
0 The beam was assumed to be circular in shape.
0 The composite material moves at a constant relative velocity.

0 Composite material has constant absorptivity that does not change with time and

_ temperature.

107

o Molten liquid pool formed in the weld zone does not interfere with the incident laser

beam.

With these assumptions, the governing differential equation for laser joining reduces to:

62T 62T NT 1. _pc,u,§:r__%ar

37+? '52?+EQ‘—K arc-a5; ‘62)

where a = p—CK—p is the thermal diffusivity. In the above equation, the time dependence
of temperature was transformed into a spatial dependence and Q was assumed to be a
continuous moving surface heat source. The basic steady state heat conduction equation
with different source terms has been analytically solved by'many authors [225]. The dual
laser beams are assumed to be parallel to the z-axis and move at a constant velocity, it;
along the positive z—axis. The primary laser beam (1500 W) was on the top surface of the
plate (in the positive z-axis) and the secondary beam (1000 W) acts on the bottom face
in the negative :c-axis. In one’case, both the beams were collinear and for the other case
the secondary beam was oﬂ'set from the primary beam by 1.0 mm along positive x-axis
(pre—weld heat treatment). Heat loss from the surface and the boundaries are assumed to
be by convective heat transfer, eventhough these losses are negligible compared to the heat

conduction into the composite material. Only the dual laser beam welding process was

modeled.

6.4.1 Finite Element Model

Figure 6.9 shows the ﬁnite element discretization of the 50 mm x 25 mm x 6 mm block
of Al-MMC composite sample. A reﬁned mesh should be used near discontinuities and
surfaces exposed to large thermal gradients. A gradual transition to a coarser mesh is
appropriate in areas removed from the thermal load. Since the center regions are subjected
to higher temperature gradients as it is located right below the laser beams, ﬁner mesh is

used in this section. There are a total of 880 3-D 8-node brick elements (Element No. 43 of

108

MARC) and 1260 nodes in the model. The ﬂux was assumed to be uniformly distributed
on each element for 0.2 s. This time increment was determined by the magnitude of the
traverse speed of the laser beam. Results are presented only for a traverse speed of 13
mm / s. A number of time step schemes were tried, before ﬁnally settling for 6 time steps
in each time increment. Automatic time stepping feature of the MARC program was not
used, since the time steps being selected were very small (about 10‘65). The material

properties listed in Table 6.6 were used for the analysis.

Table 6.6 - Average values of the thermo-physical properties of 6061/A1203/20p 6061 Al,
and A1203 [221, 222, 226, 227].

 

 

Property 6061/A1203/10p 6061 A1 A1203
Thermal conductivity (W/ m -° C) 179.9 (at 260 0C)

Mass density (kg/m3) 2810 2700 3990
Speciﬁc heat (J/kg 0C) 883 1100 750
Melting point (0 C) 660.27 (Al) 660.27 (A1) 2049
Boiling point (0C) - 2467 -
Latent Heat of Fusion L f (J/kg) - 39.54x 103 -
Latent Heat of Vaporization, L,, (J/kg) - 12.1 x 105 -

 

 

 

 

 

 

."
. '-‘r

.“u _-‘

.'.w“b;L'.

ﬁg!"

"au
. a \ . . K‘
g ..
‘Q‘rvi‘ . :‘f" "£4
recesses:
. x. ‘14.! .t‘ltr‘h
. .i - J ‘ .3“ ‘2‘ 3v

)6

7 4 w
lik‘ .‘

7 p A‘ - e
(‘3‘; “#33 ‘3‘de
'3’.“ 5&m We?“

 

Figure 6.9 — Finite Element mesh with a total of 880 “8—node” brick elements.

CHAPTER 7

RESULTS AND DISCUSSION

Aluminum alloy based composites are characterized by a low absorption factor for the
co2 laser beam and high thermal conductivity. When 002 laser cutting and joining is
used for these materials, they produce oxides with high melting points, liquid metal with
low ﬂuidity, and dross with poor peeling. As a result, these materials are very hard to
cut and join with laser beam. Cutting techniques may result in mechanical, thermal, or
metallurgical alterations on the surfaces produced. Joining process may generate excessive
porosity, microcracking, HAZ, interfacial reaction products or loss of alloying elements.
So far no detailed investigations had been undertaken to study the cut surfaces and the
weld bead integrity and quality, produced by C02 laser beam on Al-MMC composites.

The present study tries to address some of these aspects.

7 .1 Energy Required for LBM and LBW

To gain a preliminary knowledge about the power requirements for machining and joining
of 6061/A1203/10p and 6061/A1203/20p composites, a few numerical calculations, based
on a simple energy balance relationship were computed as shown below, with the data

provided in Table 6.6. From simple energy balance considerations, the amount of heat

110

111

input necessary to raise the temperature of a unit mass of material can be computed from:
q=CAT+LI or (7.1)
for a mass m = pV, we can write:
E = pV(CAT + Lf) (7.2)

Where E is the actual energy required to produce the weld nugget, p is density, and V is the
volume of the weld nugget (which is determined by shape of the nugget), L f is the latent
heat of fusion, and AT = Tm — To, where Tm is the melting point temperature, and To is
the ambient temperature. For a thermal conduction mode weld (see Figure 5.3) of 1 mm

deep nugget with hemispherical shape (typical of conduction mode welds):

-15 3 i
V— 2(371'7‘) (7.3)

where ‘r’ is the nugget depth or radius. For this bead size (1' = 1 m) it was found that
E = 3.53 J. So for an assumed weld efﬁciency of 0.5 and reﬂectance of 0.8, E = 35.3 J,
i.e., an increase by a factor of 10 ( (weld efﬁciency) x (l-reﬂectance)= 0.5 x 0.2 = 0.1).

Modifying the above equation for continuous seam welding, with
V = wld (7-4)
where ‘w’ is the width, ‘l’ is the length, and ‘d’ is the depth of the weld bead.

P = 5:- : pw-i—d(CAT + L ,) = pwvd(CAT + L,) (7.5)

112

where ‘v’ is the traverse speed. Using this equation, one could compute that for a for
2 mm deep and 0.75 mm wide weld seam, with 2.54 m/mz'n (100 ipm), the required
power would be: P = 107.15 W. So multiplying this by 10 (to account for the reﬂectance
and weld efﬁciency), would give P = 1071.5 W. For machining applications, the above
equations can be modiﬁed. Denoting E/V (see Equation 7.2) as Q (speciﬁc energy) and

assuming that the material removal process is by ablation alone, we get:

0 = p(c,(T, - T.) + L; + L.)

E = QV = led
P = g = de- = dev
P = dev ' (7.6)

where C, is the speciﬁc heat, T, is the vaporization temperature, To is the ambient tem-
perature (assumed to be 25 0C), L f is the latent heat of fusion, L” is the latent heat of
vaporization, and ‘v’ is the speed of cut. So for a kerf width of w = 0.1 mm, thickness
of d = 6.35 mm, and velocity of v = 4.25 cm/sec, gives P = 258.28 W and assuming
that ten times this power is required for cutting (as in thermal conduction welding), gives
P z 2583 W. A more reﬁned model is presented in Section 7.2.3.1.

7.2 Laser Beam Machining

With the information obtained from the preliminary calculations given above, a set of ma-
chining and joining experiments were conducted and a few signiﬁcant results are presented
below. Results of LBM of 6061/A1203/20p as—received composite are presented in this
section (hot-rolled particulate composite did not result in any signiﬁcant improvement in

thecut quality). As explained in Section 1.2, high volume percentage of the hard reinforce-

113

ment in the composite, result in rapid tool wear in traditional machining techniques. Hence
the results presented in this report are those of the 20 vol% particulate composite, since the
reinforcement content and amount is immaterial for LBM and the results for the 20 vol%
composite are identical to those of the 10 vol% composite.

For the laser beam machining, the effect of type of cover gas and its pressure on the
roughness of the cut, kerf width, and cutting speed are presented. An analytical model was
derived to determine the inﬂuence of cutting speed on thickness of the cut and to predict
the amount of molten material formed in the interaction zone. A simple theoretical model
was presented in Appendix A, to explain the role of the focal plane position with respect
to the surface of the sample and to compare with experimental result. The role of the
cutting speed on the roughness of the cut and HAZ are outlined in this section. Finally, the
inﬂuence of the type of optics on the cut surface quality is presented.

7 .2.1 Inﬂuence of Cover Gas

In Figure 7.1, the rate of weight loss for 5.0 mm thick specimens cut with 2400 W are
plotted against the cutting speed. The cover gas pressure was set at 415 kPa for all gases
(see Section 7.2.2). This plot shows an exponential decrease of the rate of weight loss, for
all types of cover gases, with the increase in cutting speed. For 02 and compressed air the
rate of weight loss is substantially higher for all the range of cut speeds. Since both 0;» and
compressed air provide oxidizing atmosphere, large amount of Al alloy will get oxidized
and removed from the kerf. The small variation in the weight loss for these two types of
cover gases was mainly due to the diluted concentration of oxygen gas in compressed air.
When the test specimens were cut in He and N2 atmosphere the loss in weight is very
small. Since both these cover gases are inert in nature, this behavior could be expected.
With He as the cover gas, the rate of loss is slightly higher in comparison to N2 case.

The reason for this might be the higher thermal conductivity of He gas than N2, which

114

 

0.45

T O Nitrogen cover gas

V Helium cover gas

*6 Compressed air cover gas
A Oxygen cover gas

0.35

I

0.25

Rate of weight loss (g/sec)

0.15

 

 

 

 

0.05 1 l r L 1 J l l l l
0.2 0.7 1.2 1.7 2.2 2.7

Cutting speed (m/min)

Figure 7.1 - Variation of rate of weight loss with cutting speed for 6061/A1203/20p com-
posite of 5.0 mm thick, with a power of 2400 W and cover gas pressure of 415

kPa.

115

aids in increasing the plasma temperature and also providing a better coupling of laser
energy with the specimen surface. Visual examination of the cut surfaces obtained with
02, compressed air, and He indicates rough cut surfaces with adherent dross at the exit side
of the cut. Taking this experimental evidence into consideration, N2 gas was considered
to be the ideal cover gas for this composite material. As discussed later, it was found that
for a 5.0 mm 6061/A1203/20p composite, the optimal cutting speed was 0.508 m / min (see
Sections 7.2.2.2, 7.2.3, and 7.2.5). At this cutting speed, the rate of weight loss in g / sec for
nitrogen, helium, compressed air, and oxygen cover gases was 0.27, 0.29, 0.35, and 0.38
g / sec respectively. The optimal cutting speed was decided based on cut surface roughness
and kerf width both at the entrance and exit side of the beam.

7 .2.2 Role of Cover Gas Pressure
7.2.2.1 Effect on Roughness.

Figure 7.2 exhibits an optical micrograph of the 6061/A1203/20p composite cut surface.
The cut was made at 0.508 m/mz‘n with N2 cover gas at 415 kPa pressure and 2400 W
power. The upper exposed surface shown in the micrograph was cut at a reduced pressure
of 300 kPa. At lower pressures, the roughness of the cut and kerf width were dramatically
increased.

The surface quality of the cut is an important aspect that requires special attention.
The pressure of the cover gas has a direct bearing on the thickness of the material that
can be cut through — it is directly proportional to the cover gas pressure. The inﬂuence
of the cover gas pressure on cutting action, and surface roughness is not very clear at this
point. It can be speculated that at reduced pressure level, the plasma cloud that forms in
the blind kerf and above it, is some what more stable than that at high cover gas pressure
levels. This stability leads to more energy being available at the cut zone, hence wider kerf

width. Large amount of molten material accumulates at the mid-zone of the kerf, due to

1l6

 

Figure 7.2 - Photomicrographdisplaying the kerf width and cut surface for a 5 .0 mm thick
6061/A1203/20p composite, with nitrogen cover gas at 415 kPa and cut at
0.508 m/mz‘n (indicated by ‘X‘). The exposed surface was cut at 300 [(1%
(indicated by ‘0‘).

117

low pressure levels, leading to non-uniform roughness. Figure 7.2 shows that straight cuts
with narrow kerf width of 0.8 mm at top edge and 0.5 mm at the bottom edge (not shown
in the micrograph) could be obtained at cover gas (N2) pressure of 415 kPa. Improved
edge and surface qualities are important processing beneﬁts at high pressures. The average
surface roughness of the cut at 0.508 m/ min was in the range of 1.0 to 1.3 pm for 415

kPa cover gas pressure and for the cut made at 300 chr, it ranged from 1.8 to 2.3 pm.

7.2.2.2 Cutting Speed and Kerf Width

Another important effect of cover gas pressure is on the cutting speed and kerf width. With
N2 as the cover gas, the cutting speed increased with cover gas pressure upto certain point
(415 kPa) and then started to reduce with increasing pressure (450 kPa) for a 5.0 mm
6061/A1203l20p composite. The gas jet “stand-oﬂ” distance was maintained at 0.5 mm.
The reduced cutting speed at high gas pressures (450 kPa) can be attributed to increased
cooling effect at high gas ﬂow rates.

Another factor that could affect the cutting performance at high gas pressures would be
the presence of discontinuities, such as shock waves, in the gas jet. These discontinuities in
the gas jet could lead to density changes in the cover gas ﬂow ﬁeld (refer to Section 7.2.4)
and effect plasma formation and stability. The kerf width increases with increasing gas
pressure as indicated in Figure 7.3 for nitrogen cover gas at 415 and 450 kPa. The data
shown in Figure 7.3 for a 5.0 mm thick 6061/A1203/20p composite was collected at a ﬁxed
cutting speed of 0.508 m/mz'n. For 2400 W power, the kerf width at 415 kPa nitrogen
cover pressure was 0.8 mm and at 450 kPa, it was 1.24 mm. This modest increase in cover
gas pressure, from 415 kPa to 450 kPa, resulted in a substantial increase (about 1.5 times )
in kerf width at all power levels. There was ample reduction in surface roughness of the cut
for samples with narrow kerf widths. For narrow kerfs, the plasma and the cover gas were
forced to exit from the bottom surface of the cut along with the molten metal and dross.

Despite of the minute increase in gas pressure (from 451 to 450 kPa), the large difference

118

in kerf widths was due to the enhanced gas ﬂow rate at the higher pressure of 450 kPa.
At high cover gas pressure, the faster ﬂowing gas would also create greater drag forces in
the kerf, leading once again to wider kerfs and higher roughness values of the cut surface.
The advantageous side effect of utilizing high gas pressures was in the sizable reduction of
the heat affected zone. It was also observed that the kerf width slightly increases with a

decrease in cutting speed, since the laser interaction time is higher at lower traverse speeds.

 

 

 

 

 

 

E 0.8 ~
3 t
29.
i
M 0.6 h
H Nitrogen gas pressure = 415 kPa
-+——+- Nitrogen gas pressure = 450 kPa
0.4 ~
thickness = 5.0 mm
0.2 -
0.0 A J u _l_ n l L _J_ A.
1600.0 1800.0 2000.0 2200.0 2400.0 2600.0

Laser power (watts)

Figure 7.3 - The variation of kerf width with laser power and cover gas pressure for a 5.0
mm thick 6061/A1203/20p composite cut at a ﬁxed cutting speed of 0.508
m / min.

119

The gas jet working distance is also an important factor. If the cover gas jet is too
close to the sample (< 0.5 mm), extreme back pressures would be created on the optics, in
addition to splattered dross particles attaching to the nozzle tip and blocking the beam from
interacting with the sample surface. If the cover gas jet (i.e. the working distance) is too
far away from the sample surface, then there would be unnecessary loss of kinetic energy

of the gas jet, which is essential in cutting process.

7.2.3 Thickness Vs. Cutting Speed

Figure 7.4 shows the relationship between the thickness of the composite specimen that can
be cut through with the laser power maintained at 2400 W. Commercial weld grade N2 at
415 kPa was used as the cover gas, since this led to lower weight loss, hence narrow kerf
width. Also the length of the clinging dross at the exit side of the cut was smaller and it
was easier to ﬁle away from the bottom surface of the cut, with N2 as cover gas. But it was .
reported in literature that mechanical removal of dross led to inferior surface properties for
alumina substrate materials used in electronic industry [228].

At a constant power of 2400 W, the thickness of the material that can be cut shows
a exponential decrease with respect to cutting rate. At higher cutting rates the interaction
time of the beam with the material is very low, hence through-cut is possible only in thin
materials, at these speeds. In other words, at a given power level, higher speeds means less
amount of energy is available to remove the material in the cut zone. Also the interaction
time of the laser beam with the material is low at higher speeds. The thickness of the
material that can be cut, in addition to the cover gas presSure, depends both on the laser
power and speed of the cut. Also for some materials combustion-assisted cutting (using for
example oxygen or mixture of oxygen and acetylene) has inﬂuence on the thickness that can
be cut. But for 6061/A1203/10p and 6061/A1203/20p composite materials, using oxygen

as the cover gas had a very deleterious effect on the cut surface, kerf width, roughness of

120

 

 

 

 

12.0
I CW carbon-dioxide laser power = 2400 W
10.0 r I
‘5:
“a
5 I
3
“U
.5 I
I
g 8 0 — I
2 I
2 I
«5 I
t I:
E I
i5 I
6.0 ’ I
I
I
4.0 i l i l i
0.0 1.0 2.0 3.0
Cutting speed (m/min)

Figure 7.4 - Dependency of thickness of 6061/A1203/20p composite on cutting speed.

121

the cut surface, and dross adherence at the exit side of the cut. Compressed air is also not
a suitable cover gas. If air is used as an assist gas, the drag lines or striations, typical of
laser cutting, are slightly curved in comparison to cuts made with N2 as the cover gas. This

variation in the striation pattern resulted in higher average surface roughness of the cut.

7 2.3.1 Mathematical Model

To compare the above experimental results with theoretical data, a simple analytical model
based on a work by Mukherjee er al. [25] was modiﬁed to apply to composite cutting, to de-
termine the maximum depth of cut for a given traverse speed. The generalized differential
equation of heat conduction in a moving medium can be simpliﬁed to a one-dimensional
model by assuming that the applied heat ﬂux is ﬁxed (time-independent, hence g— = 0)

and is given by:
E _ 921 __

V2 02 a 822

0 (7.7)

where V2 is the velocity component in the z-direction, ‘T’ is the temperature, and a is the
thermal diffusivity. The geometry and coordinate system are shown in Figure 7.5. At the
surface of the solid, 2 = 0. At a high enough energy density, the cutting process could be
approximated by mixed mode of melting and ablation process; and if the molten material
formed in the interaction zone (kerf) is constantly blown away (strong co-axial jet), the
solid at z > 0 can be regarded as moving toward the laser beam at a speed —Vz, since
relatively speaking, laser beam reaches a new surface instantaneously. Thus Equation 7.7
can be modiﬁed to reﬂect this assumption:

9: + a
62 6.22

9:.

V2 — 0 (7.8)

122

Laser heat ﬂux

dia = 2'. Circular laser beam

I, ll radiusr

 

 

 

 

 

 

3’
E
v

 

 

Vx

Figure 7.5 - Schematic diagram showing cutting parameters and the coordinate system used
for the model [25].

Solving Equation 7.8 and applying proper boundary conditions (when 2 = 0, T = Tm and

when 2 = oo, T = 0), it can shown that
f. E 0

where f, is the heat ﬂux in the z-axis. This indicates that the heat ﬂux below the plane
of laser-material interaction is zero and the external heat ﬂux F, is used to remove the
material by melting and some evaporation (at high energy density and high cutting rate)

and the molten layer is blown away from the kerf. From heat balance, we know that:
F = (20.- + pom. — Tn) V. (7.9)

where F is the applied heat ﬂux (laser) in energy per unit time per unit area, 2Q,- represents
latent heat of evaporation and latent heat of fusion, To is ambient temperature, and p is the

mass density of the composite material. Changing the power of the laser beam to power

123

density (i.e. ﬂux), Equation 7.9 can rewritten as:

P

V, = A (7.10)
(L, + pC,(T,, — m)

 

where L f is the latent of fusion, ‘P’ is the laser power in watts, and ‘A’ is the area in
cm2 of the focal spot. For laser processing, there is always some melting and also some
evaporation of the material that interacts with the laser beam. In the present model to reﬂect
this aspect, we assume that a fraction 6 of the mass goes into melting and a fraction 7 goes
into vapor state. Also 3 + '7 = 1. Incorporating these two mass fractions (6 and 7) into the

heat balance equations we get:

bl'o

 

V, = (7.11)

[p(pc,(rm — To) + L,) +1(pc,(T, — Tm) + L,)]
where 6 and 7 = 1 — ﬂ is the mass fraction of the liquid metal and vapor phase respectively,
L, is the latent heat of evaporation, T, is the boiling point. The kerf width is assumed to
be constant in this model (i. e. the volume of the material removed is a constant). In the
laser-material interaction, the absorbed power P is given by P = aP,-, where P, is the
incident power and a is the absorptivity (a can also be stated as a = (1 — R), where R is
the reﬂectivity of the material for the infra-red beam). Assuming a circular spot of radius

r and modifying the heat units of J / cm3 to J / gm, the above equation can be expressed as:

V. —_- “P‘ (7.12)

71-1'2p [,6 (Cp(Tm - To) + Lf) + 7(CP(Tv - Tm) + L”)]

 

For a through-cut model, the variables are indicated in Figure 7.5. In this ﬁgure, V, is the
velocity of the molten front in +z direction. Assuming t, max as the time for the molten
metal wave to travel from 2 = 0 to z = Zmax and t, as the time to traverse a distance 2r,

it is necessary for through-cut that t, g t, max- In the limiting case of t, = t, max, it can

124

be shown that

21'
Zmax = "V—Vu
3

Assuming V, and V, to be the same, we can rewrite the above equation as:

V. _ ZmaxVx

 

 

 

 

2r
ZmaxVx ____ GP"
21' 7”.2p[5(01,(Tm — To) + Lf) + 7(CP(Tv " Tm) + L”)]
zmax = 201’.-
nrv,p[p(c,(rm — To) + L,) + 7(CP(Tu — T...) + L)]
zmax = 2aPt (7.13)

 

7rrV,p[ﬁ(Cp(Tm — To) + L!) + {1 - ﬂ} (CP(Tv - Tm) + L,)] .

From Equation 7.13, for a given power, the following proportionalities hold:

1 1
2111an -— and Zmaxa -.
V, T

It is also clear from the above equation that the depth of cut is small for materials with
high melting point and high latent heat of fusion. A similar expression for machining was

reported in literature as indicated below [21]:

= 201’,-
ﬁdV,p(C,(Tm — To) + L ,) '

 

Zmax (7.14)
The predicted values from the Equation 7.14 are much higher than the experimental val-
ues for low traverse speeds and also at high traverse speeds the predicted values are much
lower than the experimental values. This contradicting effect can be accounted for in Equa-

tion 7.13, by assuming different values for either 6 or 7. Steen et al. [229] have also

125
provided an expression for predicting the maximum cutting speed as reported below:

aPt
nwlép(Cp(Tm — To) + L , + m’L,)

 

Zmax =

where n is the reciprocal of process efﬁciency, ‘w’ is the kerf width, and ‘m’ ’ is the mass
fraction of the molten material which subsequently evaporates. In this expression, the initial
beam spot size is completely ignored, but account is taken of the kerf width (which is harder
to measure). The analytical model presented here agrees better with the experimental data.
The values of the maximum depth of cut for a given traverse Speed were computed using I
Equation 7.13 with the material properties given in Table 6.6 and were plotted in Figure 7.6
along with the experimental data presented in Figure 7.4.

Absorptivity of a material is not a constant value as assumed (a value of 10% was
assumed for absorptivity) in the above model to plot the theoretical curves indicated in
Figure 7.6. It is a complex function of heating time, incident power, optical and thermal
pmperties of the specimen, and its surface treatment [230] . In Equation 7.13 the inﬂuence
of all these parameters on absorptivity are ignored. From Figure 7.6 it can observed that
from a cutting speed of 1.27 m/mz'n (50 ipm) to 2.8 m/ min (110 ipm), the major cutting
mechanism is by melting (60 to 80% of the material is in molten state). In this range all the
experimental data lies within the theoretical curves of 6 = 0.6 and 5 = 0.8. For a traverse
speed of 1.0 m/mz'n (40 ipm) to 0.9 m/mz’n (35 im), the major cutting mechanism is
by equal amount of melting and ablation, as indicated by the experimental data enveloped
within the 7 = 0.5 theoretical curve (lower part of Figure 7.6). The optimal traverse
speed of 0.508 m / min is the transition point, where about 10% of the material goes into
molten state and the remaining material is removed by ablation (i. e. initially liquid metal is
formed and later depending on the traverse speed, some of the molten material transforms
to vapor state. The term ablation refers to this phenomenon and not to direct solid —->

vapor transition. Refer to Section 7.2.6 for the two types of cutting mechanisms). Below

126

4G I I I I l

 

35 - 1-0 * Experimental data - -

30 - — Predicted curve -

Thickness MMC (mm)
m M
0 (II

_L
01

.5
o

 

 

 

 

 

1 .5 2 2.5 3
Cutting speed (m/min)

Figure 7.6 - Comparison of theoretical and experimental data of depth of cut versus cut-
ting speed. The values indicated on the solids lines represent ,8 values and
absorptivity is assumed to be 10%.

127

this traverse speed all the material that interacts with the laser beam will be removed by
pure ablation process. The value of 6 (and hence 7) varies with traverse speed. It would be
lower at low traverse speeds and low at high traverse speeds. In other words, as the traverse
speed increases the material removal mechanism goes from pure ablation to pure melting
process. Figure 7.7 depicts the relation between 6 (7) and the cutting speed. The [3 values
were determined by interpolation of the experimental values for depth of cut. The solid line

indicates a curve determined by polynomial ﬁt and it is given by:
p = —0.0909V,;1 + 0.7331V,3 - 2.26551;2 + 3.30341/,, — 1.1658 (7.15)

where V, is the cutting speed. This expression for 6 (Equation 7.15) is substituted back
into Equation 7.13 and the maximum depth of cut was determined for each experimental
cutting speed. These predicted values for thickness of cut are plotted in Figure 7 .8 along
with the experimental data given in Figure 7.4.

Beta: —0.0909 V“ + 0.7331 Vx"3 - 2.2655 VXAZ + 3.3034 Vx - 1.1658

 

0.9

 

 

 

o L 1 I l
0.5 1 1.5 2 2.5 3
Cutting speed (tn/min)

 

Figure 7.7 - Nonlinear relationship between ,6 and the cutting speed. Absorptivity, a =
10%.

128

Beta- -0.0909 VXM + 0.7331 VW - 2.2655 Vx"2 + 3.3034 Vx - 1.1658

14 f I 7

 

13"

12*-

11'

.5
O
I

Q
ﬁ

Thickness ot MMC (mm)
@

 

 

 

 

4 l L

0 0.5 1 1.5 5 2:5 3

Cutting speed (m/min)

Figure 7.8 - Predicted and experimental values of maximum thickness of composite that
can be cut through. Absorptivity, a = 10%.

Despite numerous limiting assumptions made in the derivation of the analytical model
presented here, the predicted values are in good agreement with experimental observation.
In Figure 7.8, the predicted values at 0.25 and 0.3 m/mz‘n cutting speed are 13.4 and
12.0 mm respectively and the corresponding experimental values are 11.0 and 10.2 mm
respectively. This deviation of the predicted values at low traverse speed is mainly because
of the assumed constant kerf width. In general the kerf width decreases with increase in
cutting speed. At low traverse speeds, the kerf width is much larger than at high cutting
speeds; the size of the heat affected zone is large; and the amount of molten material that
is formed is high. All these factors contribute to the high predicted values at low traverse
speed. The model presented here enables us to predict with great conﬁdence the amount of
material that goes into molten state (Equation 7.15).

It was reported that the absorptance increases with decreasing traverse speed [230],

129

since a nascent oxide layer forms on the Al-alloy surface as the beam proceeds along. The
porous alumina oxide layer on the surface increases the laser-material coupling, thereby en-
hancing the absorptivity. In contrast the absorptance tends to decrease with increasing laser
power, because of the removal of the surface ﬁlm of alumina oxide by evaporation. This
also explains the deviation of the predicted values of the depth of cut from the experimental
values at low traverse speeds. It is important to note that pure aluminum and aluminum al-
loys have very low absorptivity for infra-red beams when compared to the 6061/A1203/10p
and 20p composites.

7 .2.4 Position of Focal Plane

In addition to factors like thermal conductivity, reﬂectance characteristics of the surface,
cover gas pressure, speed, type of cover gas etc., the depth of cut also depends to a large
extent on the position of the focal point from the surface of the work piece (Figure 7 .9).
Operating conditions, such as cleanliness of optics and auxiliary gas pressure at the nozzle,
and the number of mirrors in the beam delivery path all aﬂect the quality and performance
of the cut. The depth of cut displays a parabolic trend with respect to the position of the
focal point from the surface of the composite material, as depicted in Figure 7.9. This
phenomenon is known to occur in all laser machining processes. As pointed out earlier
(Section 2.2.1.3), the beam waist or the depth of ﬁeld for 25.4 cm focal length was about
1.0 cm. So, when the focal plane (i.e. the midpoint of the beam waist width) is on the
surface of the specimen, it is possible to cut specimens of 359 = 5 mm only, as the beam
starts diverging rapidly after this point with decreasing energy density and hence less power
is available for cutting. In this investigation, it was possible to cut a 26.0 mm thick com-
posite, with the focal point positioned on the surface of the workpiece. When the focal
plane was moved into the specimen surface, the depth of cut increased to about 9.2 mm
corresponding to a focal point position of 0.4 mm inside the surface. After this point, the
depth of cut displayed a gradual decrease with focal point position.

130

 

10.0

8.0 _ 1

Depth of cut (mm)
\1
o

6.0 _ cutting speed=0.508 m/min

power = 2400 W
thickness = 11.0 mm

 

 

 

4.0 i i I l i J l 4 r
-0.6 -0.2 0.2 0.6 1 .0 1 .4
Focal point position from the surface of the specimen (mm)

Figure 7.9 - Variation of depth of cut with focal point position from the surface of
6061/A1203/20p composite.

131

In Appendix A, a simple model based on geometrical considerations is provided to
explain the reason for the focal plane to be positioned below the surface of the material.
This model indicated that the focal plane should be located approximately about 1/3rd (of
the thickness) from the surface of the material. But, the experimental results indicate that
the focal plane should be about ten times less than this theoretical value. This validates the
theoretical explanation, since the beam was assumed to be focussed to point rather that to
a focal spot and also the focussing lens was assumed to free from all optical defects.

One plausible explanation for the parabolic variation of the depth of cut with focal point
position was given by Steen et al. [229]. Due to high cover gas pressures and low absorp-
tion of the infra-red laser beam by the composite material, non-uniformity of pressure and
temperature ﬁelds would exist within the plasma cloud formed on the surface of the mate-
rial. Such density gradient ﬁelds (DGF) would vary the refractive index across the plasma
cloud, aﬁecting the focusability of the optical energy from the laser beam as indicated in
Figure 7.10. The formation of the DGF would result in either re-focussing (Figure 7.10(b))

(a) (cl
\/ 09515“,
, , . adient
0?: "F" . Fictdtoon , ‘ 5011920“!
.313?“ ’ ‘ Deter ussedl bea-
. -l 1*—
0;

lb)
~I t—
I}

I

Hoﬂrpiere surface

Figure 7.10 - Formation of density gradient ﬁeld on the surface of the material [229].

or scattering (Figure 7.10(c)) of the laser beam on the surface of the material. Either way,
the beam size would increase and the focal plane would shift away from the surface of the
material. This would lead to an increased incident beam diameter that results in reduced

melting efﬁciency and hence wider kerf width and low depth of cut. For the case where the

132

focal point is positioned below the surface of the workpiece, the focal position would be
moved closer to the surface by the DGF (either by beam scattering or re-focussing) and a
complex set of interactions would come into action. Since maximum energy is concentrated
at the focal point, a “blind hole” or a “key hole” will be formed inside the specimen sur-
face, as soon as the beam hits the specimen. Within this newly formed key hole, laser beam
undergoes multiple reﬂections from the molten walls of the hole creating an intense plasma
cloud (“self-focussing effect”). The plasma cloud created inside the hole will be heated to
extremely high temperatures by the incoming beam and also by the reﬂected beam. With
the help of the impinging cover gas jet, the heat ﬂux is directed to the bottom wall of the
key hole, thereby performing cutting action and increasing the depth of cut. The ability to
achieve deep cuts depends to a large extent on the optimization of the gas ﬂow parameters.
With regard to this it is essential to maintain an optimum jet “stand-off” distance from the
surface of the workpiece (DGF is effected by the stand—off distance). Capacitive feedback
surface followers to sense optimum distance from nozzle to workpiece and ensure precision
cutting of convoluted shapes, are becoming commonplace in the laser processing industry.
On the other hand, if the focal plane is moved much lower into the surface, then the cutting
depth decreases, since the re-focussed focal plane lies way inside the sample surface as in-
dicated in Figure 7.11. In Figure 7.11, both the optimal focal plane position and the effect
of having the focal plane positioned way below the surface of the sample are schematically
illustrated. In the present investigation, positioning the focal plane about 0.4 mm inside the
sample surface yielded the maximum depth of cut. The above mentioned phenomena are
inﬂuenced by a great number of parameters, some of which cannot be modiﬁed directly.
Many adjustable machining parameters inﬂuence the process performance through more or
less non-deterministic process parameters. The machining parameters are selected on the
basis of the workpiece thickness and according to the desired ﬁnal surface roughness and
HAZ width.

133

 

 

Optimal focal postion \/

workpiece

 

 

 

Figure 7.11 - Focal position relocation by density gradient ﬁeld and the original focal plane
location [229].

7 .2.5 Roughness of the Cut Surface

From the technological point of view it is not always reasonable to cut with the maximum
cutting rate because, the actual geometry of the component would differ from the pro-
grammed contour. Power of the laser beam, thickness of the workpiece, and tolerable HAZ
width limit the maximum attainable cutting speed and depth of cut. Also, it is well known
that near the cutting limit the roughness of the cut kerf increases signiﬁcantly, as outlined
in Figure 7.12. The average roughness values of the cut surface at different cutting speeds
are plotted in Figure 7.12. The upper and lower limit lines shown in the plot, are traced to
bound the data points within these limits. At all cutting speeds, the cut surface exhibited
higher average roughness at the bottom of the kerf than at the top section of the kerf. An
increase in cutting speed led to an increase in average roughness of the cut surfaces. This is
due to the smaller kerf width on the bottom side of the cut. The narrower kerf width on the
exit side of the cut, makes it difﬁcult to blow out and to exhaust material in the kerf. As a
result, a great deal of hardened and recast material accumulated in the kerf, and this led to a
higher average roughness of the cut surfaces. From the model presented in Section 7.2.3.1,
it is clear that at higher traverse speeds, a large amount of material goes into molten state

and also a signiﬁcant amount of it would be retained in molten state. This highly viscous

134

new
2960 NZ we $3 m; Mo 95305 a 3 36888 moﬁecnzzooo 5.2 8on wizzo .23 So 05 no mmoczwsoc mo 5:323 - as charm

A.-- EE: .5 88w 0.52:0
. . ._w..w.£p.v._oq.o

he

 

 

 

- eﬂmmcog
. 8 68:5

22:58.. /. _.- 8mm

o .On

     

E s n 3052;... 550on
canoe: At 8.0 .630
3 83 u 526d .83

 

 

 

8

l

I'.’

.§

—l

i/

O.

3
§ 8
mosmccmww

Ll_l

o:

 

 

 

OS

135

melt needs to be ejected out from the kerf before it has time to resolidify on the cut surface.
Also, the highly viscous molten material has to be exhausted through a narrow kerf at the
exit side of the cut, which makes it all the more difﬁcult, since laser material processing is
associated with very rapid solidiﬁcation and cooling rates. Hence, the narrow kerf width
at the exit side of the cut and the viscous nature of the large amount of molten material
formed at high traverse speeds result in increased average surface roughness.

In Figure 7.12, once the cutting limit is reached, a small increase in speed results in a
drastic increase in the roughness. This effect is also known to occur in steel and aluminum
cutting. At higher cutting rates, the assist gas pressure was not sufﬁcient to blow away
the dross collected on the cut surface, i. e., the ﬂow from the nozzle and the kerf are not in
sink with each other. The ﬂow from the nozzle must couple effectively with the recently
formed kerf to remove dross and enhance the cutting action. This might explain the rela-
tion between roughness and the cutting rate described earlier. Adaption of high pressure
machining techniques, with proper coupling of the gas jet from the nozzle might lead to
dress free cuts [231].

Figure 7.13 shows the cut surface of the specimen cut at 415 kPa N2 gas pressure and
a cutting speed of 2.0 m/ min. At this cutting speed, the surface was very rough (average
roughness was 2 pm) and it was straddled with striations throughout the cross-section of
the cut surface. For most beam processing methods, this is a common observation for
the. cut surface. Periodic striations observed on the cutting surfaces degrade the surface
quality (Figure 7.13). But quality, speed, and accuracy of the cut contour are the three
main factors that have to be taken into consideration for any machining process [21]. The
striations on the cutting surfaces are due to unsteady motion of the molten layer or melt ﬂow
oscillation. Intermittent plasma blockage could also play a part in the formation of surface
striations. It was suggested in literature that instabilities in the laser pumping mechanism
also play an important role in the creation of such structures. In LBM of composites, the

control of striations on the machined surfaces was very critical. Lower roughness values

136

 

Figure 7.13 - Photomicrograph showing rough surface cut with nitrogen cover gas at 2.0
m/‘mz’n and at 415 kPa pressure for an extruded 6061/A1303/20p composite.

137

can be achieved by aspherical lens, where the mode of machining was by ablation (see
Section 7.2.8). The stability of plasma, volume percentage, and distribution of alumina
particles play insigniﬁcant role in the characterization of machined surface ﬁnish.

So, textural parameters considerations should play an important role in deciding opti-
mal processing speed. Textural parameters of surfaces include roughness, waviness and
lay. When addressing the economics of cutting with laser beam followed by a subsequent
ﬁnishing operation, an optimum traverse speed may need to be determined. But, it may be
argued that a conventional ﬁnishing operation is more cost effective to employ on a rough
surface than lowering the traverse speed to produce a smoother ﬁnish. The waviness (stri-
ations), rather than roughness, is more important when laser beams are used to produce
parts which require subsequent ﬁnishing operation. Roughness of the cut surface would
play an important role, if subsequent ﬁnishing operation cost has to be curtailed. This
factor alone has to be taken into consideration for composite machining operations, as the
ﬁnishing operations are expensive and time consuming due to presence of brittle and hard
reinforcements in the composite. It would not be fair to say that waviness of the cut surface
is not an important factor for composites. Therefore it was reasonable to investigate the
inﬂuence of the cutting rate on roughness of the cut.

Hence an Optimal cutting rate has to be selected to minimize the macro- and micro-
scopical differences of the cut surface. High accuracy cutting is normally done at relatively
low speed, partly because of the high accuracy feed characteristic of the processing table
is operational only in the low speed range. Figure 7.14 shows the surfaces produced when
the workpiece is cut at the optimal cutting rate. In Figure 7.14, the affect of the cover gas
is also illustrated. Figure 7.14(a) shows, the sample cut at 0.508 m/mz‘n (optimal cutting
rate), with N2 as the cover gas and Figure 7.14(b) shows the sample cut with Oz as the
cover gas, under the same experimental conditions. As described earlier, cuts made with
N2 cover gas produced relatively smoother surfaces (average surface roughness was 0.98

pm) as compared to the other cover gases investigated in this report (the average surface

138

 

(b)

Figure 7.14 - Photomicrographs of as cut surface for a 6061/A1203/20p to show relative

V

Nitrogen cover gas at 415

uses. (21)

kPa and cut at 0.508 m/‘min and (b) oxygen cover gas at 415 kPa and cut

at 0.508 m/mm.

0
C

v ygen cover

roughness of nitrogen and OX

1

139

roughness of the cut shown in Figure 7.14(b) was 1.7 pm).

Figure 7.15 and Figure 7.16 are SEM micrographs of the cut surfaces shown in Fig-
ure 7.14(a) and Figure 7.14(b) respectively. In Figure 7.16, the smooth dark layer is the
recast layer. This was typical microstructure observed for all cuts made with 02. With 02
as the cover gas, the A12 03 particles are physically removed/ejected from the matrix, by the
intense heat developed (plasma cloud) in the cut zone. This clearly explains the reason for
waviness and wide kerf widths obtained with 02 assisted cuts. In contrast to Figure 7.15,
Figure 7.16 clearly depicts the secondary phenomenon of material removal (1'. e. physically
dis-lodging the particles) which is not observed when cuts are made with N 2 as the cover
gas.

There seems to be some oxidation taking place, even in cuts made with N2 (commercial
weld grade) as assist gas. This might be due to the evolution of nascent oxygen from
A1203 particulate decomposition. From Figure 7.15, however it can be interpreted that
the majority of the cutting action was by evaporation (ablation) of the material present in
the path of the beam, as their is less amount of re—solidiﬁed layer on the cut surface (see
Section 7.2.3.1).

Formation of re-solidiﬁed material along the groove walls poses a signiﬁcant disadvan-
tage; the re-solidiﬁed material reduces energy efﬁciency by absorbing beam energy which
would otherwise remove additional material. The recast layer, which also degrades the
surface quality can be minimized by using an off-axial gas jet to expel the molten material
from the cut zone. The cut edge of the beam-entry plane seems to be very smooth and
sharp. However, in the beam-exit plane, a re-solidiﬁed aluminum matrix material adheres
to the composite surface.

The results from microscopy studies indicate the existence of a critical energy density
for the cutting of the particles of alumina and matrix, Emu-t) and BMW.) respectively. The
critical energy densities, Emu-t) and E,,,(c,,-,), are the strong breakdown intensity threshold,

of both raw materials. When the laser-beam energy is over the critical value Ep(crit)a both

140

0:69:90
acm\nON_<:coo a 23 8:85:38 3sz 3 3V use 5:85:32: 32 2w 3 5X. 3 552.3 .29: .8528 2: a «one
853 2: use no.8— .282 $3 .55}: womd 3 :6 use new 550 we Smog: 5:3 605.3 :5 we he assumes? 2mm - 26 FEB...

 

141

ABE.“ 3 @2865 $6533 ccEScEwE:
kiwi 3 EV c5" ﬁx. 3 3339.: 8.323 3.8.? .3 cc. 8658? EB FO. 3 @286E :23. x33 Eb: 7.88
mi??? 9; 30% :8 2.2:}: wand 95 9% :58 mm 5ng 5:5 .ooatsm manO£<=ooc :8 ma .0 i=..ma.5:=925 - 9:. 2:3:

 

142

the particulates and the matrix are removed; however, when the laser-beam energy ranges
from Em(crit) to Epmﬂ), only the matrix is removed. The critical energy density of the
aluminum alloy is considered to be far lower than that of alumina particles.

Among the disadvantages of the laser, one can include the surface integrity factor of
the cut surface. By surface integrity we mean roughness of the cut, waviness, and ﬂaws
(are defects such as cracks, blow holes, scratches, and ridges occurring over a relatively
small portion of a surface area) on the cut surface. Surface integrity controls often result in
increased manufacturing costs and decreased production rates. Therefore, surface integrity
practices should not be implemented unless the need exists. Process parameters that pro-
vide surface integrity should be applied selectively to critical parts or to critical areas of

given parts to help minimize cost increases.

7 .2.6 Mechanism of Cutting

The mechanism by which the material and debris are removed from the cut zone are indi-
cated in Figure 7.17. Figure 7.17(a) and (b) depict the melt zone and solidiﬁed zone in the
vicinity of the kerf as the beam moves along the desired contour of the cut. Figure 7.17(c)
shows the removal of the molten material by the impinging assist gas jet.

The cutting mechanism can be classiﬁed into two main processes: ablation cutting
mechanism and fusion cutting (melting) mechanism [229]. For ablation cutting, incipient
surface melting in the interaction zone arrests heat losses by thermal conduction, thereby
causing instantaneous evaporation of the material in the cut zone. In order to observe this
phenomenon, very high power densities are required (z 108 W/cmz) and are usually avail-
able only in pulsed laser sources. Formation and sustaining of a key-hole takes place in this
kind of ablation cutting. Because of the use of high power densities, steep temperature gra-
dients are generated in the vicinity of the interaction zone, leading to high localized thermal

stresses. Moreover, extremely high surface stresses will arise from recoil pressure of the

143

 

 

laser beam laser beam
initial hole >
initial hole
*—
matrix solid
melt Joﬁdiﬁed zone

 

 

 

 

 

(a) (b)
__>cutting direction

erosion fr t

 

 

o o
0 0 melted material

a a
(C)

Figure 7.17 - Work material interaction with laser beam: (a) transverse section [232], (b)
longitudinal section [232], and (c) material removal mechanism for through
cutting [233].

144

evaporating surface atoms/ions. The high thermal conductivity of the 6061/A1203/ com-
posite does not lend itself to ablation cutting process, (because of the fair amount of heat
losses by thermal conduction) unless the process parameters are optimized as discussed
above. On the other hand, in fusion cutting process, after a small key-hole is formed, the
material is removed in the form of a molten product by the coaxially ﬂowing cover gas.
Fusion cutting process is more signiﬁcant for high thermal conductivity materials. Even-
though less power is required for fusion cutting, in general the cut surfaces are very rough,
because of the resolidiﬁed layers on the walls of the kerf as seen above.

7.2.7 Hear Affected Zone

Figures 7.18(a), (b), (c), and (d) are the microstructures of the 6061/A1203/20p polished
specimens. The particulate distribution in as received sample is illustrated in Figure 7.18(a).
Figures 7.18 indicates the presence of the heat affected zone in samples cut with N2 assist
gas at 0.508 m/mz'n cut speed. Even at this low processing speed the HAZ is only about
0.180 to 0.050 mm. It was also observed that at higher cover gas pressure (450 kPa), the
HAZ was almost negligible. At higher gas pressures, the cooling effect of the co—axial gas
jet helped in reducing the HAZ signiﬁcantly.

But as noted earlier (see Figure 7.3) high gas pressures led to wider kerf widths. The
transverse sections used for these SEM microscopical studies are shown in Figure 6.8. The
HAZ width shows a decreasing trend from top to bottom of the cut width, as schematically
illustrated in Figure 6.8(a). At the exit end of the cut there was almost no HAZ. Samples
cut at 2.0 m / min showed hardly any HAZ. The only major objection to usage of LBM
for machining of aerospace materials was the presence of HAZ and micro-cracks. But
as observed in Figure 7.15 and Figure 7.16, there was no observable cracks in the recast
and re-solidiﬁed layers. Presence of micro-cracks would lead to a reduction in fatigue life

of the components fabricated by LBM technique. Negligible HAZ width and absence of

145

 

 

 

 

 

 

 

 

 

 

 

Figure 7.18 - SEM micrographs of polished composite sample: (a) showing particulate
distribution in the as-received composite, (b) depicting Heat Affected Zone
(HAZ) at entrance side of the cut, (c) HAZ at mid—section of the cut, and (d)
HAZ at bottom—section of the cut. HAZ in (b) is large as compared to (c) and

(d)

146

micro-cracks would make this technique a feasible and economical way of carrying out
secondary processing of metal-matrix composite materials. As reported elsewhere in this
investigation, any machining process would require further ﬁnishing operations, to obtain
physically acceptable functional surfaces. The negligible HAZ width would be easily re-
moved, when the cut components are subjected to grinding, broaching, honing, or other
ﬁnishing operations required to get rid off the roughness of the cut surfaces.

The accurate machining of work pieces, especially in case of comers and small radii,
is another major problem for most of the other non—traditional machining methods. In the
case of LBM, cutting comers and small radii is limited by the inherent errors present in
any mechanical motion system. For these contours, the cutting speed would be lower than
for other types of geometries. Circular specimens with an accuracy of 0.001 mm were
successfully cut by LBM at 0.508 m / min in N2. Because of incomplete analysis of this

experiment, the results are not being reported here.

7 .2.8 Inﬂuence of Opties

In LBM of particulate MMC, the major problem was that of striations on the machined
surface, caused by melt ﬂow oscillation, intermittent plasma blockage, ejection/removal
of particles, and instabilities in laser pumping mechanism [234]. All these factors, except
for the instability introduced by the resonator cavity, can be controlled by selecting proper
laser processing parameters.

The results presented in the previous sections are for laser head equipped with 25.4 cm
FL reﬂective optics. From the discussion it is also clear that the type of coaxial cover gas
and pressure, position of the focal plane with respect the sample surface, cutting speed,
and amount of energy input have a strong bearing on the nature of the plasma control and
mechanism of cutting. VVrth reference to Figure 7.12 it is clear that the average roughness

of the laser machined surface, for a 7.0 mm thick composite was 1.1 to 2.0 urn. The

147

composite was machined at a speed of 0.508 m / min, with “reﬂective optics”ifocal plane
positioned no.4 mm inside the sample surface. Nitrogen, at a pressure of 415 kPa was
used as coaxial cover gas. The laser power was maintained at 2400 W, with a nozzle
stand-off distance of 0.5 mm.

SEM micrographs of the cut surface of a 9.2 mm thick sample (extruded sample) shown
in Figure 7.19 were produced under conditions identical to the one given above, except for
switching the optics to aspherical lens. The average roughness was z0.3 pm. In Fig-
ure 7.l9(a), the band passing across the micrograph (indicated by ‘X’), was the molten
metal that solidiﬁed on the cut surface. These resolidiﬁed bands (streaks) were caused by
unstable melt ﬂow. The high average roughness values reported for reﬂective optics were
due to excessive banding of the molten metal across the cut surface. Striations produced by
melt ﬂow instability can be avoided, if the cutting mechanism is by ablation rather than by
ejection of molten material from the kerf. The dominant feature of material removal from
the kerf by aspherical lens appears to be by ablation process. The negligible presence of
molten bands in composites machined by aspherical optics, justiﬁes this argument. This
result is consistent with the non-equilibrium conditions that exist during high energy laser
processing, limiting the melt motion. A narrow, uniform, and parallel kerf width of 0.3
mm was obtained. In the previous study, the kerf width had a slope of ’£0.90. The minor
amount of solidiﬁcation bands seen in Figure 7.19(a) were produced due to this reduced
kerf width. The HAZ for the composite cut with aspherical lens was negligible (rather no
perceptible HAZ was observed under optical microscope). Because of the instability of the

transmissive aspherical lens to high power lasers, the lens has extremely short life span.

148

Agog—Eco 95.68:. .8 2x2 .280: 2525.8 acm\n0~_<:ccc x he 02%.; 2m: 2: he :QEmEBE 35m - 3.5 233,.—

m. w, 3 a : _

 

149

7 .3 Laser Beam Welding

The primary experimental variables that determine the geometry of the weld bead, the weld
zone microstructure, and the HAZ are: laser beam quality (mode structure) and intensity
(decided by the incident beam power and spot diameter), traverse speed, type of optics,
initial surface condition of the material (dictates the absorptivity), volume percentage and
homogeneity of the particles in the composite, pressure, and ﬂow rate of the cover gas.
These laser processing parameters play major roles in the formation of plasma. Plasma
control is very essential for achieving welds of reproducible quality. The inﬂuence of laser
beam quality and intensity manifested in terms of single and dual laser beam is discussed
in this dissertation.

Since the alumina composite was produced through an ingot metallurgy route, it is
possible to fusion weld this composite as opposed to composites produced by powder met-
allurgy techniques, where generation of porosity due to entrapped gases would cause severe
degradation of mechanical properties (refer to Section 5.4).

The different process parameters, for example those of beam, cover gas, and volume
fraction and distribution of the reinforcement have an inﬂuential role on the resultant weld-
ing of aluminum composites. Due to the properties of aluminum, the threshold intensity of
producing a laser-induced plasma in connection with a key-hole is higher compared to for
example, steel. However, the system plasma capillary is necessary for an efﬁcient energy
coupling of the laser beam with the workpiece.

The control of the plasma is very essential for accomplishing welds of reproducible
quality. The laser beam parameters play an inﬂuential role in the formation of plasma.
Besides being inﬂuenced by the beam’s parameters, the effects of the generation and the
shielding of the plasma are strongly inﬂuenced by the process gases, the volume percentage
of the reinforcement particles, and the initial metallurgical structure of the material. It

will be established later in this report, that due to these factors and despite of the reduced

150

bandwidth of the process parameters, it is possible to reach autogenous welding results of

acceptable and reproducible quality for single beam LBW.

7.3.1 Single Beam LBW

The optical micrograph of a polished and unetched virgin 6061/A1203/10p and 6061/A1203
composite samples are shown in Figure 6.1 and 6.2 respectively. These micrographs
illustrate the inherent inhomogeneous distribution of the alumina particles in the ma-
trix. As mentioned earlier, Kawali et.al. [202] have reported that the presence of A1203
(6061/A1203/20p in their case) has played a crucial role in the breakdown of plasma. This
observation was in contrast to the one evidenced in the present investigation as explained

below.

7.3.1.1 Absorption of Laser Power

Sharply focussed laser beam generates high—intensity energy with low heat input and a
narrow HAZ. But accurate beam aiming is critical to avoid plasma breakdown over the
weldment. The consistency of absorbed intensity determines the uniformity of the weld
penetration. Fluctuations in average output laser power is one source of penetration varia-
tion. In highly reﬂective materials, like Al—based systems, varying absorption also results
in inconsistent weld penetration. In addition, for the A1203 particulate composite used in
this study, the inhomogeneous distribution (see Figures 6.1 and 6.2) of the particles also led
to variation of the absorption (inconsistent coupling) of the laser energy by the composite.
The interaction of the laser beam with A1203 particle caused a better coupling of the beam
with material. The absence of particles in- the path of beam traverse contributed to plasma
breakdown and as a result the weld bead was riddled with numerous pores and thermal
cracks as shown in Figure 7.20 for a 3.2 mm thick composite. The particle deﬁcit areas of

the composite had higher reﬂectivity to 002 laser, thereby creating a dense super-charged

151

 

Figure 7.20-Macrograph of bead—on—plate weld formed on a 3.2 mm thick
6061/A130gl’10p particulate composite. at a traverse speed of 0.8 mymr‘r‘
and heat input of 1200 H". (a) The fusion zone comprised of thermal cracks
(indicated by arrow). (b) non-uniform melt ﬂow, and (c) numerous pores
(indicated by 'X‘).

152

plasma, which in turn was responsible for the complete breakdown of the plasma. Studies
also indicate that for the 6061/A1203/particulate system, the amount of porosity in the ex-
truded virgin material increases with increasing reinforcement content, especially in areas
where particles are clustered [92]. Mechanical (joint geometry, weld ﬁxture constraint),
thermal (heat input and heat sinking), and metallurgical (alloy composition and volume

percentage of the reinforcement) factors play important roles in the formation of cracks.

7.3.1.2 Characterization of Weld Bead

To characterize the weldment the inﬂuence of the traverse speed and laser power input on

weld penetration and width of the weld bead have been studied in this investigation.

7.3.1.3 Eﬁect of Traverse Speed on Depth of Penetration

The dependence of the depth of penetration on traverse speed for SB-LBW is illustrated
in Figure 7.21 (also see Figure 2.5). The plots shown in Figure 7.21 are for as—received
composite samples (6061/A1203/10p) of ﬁxed thickness of 3.2 mm and laser power rang-
ing from 200 to 1000 W. A careful examination of Figure 7.21 indicates that at a given
traverse speed, the depth of penetration increases with power and decreases with traverse
speed. At a lower traverse speed, the amount of the laser power available for weld bead
formation is higher, hence the depth of penetration is large. An inadequate supply of laser
power in the weld zone, with increased traverse speed (as a result of reduced interaction
time), causes a sizable reduction in penetration depth. 0n the other hand, the traverse speed
and laser beam power has a diametrically different effect on the plasma cloud formation.
At low traverse speeds and high power levels, one would expect a signiﬁcant increase in
the depth of penetration, because of the increase in laser—material interaction time and high
energy input respectively. In Figure 7.21, for a ﬁxed power of 200 W, the depth of pene-
tration at 2.5 m/mz‘n was z0.3 mm and at 0.8 m/mz‘n it was z1.0 mm, i.e., an increase

of approximately three times, but for a power of 1000 W (ﬁve times increase in power

153

 

 

 

 

2.0
0—0 200W
H 400W
H 600 W
-l——l- 800W
<>——<> IOOOW
1.5 r L
E s
E
r:
.g
g
g 1.0 -
0
o.
H—a
o
'5
ca.
° F
D
0.5 -
0.0 r l m l l l n L L
0.5 1.0 1.5 2.0 2.5 3.0

Traverse speed of the laser beam (m/min)

Figure 7.21 - Plot of variation of depth of penetration with traverse speed, for various laser
powers. The thickness of the 6061/A1203/10p specimen was 3.2 mm (6 data
points were used to generate the range indicated by the error bars).

154

levels), the depth of penetration at 2.54 and 0.8 m / min was approximately 1.0 and 1.7
mm respectively; an increase of only 1.7 times. Both these conditions (low traverse speed
and high laser beam power) result in the formation of a dense plasma cloud, which atten-
uates the incident beam, reducing the depth of penetration. At high traverse speeds and
low power levels, the plasma formation was insigniﬁcant to affect the depth of penetration.
On an average the depth of penetration exhibited a decreasing trend with traverse speed,
for a given power. Depending on the thickness of the sample, there appears to be a crit-
ical traverse speed, below which it was difﬁcult to control the plasma breakdown, due to
excessive absorption of energy by the super-charged plasma. The lower bound of the tra—
verse speed (critical speed) can be ascertained by observing the nature of the plasma plume
above the weldment. The height and intensity of the plasma varies with the traverse speed
and heat transfer rates. At traverse speeds higher than the critical speeds, improper pene-
tration occurs, whereas at traverse speeds lower than the critical speed excessive melting

with signiﬁcant loss of material from the weld zone and weld perforation occurs.

7.3.1.4 Inﬂuence of Power on Depth of Penetration

In Figure 7.22, the variation of depth of penetration with the total heat-input for a ﬁxed
thickness as-received 6061/A1203/10p composite sample at a constant traverse speed is
shown. Figure 7 .22 complements the information provided in Figure 7 .21. In Figure 7.22,
the three plots were for 0.8 m/mz’n (30 ipm), 1.3 m/mz'n (50 z'pm), and 2.5 m/mz’n (100
z'pm) traverse speeds. These three traverse speeds were selected for further rrricrostructural
studies, as they provide the extreme conditions of the laser welding process. Sufﬁcient data
points could not be generated for the 0.8 m / min ( 30 ipm) plot in Figure 7.22, for powers
beyond 1000 W, as the specimen were being subjected to cutting action.

In addition to the inhomogeneous particle distribution (see Figures 6.1 and 6.2), a tra-
verse speed at or below the critical speed was another contributing factor to the severe

intermittent plasma breakdown. The lower ionization energy of aluminum leads to for-

155

.3on 8.8%: 2388 use GE: «.8 $05.35
33 «0 9: 103538 3382-8 no 203 839-5689 SH .8933: 55 scumbag—om he Egon mo 223:2, of. - «NH 2:»:—

333 £225 .595
0.88 odoom odomm 0.89 ado:

. J l—‘ [1

 

1

252 38 0.8m.
. . . . . _ . . . oo

 

 

md
04 M
m.
m.
m.
m w a
m
.m
QN m
- md
.E\E ad 4
.5): m... o
c E): 3d *

 

 

 

 

 

156

ASS. CC :izé:
wd mo Beam 3.8%: a 9:... \S 83 mo 5%: 838 Be .3383 E 3.86:: 53.20 5:55 35... 8.5a 5:95 2V as:
$5.5; 3222:; AS Eon .«o 8:82; 2: MEEQSE .23 2: me 8:03 9.83%: 2: E naﬁwoﬁi. 5:029 chcsom . mas 2:3...

 

157

mation of dense overheated plasma, even at lower intensities. This interruption in plasma
generated shock waves in the weld zone with resulting micro-cracks and porosity as indi-
cated in Figures 7.23 (a) and (b). This overheated plasma causes an increased absorption
of laser energy by the plasma region above the surface of the sample and hence leads to a
decreased energy coupling with the workpiece. Mth increasing welding speed the poros-
ity again decreased rapidly. This effect is already known from welding experiments of
thin metal sheets, where the gas evolution is supported by reduced welding depths. The
occurrence of porosities in the welding zone can also result from contaminates, chemical
reactions or boiling/evaporation of the molten material. Various surface cleaning proce-
dures were followed to mitigate the effect of surface contaminants as discussed elsewhere.
Since the laser welding process is characterized by a high aspect ratio (for deep penetration
welds), the gas evolution is limited by‘the viscosity of the molten material, the distance of
the blow-holes to the surface of the workpiece and the solidiﬁcation time.

In 6061/A1203/10p and 20p composites, the plasma formation is further complicated by
the low ﬂuidity (high viscosity) and low melting point of aluminum and the very high vapor
pressure of magnesium. Thus, there is a substantial need to control the plasma formation,
to eliminate material loss and weld porosity. Both the loss of constituent alloying element
(Mg alloying element in the A1 6061 alloy) and porosity led to weak weld joint. The pores
provide the potential sites for crack initiation and the magnesium loss reduces the effect of
solid solution hardening, which is reﬂected in the loss of hardness in the fusion zone. It is
a well known fact that as the magnesium content decreases, the aluminum alloy matrix will
be more prone to thermal cracking [155]. Porosity due to material loss as well as surface
oxidation are major obstacles encountered during SB-LBW of 6061/A1203 composite.

The high porosity content for the alumina composite near the critical weld speed can
be attributed to the surface reactivity of the matrix with ambient moisture and the vio-
lent breakdown of the plasma. Angular cavities observed in Figure 7.23(b) are usually

associated with higher gas content in the molten material. Small spheroidal porosity (Fig-

158

ure 7.23(a)) forms from a portion of the hydrogen that remained in the solid metal during
the primary fabrication. The gas in these small pores is at high pressure. With the increase
in power density, temperatures greater than the boiling point of the material are generated
at the surface of the weld pool (refer to Section 7.3.5). As a consequence, the vapor pres-
sure on weld pool is higher than the ambient pressure. This prevents the occluded gas in
the material to rise to the top surface and escape. .

The onset of cracks at the interface of the alumina particle and the parent metal can also
be observed in Figure 7.23(b). The smooth surface and the size of the alumina particle indi-
cates that there was excessive melting and secondary recrystallization of the matrix on the
alumina particles. The bottom of the weld bead also exhibited secondary recrystallization
of the alumina particles, with extensive grain growth (Figure 7.24), than the top portion of
the weldment. Below the critical traverse Speed, a cross-sectional view of the weld bead
exhibited clustering of alumina particles towards the bottom of the weld.

Figure 7.24 indicates the presence of recrystallized grains on the surface of alumina par-
ticles. The small faceted crystals in Figure 7 .24, on the A1203 particles are the MgA1204
spinels. It is a well established fact that A1203 particles are unstable in magnesium
containing aluminum alloys. Two possible reactions can occur in molten Al containing
Mg [90,104]:

4A1203 + 3M9 -—) 3MgA1204 + 2A1 (7.16)

A1203+3Mg —+3MgO+2Al

Eventhough the MgO reaction is thermodynamically favored in high Mg content 6061 Al
alloy, the interfacial energy for Al203/spinel is much lower; hence the spinel formation is
the most favored reaction [104]. In melt-processed composites like 6061/A1203/10p and
20p, the nucleation and growth process is controlled to limit the formation of spinel to a
thin, dense ﬁlm on the A1203 particle. This thin, dense reaction layer of spinel on A1203

. particle interface, will isolate the particle from the surrounding molten matrix during any

159

 

Figure 7.24 - SEM micrograph of the cross—section of the weld bead. indicating the bottom
half. A traverse speed of0.8 m/mm (30 rpm) and heat input of 1200 11' were
used. The recrystallized grains are indicated by arrows.

160

remelting and prevent excessive spinal reaction from occurring. In the joining process, the

formation of excessive spinel layers on A1203 particles is undesirable because:

0 The reaction indicated in Equation 7.16 removes Mg from the 6061 Al matrix. This
reduces the age hardening capability of the matrix.

0 This reaction further reduces the viscosity of the composite melt, which is highly

undesirable for autogenous welds.

0 Degradation of the particle/matrix interfacial strength.

The reduced cooling rates and rate of solidiﬁcation of the molten weld pool at the bottom
of the weld bead (Figure 7.24), favors the growth of the spinel ﬁlm on the Argo3 particles.
Whereas at the top side of the bead (Figure 7.23), the rapid solidiﬁcation and cooling rates.
should favor nucleation of thin, dense spinel ﬁlm on the particles. However, due to intense
beam power, there was substantial loss of Mg from the melt (high vapor pressure, porosity,
and prone to microcracking); this loss of Mg prevents the formation of protective thin spinel
ﬁlm on the particles (refer to Equation 7.16). To prevent excessive spinel formation (due to
reduced solidiﬁcation rates), Mg loss (due to intense heat input), super-heated dense plasma
cloud (due to Mg loss, high reﬂectivity of the base Al matrix), the SB-LBW process was
modiﬁed as discussed in Section 7.3.4.

Figures 7.25(a) and (b) indicate the presence of thermal micro-cracks in the weldment.
At a constant power, as the weld speed increases, the solidiﬁcation rate also increases. The
rapid solidiﬁcation rates associated with laser processing are extremely high at these high
traverse speeds, leading to solidiﬁcation shrinkage cracks. The shallow depth of penetration
explains the absence of porosity under these conditions. In Figure 7 .25 (a) one can observe
the shallow depth of penetration, as a result of the switch from deep penetration weld to
thermal conduction mode. As discussed earlier, a minimum power density is required in
order to form a keyhole. In the absence of keyhole formation (which is true at high traverse

- speeds), it is not possible to achieve deep penetration welds.

161

 

Figure 7.25-Weldment formed at 2.5 m/mz‘n (100 2pm) and laser power of 1700 ll".
(a)Macrograph of the transverse section indicating the shallow depth of pene-
tration and (b) SEM picture indicating micro-cracks on the top surface of the
weld bead. Thermal cracks are indicated by arrows.

162

In contrast to Figures 7.25 (a) and (b), Figures 7.26, 7 .27, and 7.28 does not indicate
presence of micro-cracks in the weldment. The micrographs in Figures 7.26, 7.27, and
7.28 are for the optimal weld speed of 1.3 m/mz‘n (50 rpm) at 1700 W laser power for
6.4 mm thick specimen. Despite of the absence of porosity and micro—cracking, the depth
of penetration for these optimal parameters was only #2.] mm. To increase the depth of
penetration and at the same time contain porosity and micro-cracking, one need to reduce
the high solidﬁcation and cooling rates of the molten weldment. This can be achieved by
CD-LBW process explained in later sections of this report.

The macrostructure of the t0p view of the weld bead face is indicated in Figure 7 .26.
In the SEM micrograph (Figure 7.27) no visible thermal cracks can be located, eventhough
it is known that thickness of the weld specimens effect the crack geometry. So allowance
has to be made when comparing weld data of the thick samples with the thin samples
used in this investigation. Figure 728 shows the typical lamellar like structure observed in
the transverse section of the weldment. The lamellar like structure was very ﬁne near the
surface of the weld and it gets coarser as one goes to the base of the weld bead in transverse
section. This lamellar structure can be explained by considering the solidiﬁcation rates of
the melt pool.

The solidiﬁcation rates of the weld pool is an important variable in providing secondary
recrystallized grains and lamellar structure in aluminum composites. It is evident that ﬁne
lamellar structures are present near the surface of the weld (Figure 7.28). As the distance
from the surface into the bulk of the weldment increases, the rrricrostructures, namely the
lamellar structure become larger. The transition from ﬁne to coarse structures may be ex-
plained on the basis of solidiﬁcation mechanics. The solidiﬁcation growth rate (R) and
temperature gradient (G) are the two parameters along with particulate concentration that
determine the type and ﬁneness of structures in the weld. The solidiﬁcation conditions
which exist in the weld pool can be characterized by the G/R ratio (thermal gradient divided
' by the growth rate of the solidiﬁcation front). The ﬁne cellular structure at the surface of

163

 

Figure 7.26 - Macrostructure of the 6061/A1203/10p composite, showing the top view of
the weld bead formed at 1700 W heat inpUt and 1.3 m/ min (50 ipm) traverse
speed.

 

Figure 7.27 - SEM micrograph of the top surface of the weldment shown in Figure 7.26

165

 

Figure 7.28 - A lamellar like structure observed in the transverse section of the weldment
shown in Figure 7.26.

166

the weld interface is due to the large G/R ratio. The amount of under-cooling present en-
hances the ability of the alumina particles to heterogeneously nucleate and grow new spinel
grains in two ways. At constant heat inputs and increasing welding speeds, which corre-
spond to lower G/R ratios (higher under-cooling), the exposure time for the particles at the
bottom of the weldment to the weld pool environment is reduced and also the elemental
Mg loss is very small (hence less plasma). The heat input in the lower half of the weldment
goes into melting the matrix. The minimal loss of Mg, along with molten Al in the bottom
of the weldment favor Equation 7.16 to proceed rapidly. The excessive spinal formation
with the reduced G/R ratio, result in a very coarse lamellar structure in the bottom half
of the weldment. It is tempting to use equilibrium phase diagrams to predict the various
features observed in the weld zone. But, in LBW, we observe non—equilibrium microstruc-
tural components reﬂecting the high cooling rates, hence it is not possible to predict the

microstructural features using ternary equilibrium phase diagrams.

7 .3.2 Effect of Optics

The effect of reﬂective optics on the weldment was also studied. All the previous results
were for the aspherical optics. Despite the difference in the focal lengths of these two
types of optics, the microstructures in terms of particulate distribution in the weld zone
was very different. Figure 7 .29 shows the optical micrograph of the transverse weld cross-
section, with depleted alumina particulates along the path of the weld. This micrograph
was obtained with reﬂective optics. The requirement on joints is for minimal disruption of
the reinforcements and little interaction between the reinforcement and matrix. The low
interaction time of the beam with the sample and rapid solidiﬁcation rates would inhibit
any undesirable chemical reaction between the reinforcement and the matrix. A signiﬁcant
loss of A1203 particles in the weld path of the beam entrance side was observed. At the

exit side of the beam,.the loss was not very considerable. The particle-free zone was wider

167

 

Figure 7.29 - Photomicrograph (at higher magniﬁcation) showing a region depleted of alu-
mina particles in the weld path. Off—axis argon gas at 345 kPa and weld
speed of 1.3 m/min.

168

at the entrance side than at the exit side of the beam. A1203 particles, on interacting with
laser beam, disassociate, generating large quantities of heat (Equation 7.16). The excessive
heat so generated leads to further decomposition of neighboring alumina particles. These
reactions along with Mg loss contribute to the build up of large amount of plasma plume
above the key hole. This dense plasma plume absorbs substantial amount of energy. Hence
towards the exit end of the beam, negligible depletion of alumina particles was observed.
This might also be the explanation for pore-free weld zone towards the exit end. Even
with “cross-ﬂow” cover gas at 345 kPa, the plasma was not completely suppressed.

The presence of dense plasma plume was closely related to the porosity formation [235]
as shown in Figure 7.30 for a 6.5 mm thick 6061/A1203/10p composite. With cross-ﬂow
cover gas, the porosity was reduced and deep below the surface, the cross-section was
pore free. The off-axis gas ﬂow was helpful in suppressing the plasma breakdown to a
large extent. The large number of pores observed at the entrance side of the beam may be
largely due to the presence of porous aluminum oxide layer on the surface of the composite. -
The melting temperature of aluminum oxide is 2049 °C, which is more than three times the
melting temperature of pure aluminum. Formation of porous oxide layer on the surface
of the composite tends to trap moisture and other surface contaminants that lead to weld
porosity. Weld porosity in aluminum is principally caused by bubbles of hydrogen that
form in the solidifying weld pool [236]. The surface porosity of the oxide layer can lead
to entrapment of water molecules on the surface of the composite. The hydrogen solubility
of aluminum increases almost twenty times at the solid-liquid transition temperature, and
continues to increase as the temperature increases. Hydrogen contamination usually comes
from the oil and water. The micrographs in Figure 7.29 and Figure 7.30 were for bead on
plate welds. It was known that many aluminum alloys exhibit cracks and hot-tear defect
when fusion welded [237]. The inability of the melt pool to support the strain imposed by
the solidiﬁcation shrinkage causes hot-tearing. With reﬂective optics, it was very difﬁcult
' to control the plasma. Hence it became necessary to use a high cross-ﬂow shielding gas

pressure of 345 kPa.

169

   

h... v. New

at.
«a.

...r.,....r..y

Figure 7.30 — Photomicrograph of alumina composite showing presence of porosity at the

top side of the weld zone (off-axis argon gas at 345 kPa and weld speed of

1.3 Try/min).

170

As a comparison of the welding process, experiments were also carried out on SiC
whisker reinforced composites. Neither aspherical nor reﬂective optics could help to con-
trol the formation of excessive porosity in this composite. Figure 7.31 shows the micro—
graph of butt-welded SiC whisker reinforced composite, with reﬂective optics. This com-
posite exhibited a very high degree of porosity levels and shallow penetration depth. To
address this observed phenomenon, further study has to be carried out. The high porosity
levels could be due to the generation of carbon monoxide. This composite was processed
through powder metallurgy route, leading to possible entrapment of moisture. It is known
that SiC shows sublimation at high temperatures.

SiC+02——>Sz'+CO

The irregular nature of the pores also agree with this explanation. In contrast, in Fig-

ure 7.30, the pores are spherical in shape.

7.3.3 Role of Particle Distribution

As mentioned above (Section 7.3.1.1), the as-received material exhibited inhomogeneous
distribution of the A1203 particles in the matrix (compare Figures 6.1 and 6.2 with Fig-
ures 6.3). The clustered nature of the A1203 particles manifest in the variable absorption of
laser energy by the composite, leading to numerous pores and thermal cracks. The interac-
tion of the laser beam with A1203 particles caused a better coupling of the beam with the
composite, whereas particle deﬁcit areas exhibited inferior coupling with the laser beam,
due to the high reﬂectivity of alurrrinum for C02 laser radiation. The high reﬂectivity, com-
bined with the low ionization energy of aluminum, lead to the formation of overheated and
super—charged (dense) plasma envelope over the weldment. The existence of this unstable
and stagnant plasma envelope leads to intermittent disruption of the weld bead. The role
of the plasma envelope is contradictory in the sense that a stable plasma cloud is needed to

' protect the weldment, whereas an unstable plasma causes porosity and rrricro—crack forma-

171

 

Figure 7.31 - Photomicrograph of SiC whisker reinforced composite showing presence of
porosity through out the weld zone (off—axis argon gas at 345 kPa and weld
speed of 1.3 m/nnn).

172

tion in the weld zone. As a passing remark, it is important to note that this unstable plasma
can be gainfully employed in the machining of this composite as explained elsewhere, but

plasma control is very essential for achieving welds of reproducible quality.

7.3.3.1 Effect of Hot-Rolled Composite

Various process parameters were optimized to inhibit the formation of unstable and stag-
nant plasma cloud over the weldment. Of all the process parameters, that affect plasma
stability, the distribution and volume percentage of the A1203 particles are the two most
important factors. Homogeneous distribution of the particles was achieved by hot-rolling
composites sheets (40 to 60 percentage reduction). Bead-on-plate welds were produced on
6061/A1203/10p and 6061/A1203/20p hot-rolled composite material to study the effect of
the above mentioned factors. In view of the superior results obtained with aspherical lens,
all the remaining single beam weld experiments were carried out with laser head equipped
- with aspherical lens. For any given unfocused laser beam diameter, the aspherical lens has
a smaller spot size than the transmissive or reﬂective optics (see Figure 6.6). Hence the

intensity of the laser power at the sample surface was very large.

7.3.3.2 Depth of Penetration in As-Received and Hot-Rolled Composites

Some of the signiﬁcant results of 6061/A1203/10p and 6061/A1203/20p composites are
reported here. Figure 7.32 illustrates the relationship between depth of penetration and
laser power, at various traverse speeds, for as-received and hot-rolled composite (see Fig-
ure 7.22).

The known inverse relationship between the depth of penetration and traverse speed, at
a constant laser power, can be conﬁrmed from Figure 7.32. The difference in the depth of
penetration at higher traverse speed (2.5 m/mz‘n), for as-received and hot-rolled composite,
was not very signiﬁcant, since plasma control was not that critical at these higher processing

speeds. At lower traverse speed (1.3 m / min), the difference in depth of penetration was

173

 

  

 

 

3.0
2.5 -
L 9K
2.0 ~
E
E
a“
.2
3
a
g 1.5 ' 9K
8.
'8
5
G.
a 1_o _ 02.5 m/min, as received
x 2.5 rn/min, hot rolled
I 1.3 m/min, as received
9461.3 m/min, hot rolled
A 0.8 m/min, as received
0.5 —

 

0.0 n r r l r r ;1 l r r r l 1 r r L r r r l L
0.0 400.0 800.0 1200.0 1600.0 2000.0 2400.0 2800.0
Laser Power, Watts

Figure 7 .32 - The variation of depth of penetration with heat-input, for bead-on-plate weld
of a ﬁxed thickness sample (2.4 mm) of 6061/A1203/10p as received and hot

rolled composite.

174

signiﬁcant enough to warrant an in-depth analysis. This difference can be attributed to the
distribution of the particles. The hot-rolled composite had a more uniform, cluster (band)
free particulate distribution, thereby providing better beam coupling with the material and
result in a stable, continuous plasma shield over the weldment (Figures 6.1, 6.2, and 6.3).

For the as-received composite, in addition to the effect of the inhomogeneous particle
distribution, traverse speed at or below the critical speed, was another contributing factor
to the severe intermittent plasma breakdown. This interruption in plasma, generated shock
waves in the weld zone which gave rise to micro—cracks and porosity. The overheated
and super-charged plasma prevents the incoming energy from reaching the surface of the
sample, and leads to a decreased energy coupling with the workpiece. In Figure 7.33,
'the depth of penetration results for a 6061/A1203/20p hot-rolled composite, along with
the data presented in Figure 7 .32 for 6061/A1203/10p hot-rolled composite, are plotted for
comparison.

Figure 7.34 and Figure 7.35 show optical micrographs of the transverse section of the
etched weldment in hot-rolled 2.4 mm thick 6061/A1203/10p composite at a traverse speed
of 1.3 and 2.5 m/mz'n respectively with a laser input power of 400 W.

In Figure 7.34, three zones corresponding to fusion (molten) zone, HAZ, and the base
matrix are visible. The fusion zone corresponds to about 150—175 pm, the HAZ (charac-
terized by the ﬂow lines in the micrograph) measures approximately 350 pm. Because of
the rapid solidiﬁcation nature of the LBW process, the heating and quenching cycles are
moderate at a traverse speed of 1.3 m / min and as a consequence the fusion zone exhibits
very ﬁne, equiaxied grain structure. In the HAZ, the thermal diffusion was slow enough
to allow limited solid state diffusion, thereby displaying partially recrystallized elongated
grains. The severely distorted grain structure produced during hot-rolling are still retained
in the HAZ.

In Figure 7.35, the fusion zone and HAZ of the composite, processed at 2.5 m / min are
z25 pm and @4255 pm respectively. The low heat-input at these high processing speeds

175

 

     

 

 

2.5
x
2.0 —
E
. 1.5 ~
r:
8
E
8
a:
“5 O 20%(p) and 2.5 m/min
5 l 0 _ x 10%(p) and 2.5 m/min
g ' ,K 720%(p) and 1.3 m/min
*10%(p) and 1.3 m/min
x
0.5 — hot-rolled samples
thickness = 2.4 mm

 

0.0 l 1 l 1
100.0 500.0 900.0 1300.0 1700.0 2100.0 2500.0

Laser power, watts

Figure 7.33 - Comparison of depth of penetration with heat-input, for bead-on-plate weld of
a ﬁxed thickness sample (2.4 mm) of 6061/A1203/10p and 6061/A1203/20p

hot rolled composites.

176

 

Figure 7.34-Optical micrograph of the transverse-section of hot-rolled 6061/A1203/10p
composite. processed at traverse speed of 1.3 m/mz'n. The fusion—. heat-
affected—, and base-matrix zones can be observed.

177

 

Figure 7.35 - Optical micrograph of the transverse-section of a hot-rolled 6061/A1203/10p
composite. processed at a traverse speed of 2.5 m/mm. Note that. in com-
parison with Figure 7.34. the fusion zone was smaller in size.

178

limits the size of the fusion zone, as a result shallow depth of penetration was produced.
The size of the HAZ was also limited, due to insufﬁcient thermal diffusion (low heat input).
This banded nature in the matrix was observed for all processing conditions in the hot-rolled
samples.

In contrast to LBM, LBW process was drastically affected by the unstable and stagnant
plasma envelope. The volume percentage and distribution of the alumina particles play sig-
niﬁcant roles in the production of reproducible quality welds. Hot-rolling of the extruded
composite samples had inﬂuential role on the quality of the weld, determined in terms of
depth of weld, porosity, width of the weld bead, micro-cracking etc. The processing speed
of the weld was also one of the primary variable in the LBW of MMC. Welding of MMC
below or above the Critical speed generated weldments with high porosity and microcrack-
ing. The critical speed was primarily controlled by the thickness of the composite material.

7.3.3.3 Width of the Weld nugget in As-Received and Hot-Rolled Composite

Figure 7.36 depicts the relationship between the width of the weld bead and depth of pene-
tration for a constant power of 1700 W and sample thickness of 3.5 mm. The width of the
weld bead varies in a parabolic fashion with the traverse speed. Based on the capability of
control on the plasma breakdown, porosity levels, thermal cracks, and particulate segrega-
tion in the weld zone, it was determined that for a laser power of 1700 W at 0.8 m / min (30
im) traverse speed, optimum microstructural features can be obtained for a 3.5 mm thick
composite. From Figure 7.36, it can be interpreted that the aspect ratio of the weld seam is
high at around 1.5 m / min (60 im) traverse speed. The welding mode switched to thermal
conduction mode at an approximate traverse speed of 0.8 m / min (30 z'pm) and 2.5 m / min
(100 11pm). In between these two weld speeds, the SB-LBW process exhibited deep pene-
tration weld mode. The transition points of 0.8 m/min (30 z’pm) and 2.5 m / min (100 im)
traverse speeds, for the mode change, were utilized to understand the SB-LBW process for

this composite. At low traverse speeds, the amount of energy absorbed (with the help of

179

 

 

4.0 4.0
0 width of the weld nugget
3.5 r *6 depth of penetration ‘ 35
9k-
power = 1700 watts
3 0 r - 3 0
U
E 8
'6 5'"
o
to H.
B a
g 2 5 — 2.5 a
“5 .3.
g .3
a E
2.0 r 2.0
1.5 L 1.5
1.0 ‘ ’ ' ‘ 1.0

 

 

 

0.5 0.8 1.0 1.2 1.5 1.8 2.0 2.2 2.5
Traverse speed of the laser beam (m/min)

Figure 7.36 - The dependence of width of the weld bead and depth of penetration on tra-
verse speed for a 3.5 mm thick sample of 6061/A1203/20p composite.

180

the super-charged plasma) by the material is much higher than at higher traverse speeds.
This differential energy absorption generates varying amount of molten metal in the weld
pool. The large volume of the molten metal present at low traverse speeds, leads to higher
width of the weld bead than at higher traverse speeds, as depicted in Figure 7.36. At lower
traverse speeds, the plasma was much more stable than at higher traverse speeds, leading to
pore-free weldments. Microﬁssures are far more dangerous than spherical, blunted pores,
since these tiny ﬁssures can cause catastrophic failure of the component.

Figure 7.25(a) and (b) indicate the presence of thermal micro—cracks in the weldment
produced at the optimal processing parameters of 1700 W and 2.5 m/mz’n. The rapid
solidiﬁcation rates associated with laser processing are enhanced by the presence of large
melt pool, super-charged and stable plasma envelope over the molten metal, presence of
solid phase (at much lower temperature than the melt pool) at the root of the weld bead
etc., resulting in solidiﬁcation shrinkage cracks. The shallow depth of penetration explains
the absence of porosity under these conditions. The shallow depth of penetration, observed
in Figure 7.25(a) is the result of the switch ﬁom thermal conduction to deep penetration
weld mode. It is well known that a critical power density is required to form a keyhole. In
the absence of keyhole formation (which is true at high traverse speeds and lower powers),

it is not possible to achieve deep penetration welds.

7 .3.4 Dual Laser Beam Welding

In view of these results for SB-LBW, it was proposed to carry out experiments using the
novel approach of CD-LBW set-up depicted in Figure 6.7. The major advantages of having
two laser beams interacting from either side of the composite sample, is to avoid the pres-
ence of solid phase at the root of the weldment that leads to micro-ﬁssures in the weldment.
It also results in less HAZ width, through penetration, absence of porosity, less visible dis-

tortion, reduced cooling and solidiﬁcation rates etc.

181

In Figure 7.37, a transverse section of the weld produced by CD-LBW is shown. The
2.5 kW laser power was split into two beams of 60% (1500 W, referred as beam or side A)
and 40% (1000 W, referred as beam or side B) of the power. The tap half of Figure 7.37
corresponds to the beam A arid the bottom half to beam B. On side B of the weldment,
porosity of diameter z 30 to 100 ,um can be seen and top half (side A) of the weldment
exhibits a signiﬁcant loss of A1203 particles in the fusion and HAZ. The laser beam focal
point was positioned about 2.0 mm inside the sample on side A and on side B, the focal
point was on the surface of the sample. Under these process parameters, the sub-surface
porosity being driven from side A did not have enough time to escape out from side B, since
the processing speed was quite high (1.3 m/mz’n.) and only a small depth of the bottom
face was in molten state, which solidiﬁed before the porosity was completely removed. The
depletion of alumina particles in the interaction zone of side A will be explained later.

The fusion zone, HAZ, and unaffected parent matrix were characterized by the grain
size reﬁnement, change in hardness, and presence of incoherent precipitates. A typical
optical micrograph of a polished and etched transverse section of the weldment exhibiting
the HAZ and the parent matrix is shown in Figure 7.38. The HAZ is characterized by
the presence of columnar grains and in the interface between the HAZ and the unaffected
parent matrix, the grains are more spherical in shape. Beyond this interface, the grains are
very ﬁne requiring larger etching times to be revealed. “Within the HAZ area, the small gray
phase distributed through out the microstructure is the precipitate of Mggsi (conﬁrmed by
EDS) that has coarsened due to the large temperature gradients. Since the heat input is
being provided from both sides of the composite sample, the root of the weld bead formed
on side A is almost at the same temperature as the molten pool present in the fusion zone
and vice versa for the root of the weldment formed on side B. Because of this phenomenon
the temperature gradients generated in the weldment are not that restrictive to cause micro-
ﬁssuring. The low temperature gradients also are responsible for the reduced solidiﬁcation

and cooling rates in the fusion and HAZ respectively, resulting in columnar grains in the

182

us». .
,....._ao.¢c ”1..

H. s x. r. . I
c w as .. 1.‘
.. . . U ..:
. _ . .‘i S H but
.2, . . ..
5}». ...N. ... a:
... a .. 4....
a a xx .1 to .
.5“ x. .. r. .. CD.
..v<...}.~ . D. 0; Who
1%.. 2‘. to «1 a
v \ an. . It 4
I .. . . . . x a i

.5. 1.1.0.? .1. 4M
x . 1.
w -.v.. ..._M.V,o—,...-ne— 1‘ .
« 4:3..- an... a... 2.. 9.1
(1...? ....:.. s 7...:
‘u. s v“. Q . .4
.L . . .5. o awe

      

. . e s s
(a. t... 3... . .4, . aw...
. . A ... iv: .4: a... “0&4. . m... 4%..
ta .: . ).u\. .s
. .. p a . . 5.... 2.... .
’. A . - l a .
.¥ .'ﬂ~. are—z.’3“ s». 7. .7 . \p.
pi .aﬂ ...r. i .1 n. ’. $.00
a“ ‘ﬁl\y\ . _. . A" I. ‘.
. b was .. a. .7 c c i.

 
  

s . .. ... .
Q u . p
I .. a .2 .k-«T o. 5.x... ent; y. .t 01.1 4 . .
.n.v.¢ .mz.\.w.«...‘ 91,... ”sum Hie. 3......1 25.. r on... ., a»? it”

.5 mm (particle deﬁcit area is in-

dicated by ‘X’ and pores are indicated by arrow ).

3

q
.z

Og/lOp composite of thickness

al micrograph of the transverse section of the polished and unetched

weldment produced by CD—LBW at a traverse speed of 1.3 m/mz’n in

6061/Ala

7.37 - Optic

Figure

183

.39 .3 3853
anE 8259 380:2: 2: 28 N<= 2: e338: 08%an 23 PO. 3 88:03 N<= mascaoc €83.38 Bongo - was 2:»:—

 

184

HAZ and spherical grains in the interface between HAZ and the parent matrix as evidenced
in Figure 7.38.

SEM micrographs of the HAZ and the parent matrix are given in Figure 7.39(a) and (b).
In Figure 7 .39(a) both the grain boundaries and the precipitates are clearly delineated. In
Figure 7.39, the precipitates of Mgzsi are very ﬁne to be visible at the magniﬁcation shown
in the ﬁgure. These micrographs are typical of all weldments produced in CD-LBW.

In contrast to SB-LBW, in CD-LBW (even at a process speed of 1.3 m/min) the
absence of sub-surface porosity and micro-ﬁssures, make this CD-LBW process more
amenable for full-penetration welds at high processing speeds. But the advantage of high
process speed is more than compensated by the sizable loss or depletion of A1203 particles
in the fusion zone of side A. A slight reduction in the traverse speed (from 1.3 m/mz’n
to 1.0 m / min) resulted in less depletion of particles in the fusion zone as shown in Fig-
ure 7.40(a).

With further decrease in traverse speed to 0.8 m/mz’n, there is not only a signiﬁcant
drop in the particle depletion zone width (see Figure 7.40(b)), but also the fusion zone on
side B did not retain any surface porosity (see Figure 7 .40(c)). In all the micrographs the
weldment produced on side B, did not display any particle depletion zone. On side B, the
input laser power is much lower than on side A (1000 Vs. 1500 W), hence the laser power
on side B was utilized more efﬁciently both for forming molten pool and in heating the
particles. This efﬁciency in laser power absorption on side B is due to the larger heat input
from side A of the weldment. In Figure 7.40(a), the amount of precipitates is much larger
than in Figure 7.40(b). Again this can be explained in terms of the process Speed, which
indirectly affects the heat input into the matrix and the alumina particles. Since alumina
particles have much lower thermal conductivity than the base aluminum matrix, there is
not sufﬁcient time for the heat energy to be transfered to the particles by conduction at
high traverse speeds. Hence all the energy that is absorbed is utilized in melting the matrix,

thereby forming large zones of precipitates. At these high temperatures, the molten pool

185

£229:— eEEEe 2: 8:665 .x. 88:38.5 035$ 38 bo> £3» .anE :55:— eoaootaca 2: 3V 95 A323 .3 .65:
-03 5.me me 33:90er Ea wetness: 58w 5:» .N<m A3 ”525% 82:22: 8.55%: 3:89 me .38?qu Sim - ans 932..

 

186

 

Figure 7.40 - Optical micrographs of (a) fusion zone in a 6061/A1303/20p composite with
a process speed of 1.0 m/mm. (b) fusion zone in the same composite at a
process speed of0.8 m/min (Contd...).

187

 

Figure 7.40 — {’Contd...) (c) fusion zone and HAZ on side B of the same composite pro-
cessed at 0.8 mfnzm.

188

of the matrix is less viscous. The increased ﬂuidity of the matrix alloy helps the alumina
particles to agglomerate in areas far removed from the fusion zone. This explains the
sizable large area of particle-deﬁcit fusion zone.

Based on these observations of sub-surface porosity, surface porosity, micro-ﬁssures,
size of the particle deﬁcit area, hardness of the fusion zone and HAZ, the optimal process
speed of weldment for CD-LBW was found to be around 0.8 m / min for a 3.5 mm thick
6061/A1203/20p composite sample. Optical microstructures of the transverse section of
the weldment processed at these optimal conditions is shown in Figure 7.41. For this
case, the focal point of the side A was about 2.0 mm inside the sample and the focal
point of the side B was about 0.5 mm inside the sample. Figure 7.41(a) indicates that the
penetration is complete and the weld obtained is a conduction limited weld (indicated by

the hemispherical shape of the bead).

7.3.4.1 Microhardness of the Weldment

The microhardness values also point out that 0.8 m / min is the optimal process speed for
a 3.5 mm thick 6061/A1203/20p composite sample. Figure 7.42 indicates the variation of
microhardness with the process speed for the 6061/A1203/20p composite. The microhard-
ness values in fusion on both sides of the composite exhibit identical trends. The micro-
hardness values increase rapidly with process speed till 0.8 m / min, thereafter the increase
is very slow. This increase in hardness values of the fusion zone correspond to a increasing
trend of the hardness values in the HAZ upto the optimal process Speed. As the traverse
speed increases, the microhardness values in the HAZ recover back to almost the parent
matrix hardness values. This clearly indicates that at low traverse speeds (03,04, and
0.5 m / min), the strengthening precipitates in the 6061 base matrix dissolve into the solid
solution and upon cooling, because of extensive loss of the major alloying element Mg,
forms non-strengthening, incoherent precipitates. Moreover the minor amount of strength-

ening precipitates that may form, tend to grow in size (overaging), thereby reducing the

189

 

Figure 7.41 - Optical micrographs of the polished and etched transverse section of the weld
bead (6061/A1203/20p) (a) the cross-section of the weld bead at lower mag-
niﬁcation showing the proﬁle of both the side A and B weld beads (Contd...).

190

 

Figure 7.41 -(Contd...) (b) micrograph of the fusion zone of side A. and (c) HAZ of the
weld bead formed on side A.

191

 

 

 

 

 

 

 

140.0
0 fusion zone of side A (60%)
9K fusion zone of side B (40%)
O HAZ
V parent/base matrix
120.0 -
V

a v

8

E“ v

:1: v 'z/I’V

5 100.0 — v

E:

a

.r:

8

.2

2

80.0 L
o
60.0 L l r l .l l 1 l g
0.2 0.4 0.6 0.8 1.0 1.2

Traverse speed of the laser beam, m/min

Figure 7.42 - The variation of microhardness (Vickers’) with process speed in the fusion
zone of side A and B, HAZ, and the parent matrix (6061/A1203/20p).

192

hardness. At high traverse speeds of 0.8 and 1.0 m/mz’n, the dissolved precipitates, upon
cooling may precipitate out as strengthening phases; hence resulting in enhanced hardness
values. At the optimal speed of 0.8 m / min, the microhardness values of the fusion and
HAZ are very close. This also indicates that the fusion and HAZ are able to sustain any
residual thermal stresses that arise during cooling of the weldment. Hence no microﬁs-
sures and porosity were observed in Figure 7.41. In general, high energy density processes,
like LBW and electron beam welding, result in relatively higher hardness values in the fu-
sion/molten zone when compared to HAZ and base metal hardness values [225]. The fast
solidiﬁcation and cooling rates associated with SB-LBW, result in high hardness values in
all regimes of the joining process when compared to the CD-LBW process. For this case,
the ﬁne precipitates lack sufﬁcient time to go into growth stage. Also the high rates of so-
lidiﬁcation and cooling prevent the weld pool to accommodate the high thermal shrinkage
strains, that are necessary to survive the solidiﬁcation experience without cracking. In the
case of CD-LBW, the preheat beam reduces the cooling rates of the fusion zone, resulting

in reduced hardness values in the fusion zone and negligible micro-cracking.

7 3.4.2 Size of Fusion Zone and HAZ

The most commonly used method for determining the width of the HAZ is by measuring
the hardness across this zone. For this composite, with 6061 Al base matrix, the HAZ can
be distinguished by nature of growth of the precipitates (for other heat—treatable aluminum
alloys dissolution of the precipitate may take place, e. g. AA2219 alloy) [155]. The 6061
aluminum alloy is a precipitation hardenable alloy. The variation of microhardness across
the weldment for the dual laser weld on the primary beam side is shown in Figure 7.43.
From these hardness values, it was determined that on the primary beam side (side A), the

fusion zone and HAZ was approximately 2.9 mm and 1.6 mm respectively.

193

 

110.0

105.0 ‘ Speed = 0.8 m/min , Thickness = 3.5 mm

100.0 . g 5 E

95.0- .

:0 co

5 fusion zone

90.0 - 9 o ,
5 o

35.0 - o

Hardness, Hv (100 g)

80.0 ~ ,
75.0 . §—----—--§ - ...... ..
70.0 . o,

65.0 - t---__--____--§

 

 

 

60.0 . l l . i . l . l . ' . 5
0.0 1.0 2.0 3.0 4.0 5.0 6.0 7.0 8.0

Distance (mm)

Figure 7.43 - The variation of microhardness with distance in a hot-rolled 6061/A1203/20p
composite.

194

7 .3.5 Results of Finite Element Analysis

In Section 6.4, the ﬁnite element model of the collinear dual laser beam (including pre-heat
conﬁguration) joining was described. A few representative results from FEA are discussed
below. Figure 7.44 and Figure 7.45 illustrate the contour line mapping of temperature with
time. At the start (0.13) of the joining process, the temperature rises to a maximum of
2863 00 (Figure 7.44(a)) at the point of interaction of the laser beam and at distances far
removed from the source, the composite plate remains at an initial room temperature of
25 0C'. At an identical time increment of 0.13, the maximum temperature for the pre-heat
beam conﬁguration was 3200 00 (ﬁgure is not given). The temperature dispersion through
the thickness is clearly shown in Figure 7.44(b), after a time interval of 0.33. For this time
increment, the maximum temperature was 2541 0C', where as for the pre-heat conﬁguration,
the maximum temperature was 2890 00.

From Figures 7.44(a) and (b) and the data for the pre-heat conditions, it is known that
the pre-heating of the weld sample by the secondary laser beam has a signiﬁcant effect on
the temperature distribution through the sample. In the case of collinear beam conﬁgura-
tion, the limited temperature rise (2863 Vs. 3200 0C) was mainly due to a signiﬁcant part
of the laser energy getting reﬂected ﬂ'om the sample surface and not being utilized for melt-
ing the material by thermal conduction. In the case of pre-heat, the temperature spread was
narrow and was conﬁned to the interaction zone. This resulted in a higher peak temperature
(3200 00) in the sample. It is well-known that the thermal conductivity for most alloys and
oxides, decreases with increasing temperature [141, 142, 238, 239]. Despite assumption of
a constant thermal conductivity, it is important to note that the heat ﬂow was better con-
trolled in the pre-heat case. By pre-heating the sample, the coupling between laser and the
material was enhanced. This phenomenon was responsible for heat ﬂux to travel largely in
the z-axis direction.

The temperature distribution after about three-fourth (2.95) of the time and just before

195

 

i

.3,
2i

2%???

Ilka

 

Elli

 

wmdmu-ll

 

 

 

(a)

 

i

1

0.770...

 

V

\4

(b)

Figure 7.44 - Contour line temperature maps (a) for collinear conﬁguration at t = 0.1 s (b)
att = 0.33.

i E E N
S E

 

minimum

 

 

 

196

end (3.63) of traverse are shown in Figures 7.45(a) and (b) respectively. These two ﬁgures
are for the collinear conﬁguration. At the end of 2.93 time increment, the peak temperature
was 2391 °C and the minimum temperature was 451 00. At the end of 2.93, for the pre-
heat condition, the peak temperature was 2760 °C and the minimum temperature was 150
06'. Again these temperature values clearly indicate that in the pre-heat condition, the
heat ﬂux was conﬁned to a very narrow region in the interaction zone. For the 3.63 time
increment, the maximum temperature was 2541 °C and the minimum was 451 00 (see
Figure 7.45(b). For this same time interval, in pre-heat condition, the peak temperature
was 2910 °C and the minimum was 220 0C'. These minimum temperature values indicate
that for the collinear case, a large part of the sample was at a higher temperature than the
pre-heated sample, which implies that a sizable amount of heat was conducted away from
the interaction zone. This loss of heat in lateral directions would lead to larger weldment
width and shallow depth of penetration. These results are in agreement with experimental
results reported elsewhere. From these contour proﬁles, one can conclude that pre-heating
the sample, helped in tying down the heat ﬂux to a narrow zone around the interaction point
of the laser beam with the material.

To study the inﬂuence of pre-heating, temperature proﬁles along the width and depth of
the weldment were generated for both collinear and pre-heat modes. The previous contour
proﬁles presented in Figure 7.44 and Figure 7 .45 did not reveal much about the depth and
width of the weldment. They present an overall view of the composite sample. The vari-
ation of temperature with time increment for a given location of a: = 25mm are recorded
along Ay = 0 to Ay = 1.25mm with an increment of 0.25 mm for width and z = 0 to
z = 6m with an increment of 1.5 mm for depth of penetration for the weldment. In Fig-
ure 7.46, the variation of temperature for :c = 25 and y = 12.5, with ‘z’ varying from 1.5,
3.0, 4.5, and 6.0 mm, for both collinear and pre-heat conditions are given. In Appendix B,
Figure B] and Figure B.2, illustrate the temperature variation for collinear and pre-heat

modes respectively. Figure 7.46 is combination of Figure 8.1 and Figure B.2, without the

197

 

 

 

mullet-ht”

 

 

 

 

(a)

 

 

... L
(MW '

Figure 7.45 - Contour line temperature maps (a) for collinear conﬁguration at t = 2.93 (b)
at t = 3.6.9. ,

 

 

 

 

198

peak temperature curve observed for the top surface of the composite plate. The same is
true for Figure 7.47, which is a combination of Figure B3 and Figure B.4, without the
peak temperature curve. In Figure 7.46 and Figure 7.47, the maximum temperature curve
was not plotted, inorder to show more detail in the other temperature variation curves. Fig-
ure 7.48 illustrates the nature of temperature proﬁles for a single beam laser welding of a
2.0 mm thick Ti with 1500 W at a traverse speed of 7.5 mm/s. The proﬁle indicated in
Figure 7.48 closely resembles the proﬁle for CD-LBW weldments in the width direction
(see Figure 7.47).

Figures 7.49 and 7.50 illustrate the predicted temperatures along the depth of penetra-
tion and width of the weld bead at various time intervals respectively. From Figure 7.49,
it can be seen that the preheat condition of the dual laser beam weld results in large depth
of penetration than the collinear mode (5.2 mm vs. 4.3 mm). This clearly agrees with
the experimental results. From a practical point of view, for the collinear mode, the laser
beam alignment is a time consuming process, since perfect alignment between the two split
beams is very diﬂicult. In Figure 7.51, the symmetric temperature distribution along the
width of the weldment at a particular location and time is given. This bell-shaped curve is

a typical distribution of temperature for all fusion welds.

199

06

dom 45885 08E.

od

ON

0.?

458203 05 we 5&3 H8 0:5 5; counter, 2302.83 mo :Steaaou - 3.5 9.53..—

06

 

 

—

885 .EE 3. I
.8265 .EE 3. T...
“8:65 .EE 0.» I
:8ch 6.5. m; I
5858 .86 cc T...
895.8 .55 mé ele

 

 

 

305:8 .EE ad I 1
595.8 .EE m4 :1.

 

 

 

 

q

in.

g

l

 

 

ed

0600

0.08.

;) mendmoj,

Odom P

odoow

200

.88 203 05 m0 523 .8: 080 553 805.5, 888880: :0 85.80800 - 3.6 0.53.,—
.08 488085 25:.

3 as as 3 od .
. _ . . . . . o o

 

   
   

i 0.08

3 mundane;

 

$0.03 .55 9N: one
.885 .55 02.71.. 989
$05.05 .55 who. I
28:05 .55 mum. I
60:05 .55 mg. I
305:8 .55 0.9 I
305:8 .55 mm: ole .
505:8 .EE mum. I
505.8 :5: mum. «to
805:8 .EE mum. one

 

 

 

Odom:

201

 

8

 

 

 

 

n T 1 I
9 At the Surface
§ 0.275 mm Below -—
.5 At the Bottom ”1
2.’
2 10
E
0
n ‘
5 /
l— /
O _ , ,- //////// .
0 0.5 1.0 1.5 2.0 2.5
Time, seconds

Figure 7.48 - Mathematical predictions of temperature proﬁles for a single beam laser
welded 2.0 mm thick Ti sample [240].

Temperature K '

202

2500.0

 

.. G—e inc-19, collinear
r H inc=19, pre-heat
A H inc=20, collinear
V—V inc=20, pre-heat
+——+— inc=21, collinear
I H inc=21, pre-heat
2000.0< _ . 0—0 inc=22, collinear
meltmg PM" <1—<l inc=22, pre-heat

    
 

1500'” ’ collinear, 4.3 mm

preheat, 5.2 mm

1000.0

 

 

 

 

500.0 - ' . 1 ' .
0.0000 0.0020 0.0040 0.0060

Distance along the depth of sample, in

 

Figure 7.49 - The depth of the weld penetration for diﬂ'erent time increments.

2500.0

203

 

Q.‘

2300.0 1,

——""

2100.0
1900.0 '3‘ \

1 700.0

Y ‘I Y T

1 500.0

1

1300.0

Temperature, C

U I r U

 

 

1100.0

V I U

900.0 1
700.0

500.0 -

 

300.0
0.000

Figure 7.50 - Temperature variation on the top surface along the width of the sample.

 
  
  

o—-—e inc-19, collinear
H inc-19, pro-heat
5—5 inc-20, collinear
H inc-20, pro-heat
i—l- inc-21, collinear
ale—ale inc=21, pre-heat
0—0 inc-22, collinear
<>—<> inc-22, pro-heat

melting point

weld speed - 0.8 mlmin

width of the weld nugget
(approximately 4.6 mm - preheat)

4.4 mm - collinear

 

  
 

0.003 0.005 0.007 0.010 0.013
Distance along the width of the sample, in

204

 

 

 

me: 10 ‘ r _ ere- _ M
we: 0

101nm

Plummet»

Tum-ur- (x1000)

2138 340

 

 

 

Figure 7.51 - Symmetric distribution of temperatures along the width of the sample.

CHAPTER 8

CONCLUSIONS

8.1 Machining

The following signiﬁcant conclusions can be drawn from the present investigation on laser

beam machining of 6061/A1203/10p and 20p aluminum composites:

0 The cover gas plays a crucial role in the machining and joining of composites. Ni-
trogen is the ideal gas for machining. Cuts made with N2 gas produced smooth out
surfaces with low weight loss and hence narrow kerf width. For a 5.0 mm thick
6061/A1203/20p composite, with N2 cover gas at 415 kPa, the weight loss was 0.27
g / sec and the kerf width at the entrance side of the beam was 0.8 mm and on the exit

side of the cut it was 0.5 mm.

o The pressure of the cover gas also has a signiﬁcant inﬂuence on the roughness of
the cut surface, depth of cut, and width of the kerf. At low pressures (for e. g. 300
kPa) the as-cut surface was very rough (average roughness was in the range of 1.8
to 2.3 pm, for a 5.0 mm thick 6061/A1203/20p composite cut at 0.508 m / min), due
to incomplete and unsteady ejection of the molten material from the kerf. For a 5.0
mm thick 6061/A1203/20p composite, the optimum N2 cover gas pressure was 415

kPa, with an average roughness of the cut surface in the range of 1.0 to 1.3 pm.

205

206

The depth of cut increases with cover gas pressure upto a certain critical pressure
(415 kPa), with drastic decrease in the cut depth beyond this critical pressure, due
to increased cooling effect and hence reduced ejection of the liquid metal from the
interaction zone. The kerf width increased from 0.8 to 1.24 mm for cover gas pres-
sures of 415 kPa and 450 IcPa respectively, for a 5.0 mm thick sample cut at 0.508
m / min with 2400 W. Narrow kerf width cuts also resulted in reduced average sur-

face roughness. The kerf width increases with decreasing cutting speed.

c There is a direct relation between the cutting speed and the thickness of the sample
that can be cut through.

c The position of the focal point from the surface of the specimen is a key factor in
deciding the allowable thickness that can be cut. Samples of 9.2 mm thick were cut
with focal point positioned at about 0.4 mm inside the surface, with 2400 W power
and 0.508 mlmin traverse speed and N2 cover gas pressure of 415 kPa.

0 Theoretical predictions indicate that the focal plane should be positioned about 3.0

mm inside the sample surface, for a 9.2 mm thick composite.

0 As the speed of cut increased, the roughness of the cut surface tends to increase.

8.2 Welding

8.2.1 Single Beam Welding

0 Laser Beam Welding is possible for some composites with a result superior to con-
ventional welding.

0 An increase in laser power resulted in an increase in the depth of penetration for a
constant traverse speed, and an increase in traverse speed led to a decrease in the

depth of penetration for a given power.

207

o For a smooth, clean and narrow weld it was necessary to focus the laser beam to
about 2 mm inside the sample.

0 Selection of proper co-axial and auxiliary cover gas are very important to control
the production of excessive plasma, which absorbs lot of beam energy and to control
splatter and loss of the molten material. A mixture of argon and helium was found
suitable for joining this type of composite.

o Traverse speed and heat-input have a signiﬁcant effect on the depth of penetration.
They also affect the size and number of pores in the weld bead.

0 Optimum welds were possible at a heat-input level of 1700 W and a traverse speed
of 2.1 cm/sec (50 ipm), for a 6.35 mm thick composite.

0 Quality welds with reproducible results were possible with aspherical optics for both
alumina and SiC composite.

o For both A1203 particulate and SiC whisker reinforced composite, plasma suppres- ,
Sion was a major problem, with reﬂective optics. Using off-axis (auxiliary) cover gas
(argon), the plasma plume was partially contained, so as to obtain minimum porosity
in the alumina composite.

o The presence of alumina particles in the composite helped in better coupling of the
laser beam with the composite. The higher the volume fraction of the particles, the
better the coupling and plasma control.

0 Butt-Welded SiC whisker reinforced composite gave inferior welds both in terms of
porosity and penetration depth.

0 The soundness of the weldments produced by LBW depend on numerous laser pro—
cess parameters (e. g., process speed, focal point position coaxial and auxiliary cover
gas etc.) and material characteristics like initial microstructure, distribution of the

reinforcement etc.

208

8.2.2 Dual Beam Welding

o The traverse speed of the LBW process dictates the solidiﬁcation and cooling rates,

which in-tum affect the formation of microﬁssures and porosity in the weldment.

o The inﬂuence of the thermal gradients on formation of porosity and microﬁssures
were studied from the point of view SB-LBW and CD—LBW. In SB-LBW, it was very
difﬁcult to avoid microﬁssures, eventhough porosities were completely eliminated.
In CD-LBW, which exhibits true conduction mode of weld, the depth of penetration
is much higher than SB-LBW. Moreover, for CD-LBW process, the particle deﬁcit
areas in the weldment can be reduced to negligible levels. Uniformity in the hardness
values of the fusion and HAZ contributed to better welds. Also the heat transfer rates

are more efﬁcient in CD-LBW when compared to SB-LBW.

APPENDICES

APPENDIX A

Theoretical Aspects of Focal Plane

Position

For a perfect lens ( no spherical aberrations), the monochromatic laser beam should be
focussed to a ﬁne spot as indicated by dotted lines in Figure A.l. In practice, one has a
focal plane and not a focal point, which results in a barrel shaped beam bundle, as shown
schematically in Figure A. l. The hatched area in Figure A.1 represents the difference in

area (volume) between the these two focussed beam bundles.

 

 

 

assumed
beam focus path

I
I
l
1.4
l‘
l
I

Figure A.l - Actual and theoretical shape of the focussed laser beam.

209

210

y = tane(h1-x)

focal point

  
 

y = tane(x-h1)

/

Laser beam

 

 

 

 

 

 

 

Figure A.2 - Geometrical model for the maximum position of the focal plane.

From geometrical considerations, one can derive a simple analytical equation to pre-
dict the maximum allowable position of the focal plane, for a maximum volumetric mass
removal. In Figure A.2, the laser beam is assumed to be interacting with the material in the
positive 2: axis, with focal plane (point) located at a distance of hl from the surface of the
material. The beam is assumed to be focussed to a point. The line equations for the surface
of the beam cone bundle are given in Figure A.2. In Figure A.2, the ﬁrst quadrant cone is
depicted. The volume (V) swept by this cone can be computed as follows:

In ’12
V = / 1ry2 dz: +/ 1ry2 dz: (A.1)
0 hr

’11 ’12
V = f 7r tan2 9(h1 — :0)2 dx +/ 7r tan2 0(1: —— h1)2 dz
0

hr

V = ntanzﬁ [Ah1(h1 —z)2dx+/h2($— h1)2d:r]

hr

V I” h:
V' = —— = h2 2 _ 2h 2 2 _ .
1rtan29 l0. ( 1+3 1x)d:i:+ h) (a: +h1 2hlr)dx (A2)

211

After integration and simpliﬁcation we get,

I ha 2 2

Differentiating the above equation with respect to hl and setting it to zero (to maximize the

volume) results in:

’12 = 2,11

Substituting, h2 = (h — hl) in the above equation:
h = 3’11 01' ’21 = 0.33h

This indicates that to achieve maximum penetration or mass removal, the focal plane should

be located approximately 1 / 3 (of the thickness) from the surface of the material.

APPENDIX B

Additional Finite Element Results

212

213

 

Whammy-n m
Tmmm

 

 

3.0m t—o—oilm um
O—O—DODn-n “‘

 

 

 

FigureBJ - Variation of temperature with time for a: = 25, y = 12.5, and 0 g z 5 6.0 for
collinear mode.

214

 

 

Tuna-nun (x1000)

 

3.0m O—‘—‘18m 4.5m
o—o—oOJM .- 1

 

 

 

 

FigureBJ - Variation of temperature with time for :1: = 25, y = 12.5, and 0 5 z s 6.0 for
pre-heat mode.

215

2'. Now in” o: :55»; .

hmwtmbm meta,

 

Figure 3.3 - Variation of temperature with time for a: = 25, 0 3 Ag S 1.25, and z = 0 for
collinear mode.

216

 

Figure 3.4 - Variation of temperature with time for a: = 25, 0 3 Ag 5 1.25, and z = 0 for
pre-heat mode.

BIBLIOGRAPHY

BIBLIOGRAPHY

[1] U. Dilthey, D. Fuest, A. Huwer, J. Schneegans, and L. J acobskétter, “High energy
laser beam welding,” in Joining/Welding 2000, pp. 89-96, Pergamon, Oxford, 1991.

[2] J. E. Allison and G. S. Cole, “Metal-matrix composites in the automotive industry:
Opportunities and challenges,” 10M, vol. 45, no. 1, pp. 19-24, 1993.

[3] K K. Chawla, Composite Materials. New York: Springer-Verlag, 1987.

[4] M. Taya and R. J. Arsenault, Metal Matrix Composites - Thermomechanical Behav-
ior. Oxford: Pergamon, First ed., 1989.

[5] J. J. Schuldies and J. A. Branch, “Ceramic composites: Emerging processes, appli-
cations,” Ceramic Industry, pp. 43-46, May 1992.

[6] K. P. Rajurkar, R. F. Ross, B. Wei, J. Kozak, and R. E. Williams, “The role of
nontraditional manufacturing processes in future manufacturing industries,” in Proc.
of Manufacturing International (D. Durham, Ed.), pp. 23—37, ASME, 1992.

[7] S. Abrate and D. A. Walton, “Machining“ of composite materials. Part H: Non-
traditional methods,” Composites Manufacturing, vol. 3, no. 2, pp. 85-94, 1992.

[8] G. F. Benedict, Nontraditional Manufacturing Processes. New York: Marcel
Dekker, 1987.

[9] S. Abrate and D. A. Walton, “Machining of composite materials. Part 1: Traditional
methods,” Composites Manufacturing, vol. 3, no. 2, pp. 75—83, 1992.

[10] H. E. Chandler, “Machining of metal-matrix composites and honeycomb structures,”
in Machining, Metals Handbook, 9th ed., vol. 16, pp. 893-901, ASM, 1989.

[11] J. Mazumder, “The state-of-the—art of laser materials processing,” in Interdisci-
plinary Issues in Materials Processing and Manufacturing, vol. 2, pp. 599—630,
Boston: ASME Annual Meeting, 1987.

217

218

[12] G. R. LaFlamme and D. E. Powers, “A detailed comparison of the EBW and LBW
processes,” in Recent Trends in Welding Science and Technology TWR ’89—Proc. of
the 2nd International Conference on Trends in Welding Research (S. A. David and
J. M. Vitek, Eds.), pp. 481-485, American Welding Society and Welding Research
Council, ASM International, Materials Park, Ohio, May 1989.

[13] E. A. Metzbower, D. W. Moon, and P. E. Denney, “Laser beam welding of HSLA
steels,” in Weldability of Materials (R. A. Patterson and K. W. Mahin, Eds.), pp. 191-
200, The Welding Metallurgy Committee of the Joining Division of ASM, ASM
International, October 1990.

[14] Y. Arata and I. Miyamoto, “Laser welding,” Technocrat, vol. 11, no. 5, pp. 33—42,
1978.

[15] T. Lehecka, N. C. Luhmann, Jr., W. A. Peebles, J. Goldhar, and S. P. Obenschain,
“Gas lasers,” in Handbook of Microwave and Optical Components (K. Chang, Ed.),
vol. 3, pp. 451—596, John Wiley & Sons, 1990.

[16] B. A. Lengyel, Lasers. New York: Wiley-Interscience, 2 ed., 1971.

[17] J. T. Luxon, D. B. Parker, and P. D. Plotkowski, lasers in Manufacturing. New
York: IFS (Pub) Ltd. and Springer-Verlag, 1987.

[18] W. W. Duley, C02 Lasers - Effects and Applications. New York: Academic Press,
1976.

[19] J. F. Ready, Industrial Applications of Lasers. New York: Academic Press, 1978.

[20] S. S. Charschan and R. Webb, “Considerations for lasers in manufacturing,” in Laser
Materials Processing (M. Bass, Ed.), ch. 9, pp. 439—473, North-Holland Pub. Co.,
1983.

[21] G. Chryssolouris, Laser Machining: Theory and Practice. New York: Springer-
Verlag, 1991.

[22] J. E. Harry, Industrial Lasers and their Applications. New York: McGraw—Hill,
1974.

[23] W. Koechner, Solid—State Laser Engineering. New York: Springer-Verlag, 2 ed.,
1988.

[24] Coherent, Inc. Staff, Lasers: Operation, Equipment, Application, and Design. New
York: McGraw-Hill, 1980.

219

[25] K Mukherjee, H. Huber, V. Bamekov, S. Kudapa, and J. Howard, “Charring,
smoothness, and parallelity of kerf in laser cutting of hardwoods,” in The 2nd Inter-
national Conference on Automated Lumber Processing Systems and Laser Machin-
ing of Wood and Composites, pp. 19-31, Dept. MSM and Dept. Forestry, Michigan
State University, 1994.

[26] D. C. O’Shea, W. R. Callen, and W. T. Rhodes, Introduction to Lasers and Their
Applications. Reading, MA: Addison-Wesley, 1977.

[27] H. M. Muncheryan, Laser Fundamentals & Applications. Indianpolis: Howard
W. Sams & Co., 1975.

[28] B. A. Lengyel, Introduction to Laser Physics. New York: John Wiley & Sons, 1967.

[29] K. Mukherjee and J. Mazumder, Lasers in Metallurgy. Warrendale, PA: AIME Pub-
lication, 1981.

[30] K. Mukherjee and J. Mazumder, Laser Processing of Materials. Warrendale, PA:
AIME Publication, 1985.

[31] J. Mazumder and K. Mukherjee, Laser Materials Processing III. Warrendale, PA:
TMS Publication, 1989.

[32] J. A. Vaccari, “Advances in LASERS - A special report,” American Machinist,
vol. 136, pp. 37—48, March 1992.

[33] J. Mazumder, K Mukherjee, and B. L. Mordike, Laser Materials Processing IV.
Warrendale, PA: TMS Publication, 1994.

[34] M. Pellecchia, “Lasers: The manufacturer’s outlook,” in Nontraditional Machining
(C. Wegner, Ed.), pp. 75—78, ASM, Dec 1985.

[35] R. Migliroe, “Laser cutting,” in Engineered Materials Handbook, Composites (T. J.
Reinhart and C. A. Dostal, Eds.), vol. 1, pp. 676—680, ASM International, 1988.

[36] R. R Schrelber, “Lasers: Technology advances are boosting the usefulness of C02,
Nd:YAG, and excimer lasers in manufacturing. Part I,” Manufacturing Engineering,
vol. 106, pp. 41-44, January 1991.

[37] R. R. Schrelber, “Lasers: Lasers can help you cut up the competition. Part 11,” Man-
ufacturing Engineering, vol. 106, pp. 281—290, February 1991.

[38] C.—W. Chen, Use of Nd: YAG Laser for Processing High Tc Superconductors. PhD
thesis, Michigan State University, 1993.

220

[39] K. Tsushima, H. Miura, and K. ichi Shiroki, “YAG laser and its applications,”
in Proc. of the Materials Processing Symposium, ICALEO ’84, LIA volume 44
(J. Mazumder, Ed.), pp. 269—275, 1984.

[40] N. Croxford, “Development of high average power solid-state lasers for the engineer-
ing industries,” in Lasers in Manufacturing (MM 6), Proc. of the 6th International
Conference (W. M. Steen, Ed), pp. 11—21, IFS Conferences, Bedford, UK 1989.

[41] P. V. Ionov, “Technological and design features of YAG laser application for cutting,
marking and pulse welding,” in Industrial Lasers and Laser Material Processing:
Russia National Conference (V. Y. Panchenko and V. S. Golubev, Eds.), pp. 39-42,
Proc. SPIE 2257, 1994.

[42] P. W. Milonni and J. H. Eberly, Lasers. New York: John Wiley & Sons, 1988.

[43] I. Miyamoto, T. Ooie, Y. Hirota, and H. Maruo, “Mechanism of laser ablation -
Ablation process and debris formation,” in Laser Materials Processing, ICALEO ’93
(P. Denney, I. Miyamoto, and B. L. Mordike, Eds.), vol. SPIE 2306, pp. 1—1 1, Laser
Institute of America and SPIE, Laser Institute of America, 1993.

[44] K Haruta, “Development of high power excimer laser under AMMTRA project,”
in Laser Materials Processing, ICALEO’93 (P. Denney, I. Miyamoto, and B. L.
Mordike, Eds.), vol. SPIE 2306, pp. 270—277, Laser Institute of America and SPIE,
Laser Institute of America, 1993.

[45] M. Stehle, “W111 kilowatt excimer lasers become future industrial tools?,” Laser Fo-
cus World, vol. 29, no. 6, pp. 135-139, 1993.

[46] Hi. Montague, Th. Samet, G. Inglesakis, and M. Autric, “High energy, long-pulse,
electron-beam driven KrF laser in laser-matter interaction,” in Laser Materials Pro-
cessing, ICALEO’93 (P. Denney, I. Miyamoto, and B. L. Mordike, Eds.), vol. SPIE
2306, pp. 620—629, Laser Institute of America and SPIE, Laser Institute of America,
1993.

[47] D. W. Trainor, “Military excimer—laser technology seeks real-world uses,” Laser Fo-
cus World, vol. 29, no. 6, pp. 143—149, 1993.

[48] Th. Samet, J. E. Montague, G. Inglesakis, M. Autric, L. Barrallier, and G. Bar-
reau, “Characterization of metal surfaces irradiated by a long-pulse KrF excimer
laser,” in Laser Materials Processing, ICALEO’93 (P. Denney, I. Miyamoto, and
B. L. Mordike, Eds.), vol. SPIE 2306, pp. 867-874, Laser Institute of America and
SPIE, Laser Institute of America, 1993.

221

[49] H. C. Man, M. Li, and T. M. Yue, “Surface pretreatment of polymer composites
by excimer laser for adhesion bonding,” in Proc. of ICCM-IO, vol. III, pp. 559-564,
The Tenth International Conference on Composite Materials Society, Whistler, BC,
Canada, Woodhead Pub. Ltd, August 1995.

[50] M. Bass, “Lasers for laser materials processing,” in Laser Materials Processing
(M. Bass, Ed.), ch. 1, pp. 1—14, North-Holland Pub. Co., 1983.

[51] K. Mukherjee, C. W. Chen, J. Yoo, 1. Oh, and S. Kudapa, “Laser processing of
high Tc superconductors,” in Laser processing: Surface treatment and ﬁlm deposi-
tion (J. Mazumder, O. Conde, R. Villar, and W. Steen, Eds.), pp. 637—664, Kluwer
Academic. Publishers, 1996.

[52] E. M. Breinan and B. H. Kear, “Rapid solidiﬁcation laser processing at high power
density,” in Laser Materials Processing (M. Bass, Ed.), ch. 5, pp. 234—295, North-
Holland Pub. Co., 1983.

[53] R. Miller, “Automatic laser scribing aircraft panels,” Industrial Laser Review,
vol. 10, no. 6, pp. 11-12, 1995.

[54] 2. Sun and J. C. Ion, “Review: Laser welding of dissimilar metal combinations,”
J. Materials Science, vol. 30, no. 17, pp. 4205-4214, 1995.

[55] A. V. La Rocca, “Laser applications in manufacturing,” Scientiﬁc American,
vol. 246, no. 3, pp. 94-103, 1982.

[56] J. R. Koelsch, “Use the beam for better cutting,” Manufacturing Engineering,
vol. 106, pp. 51—55, January 1991.

[57] P. M. Moaker, “Light but tough,” Manufacturing Engineering, vol. 110, no. 1,
pp. 28-34, 1993.

[58] A. P. Divecha, S. G. Fishman, and S. D. Karmarkar, “Silicon carbide reinforced
aluminum - A fonnable composite,” J. Metals, vol. 30, no. 9, pp. 12—17, 1981.

[59] R. J. Arsenault, “The strengthening of aluminum alloy 6061 by ﬁber and platelet
silicon carbide,” Materials Science and Engineering, vol. 64, no. 2, pp. 171—181,
1984.

[60] T. G. Nieh and D. J. Chellman, “Modulus measurements in discontinuous reinforced
aluminum composites,” Scripta Metallurgica, vol. 18, pp. 925-928, 1984.

222

[61] S. V. Nair, J. K. Tein, and R. C. Bates, “SiC - Reinforced aluminum metal matrix
composites,” Int. Metals Reviews, vol. 30, no. 6, pp. 275—290, 1985.

[62] D. L. MacDanels, “Analysis of stress - strain, fracture, and ductility behavior of alu-
minum matrix composites containing discontinuous silicon carbide reinforcement,”
Metallurgical Transactions A, vol. 16A, no. 6, pp. 1105-1115, 1985.

[63] K. Stanford, “Metal matrix composites in Great Britain,” Powder Metallurgy Inter-
national, vol. 21, no. 5, pp. 40—43, 1989.

[64] T. Kyono, T. Hanano, and K. Shimizu, “Study on compatibility between carbon
ﬁber “TORAYCA” and aluminum matrix,” in New Materials and Processes for the
Future, Proc. of The First Japan International SAMPE Symposium and Exhibition
(N. Igata, I. Kimpara, T. Kishi, E. Nakata, A. Okura, and T. Uryu, Eds.), pp. 964-
969, Society for the Advancement of Material and Process Engineering, November
1989.

[65] P. Fleischmann, Y. Kagawa, and R. Fougeres, “Measurement and analysis of the
stiffness and Bauschniger effect during loading and unloading of a carbon ﬁber
and aluminum matrix composite,” in Developments in the Science and Technol-
ogy of Composite Materials - Fourth European Conference on Composite Materials,
ECCM4 (J. Fuller, G. Gruninger, K Schulte, A. R. Bunsell, and A. Massiah, Eds.),
pp. 275—280, Elsevier Applied Science, September 1990.

[66] K Schulte, A. Bockheiser, F. Girot, and G. K. Kim, “Characterization of aluminum
oxide (FP) ﬁbre reinforced Al-2,5Li - composites,” in Developments in the Science
and Technology of Composite Materials - Fourth European Conference on Com-
posite Materials, ECCM4 (J. Fuller, G. Gruninger, K. Schulte, A. R. Bunsell, and
A. Massiah, Eds.), pp. 293—300, Elsevier Applied Science, September 1990.

[67] E. E. Gdoutos, D. Karalekas, and I. M. Daniel, “Thermal stress analysis of a sili-
con carbide/aluminum composite,” Proc. of the Society for Experimental Mechanics
(Formerly the Society for Experimental Stress Analysis), vol. XLVIII, pp. 202-208,
September 1991.

[68] F. A. Girot, J. M. Quenisset, and R. Naslain, “Discontinuously reinforced aluminum
matrix composites,” Composite Science and Technology, vol. 30, no. 3, pp. 155—184,
1987.

[69] S. Coleman, B. McEnaney, and V. Scott, “Corrosion studies on metal matrix com-
posites,” in Developments in the Science and Technology of Composite Materi-
als - Fourth European Conference on Composite Materials, ECCM4 (J. Fuller,

223

G. Gruninger, K Schulte, A. R. Bunsell, and A. Massiah, Eds.), pp. 287-292, Else-
vier Applied Science, September 1990.

[70] M. Mehta, “Composite materials in reciprocating engines: A state-of-the-art re-
view,” in Advanced Composites X, Proc. of the 10th Annual ASM/BSD Advanced
Composites Conference, pp. 141-153, ASM/EDS, ASM International, 1994.

[71] K B. Das, “Joining of metal matrix composites,” in International Encyclopedia of
Composites (S. M. Lee, Ed.), vol. 2, pp. 460-509, VCH Publishers Inc., 1989.

[72] L. J. Masur, A. Mortensen, J. A. Cornie, and M. C. Flemings, “Pressure casting of
ﬁber reinforced metals,” in Proc. of the sixth international conference on composite
materials( I CCM- VI), American Institute of Mining, Metallurgical and Petroleum
Engineers, 1987. -

[73] A. Mortensen, J. A. Comic, and M. C. Flemings, “Columnar dendritic solidiﬁcation
' in a metal-matrix composite,” Metallurgical Transactions A, vol. 19A, pp. 709—721,
March 1988.

[74] R. Trumper and V. Scott, “Cast ﬁbre reinforced aluminium alloy microstructures,” in
Developments in the science and technology of composite materials: Proceedings of ‘
the third European conference on composite materials, Bordeaux, France, pp. 139-
144, 1989.

[75] J. Lecomete-Beckers, D. Coutsouradis, and M. Lamberigts, “SiC ﬁbers and parti-
cles reinforced Aluminium alloys produced by stir casting,” in Developments in the
Science and Technology of Composite Materials - Fourth European Conference on
Composite Materials, ECCM4 (J. Fuller, G. Gruninger, K. Schulte, A. R. Bunsell,
and A. Massiah, Eds.), pp. 387-392, Elsevier Applied Science, September 1990.

[76] B. R. Henriksen, “The microstructure of squeeze-cast SiC,-reinforced A4Cu base
alloy with Mg and Ni additions,” Composites, vol. 21, pp. 333-338, July 1990.

[77] G. A. Chadwick, “Squeeze casting of metal matrix composites using short ﬁbre pre-
forms,” Materials Science and Engineering, vol. A135, no. 4, pp. 23—28, 1991.

[78] C. G. Kang and G. S. Ku, “An experimental investigation on inﬁltration limit and
the mechanical properties of A12 03 short ﬁber reinforced metal matrix composites
fabricated by squeeze casting,” J. Composite Materials, vol. 29, no. 4, pp. 444—462,
1995.

224

[79] H. Asanuma and A. Okura, “Studies on the fabrication of carbon ﬁber reinforced
aluminum composites by the plasma spray and roll diffusion bonding method (in
Japanese),” J.Japan Inst.Metals, vol. 48, no. 11, pp. 1119-1126, 1984.

[80] A. Mortensen, J. A. Comic, and M. C. Flemings, “Solidiﬁcation processing of metal-
matrix composites,” J. Metals, vol. 40, no. 2, pp. 12—19, 1988.

[81] H. D. Steffens and R. Kaczrnarek, “Metal matrix composite made by thermal spray-
ing,” Powder Metallurgy International, vol. 23, no. 2, pp. 105-107, 1991.

[82] C. M. Tung, D. S. Gnanamuthu, R. J. Moores, and C. L. Leung, “Effect of laser
radiation on the rapid processing of graphite - reinforced composites,” Advanced
Materials Technology ’87, 32nd International SAMPE Symposium and Exhibition,
vol. 32, pp. 476—483, 1987.

[83] L. L. Clements, H. S. Lakarraju, R. Kellman, D. R. Osment, and B. A. Wilson,
“Laser curing of epoxies and epoxy - matrix composites,” Advanced Materials
Technology ’87, 32nd International SAMPE Symposium and Exhibition, vol. 32,
pp. 1264-1275, 1987.

[84] M. Okumura, S. Murakami, S. Utsunomiya, and T. Morita, “Study on fabrication of
MMC with laser beam,” in New Materials and Processes for the Future, Proc. of The
First Japan International SAMPE Symposium and Exhibition (N. Igata, I. Kimpara,
T. Kishi, E. Nakata, A. Okura, and T. Uryu, Eds.), pp. 1002-1007, Society for the
Advancement of Material and Process Engineering, November 1989.

[85] M. Okumura, S. Murakami, S. Utsunomiya, T. Morita, and S. Hiramoto, “Mechan-
ical pr0perties of MMC fabricated with laser beam,” SAMPE Quarterly, pp. 56—63,
July 1990.

[86] F. Ferraro, G. D. Vita, M. Marcchetti, A. Cutolo, and L. Zeni, “Laser processing of
thermoplastic matrix ﬁlament wound composites,” in Developments in the Science
and Technology of Composite Materials - Fourth European Conference on Com-
posite Materials, ECCM4 (J. Fuller, G. Gruninger, K. Schulte, A. R. Bunsell, and
A. Massiah, Eds.), pp. 89—94, Elsevier Applied Science, September 1990.

[87] C. Hu and T. N. Baker, “AA6061 Al-SiC, surface MMC’s produced by laser pro-
cessing,” in Proc. of ICCM-I 0, Metal Matrix Composites (A. Poursartip, Ed.), vol. II,
pp. 183-190, The Tenth International Conference on Composite Materials Society,
Whistler, B.C., Canada, Woodhead Pub. Ltd, August 1995.

225

[88] C. Hu and T. N. Baker, “Laser processing to create in-situ Al-SiCp surface metal
matrix composites,” J. Materials Science, vol. 30, no. 4, pp. 891—897, 1995.

[89] C. Hu, H. Xin, and T. N. Baker, “Laser processing of an aluminium AA6061 al-
loy involving injection of SiC particulate,” J. Materials Science, vol. 30, no. 23,
pp. 5985—5990, 1995.

[90] J. C. Lee, Mechanical Working of (A1203)p/Al Composite. PhD thesis, Michigan
State University, 1993.

[91] D. Lloyd, “Fabrication of ﬁbre composites using an aluminium superplastic alloy as
matrix,” J. Materials Science, no. 19, pp. 2488—2492, 1984.

[92] C.-C. Pemg, J.-R. Hwang, and J .-L. Doong, “High strain rate tensile properties of
an (A1203 particles)-(Al alloy 6061-T6) metal matrix composite,” Materials Science
and Engineering, vol. A171, pp. 213-221, November 1993.

[93] H. J. McQueen, E. V. Konopleva, M. Myshlyaev, and Q. Qin, “Effect of matrix alloy
6061, 7075, A356 on hot workability of particle composites,” in Proc. of ICCM-
10, Metal Matrix Composites (A. Poursartip, Ed), vol. II, pp. 423-430, The Tenth
International Conference on Composite Materials Society, Whistler, B.C., Canada,
Woodhead Pub. Ltd, August 1995.

[94] P. R. Smith and F. H. Froes, “Developments in titanium metal matrix composites,”
J. of Metals, no. 36, pp. 19-26, 1984.

[95] W. Wei, “High temperature low cycle fatigue of a SiC/Ti matrix composite,” in
Developments in the Science and Technology of Composite Materials - Fourth
European Conference on Composite Materials, E CCM4 (J. Fuller, G. Gruninger,
K. Schulte, A. R. Bunsell, and A. Massiah, Eds.), pp. 319—324, Elsevier Applied
Science, September 1990.

[96] P. Backe, H. Schurmans, and J. Verhoest, Inorganic Fibre and Composite Materials.
Oxford: Pergmon Press, 1983.

[97] C. A. Calow and A. Moore, “Are metal matrix composites viable?,” in Practical
Metallic Composites, pp. B17-B25, Institution of Metallurgists, London, 1974.

[98] S. Caron and J. Masounave, “A literature review of fabrication techniques of partic-
ulates reinforced metal composites,” in Fabrication of Particulates Reinforced Metal
Composites (J. Masounave and F. G. Hamel, Eds.), pp. 79—86, ASM International,
Ohio, 1990.

226

[99] V. Laurent, D. Chatain, and N. Eustathopoulos, “Wettability of SiC by aluminium
and Al-Si alloys,” .1. Materials Science, vol. 22, pp. 244-250, January 1987.

[100] K. Yu, V. Dollohopf, and R Kochendorfer, “Wettability and mechanical properties of
CVD SiC ﬁlament and reinforced aluminium in a vacuum suction casting process,”
in Developments in the Science and Technology of Composite Materials - Fourth
European Conference on Composite Materials, ECCM4 (J. Fuller, G. Gruninger,
K Schulte, A. R. Bunsell, and A. Massiah, Eds.), pp. 399-404, Elsevier Applied
Science, September 1990.

[101] S. W. Ip, M. Kucharski, and J. M. Toguri, “Wetting behaviour of aluminium and
aluminium alloys on A1203 and CaO,” J. Materials Science Letters, vol. 12, no. 21,
pp. 1699-1702, 1993.

[102] G. R. Cappleman, J. F. Watts, and T. W. Clyne, “The interface region in squeeze-
inﬁltrated composites containing 6-alumina ﬁbre in an aluminium matrix,” J. Mate-
rials Science, vol. 20, no. 6, pp. 2159-2168, 1985.

[103] J. G. Legoux, G. L’Espérance, L. Salvo, and M. Suéry, “Inﬂuence of particle oxida-
tion on the interfacial structure of SiC reinforced Al-1% Mg alloys,” in Fabrication
of Particulates Reinforced Metal Composites (J. Masounave and F. G. Hamel, Eds.),
pp. 31-39, ASM International, Ohio, 1990.

[104] D. J. Lloyd, I. Jin, and G. C. Weatherly, “Controlling the interface reaction in alu-
mina reinforced aluminium composites,” Scripta Metallurgica et Materialia, vol. 31,
no. 4, pp. 393-396, 1994.

[105] I. Dutta, S. M. Allen, and J. L. Haﬂey, “Effect of reinforcement on the aging re-
sponse of cast 6061 Al-A1203 particulate composites,” Metallurgical Transactions
A, vol. 22A, no. 11, pp. 2553-2563, 1991.

[106] J. Wang, M. Furukawa, Z. Horita, M. Nemoto, Y. Ma, and T. G. Langdon, “The
age-hardening characteristics of an Al-6061/A1203 metal matrix composite,” Metal-
lurgical Transactions A, vol. 26A, pp. 581-587, March 1995.

[107] V. C. Nardone, “Assessment of models used to predict the strength of discontinuous
silicon carbide reinforced aluminium alloys,” Scripta Metallurgica, vol. 21, no. 10,
pp.l313—l318,1987.

[108] R. H. Jones, C. A. Lavender, and M. T. Smith, “Yield strength-fracture toughness
relationships in metal matrix composites,” Scripta Metallurgica, vol. 21, no. 11,
pp. 1565—1570, 1987.

227

[109] R. G. C. Arridge, “Particulate composites,” in International Encyclopedia of Com-
posites (S. M. Lee, Ed), vol. 6, pp. 235-250, VCH Publishers Inc., 1989.

[110] H. X. Li, J. Lo, and X. R. Xiao, “Strengthening of A1203l6061 Al particulate com—
posite at liquid nitrogen temperature,” Scripta Metallurgica et Materialia, vol. 33,
no. 9. PP. 1423—1429, 1995.

[1 l l] R. J. Arsenault, “Microstructure of ﬁber and particulate SiC in 6061 Al composites,”
Scripta Metallurgica, vol. 17, no. 1, pp. 67-71, 1983.

[112] C. G. K Nair, M. R. Krishnadev, and D. Dutta, “Development of particulate rein-
forced high strength aluminum alloy for aerospace applications,” Advanced Materi-
als Technology ’87, 32nd Intemational SAMPE Symposium and Exhibition, vol. 32,
pp.889—901,1987.

[113] C. H. Andersson and R. Warren, “Silicon carbide ﬁbres and their potential for use in
composite materials. Part I,” Composites, vol. 15, pp. 16—24, January 1984.

[114] C. H. Andersson and R. Warren, “Silicon carbide ﬁbres and their potential for use in
composite materials. Part II,” Composites, vol. 15, pp. 101—111, April 1984.

[115] R. N. Singh, “Mechanical properties of a zircon matrix composite reinforced with
silicon carbide whiskers and ﬁlaments,” J. Materials Science, vol. 26, pp. 1839-
1846, 1991.

[116] M. M.Schwartz, “Metal-matrix composite processing,” in Composite Materials
Handbook (M. M. Schwartz, Ed), pp. 4. 71—4. 94, McGraw-Hill Co., 1992.

[117] K G. Satyanarayana, B. C. Pai, M. R. Krishnadev, and C. G. K. Nair,_“Aluminium
alloy metal matrix composites for engineering applications,” Advanced Materials
Technology ’87, 32nd International. SAMPE Symposium and Exhibition, vol. 32,
pp. 880—888, 1987.

[118] T. Stephenson, Y. Lepetitcorps, and J. M. Quenisset, “The interfacial role of silica
on ceramic ﬁbre-reinforced aluminum alloy composites,” in Developments in the
Science and Technology of Composite Materials - Fourth European Conference on
Composite Materials, ECCM4 (J. Fuller, G. Gruninger, K. Schulte, A. R. Bunsell,
and A. Massiah, Eds.), pp. 255—260, Elsevier Applied Science, September 1990.

[119] C. Friend, C. Naish, T. M. O’Brien, and G. Sample, “The corrosion characteristics of
a model C/Al metal matrix composite,” in Developments in the Science and Technol-
ogy of Composite Materials - Fourth European Conference on Composite Materials,

228

ECCM4 (J. Fuller, G. Gruninger, K Schulte, A. R. Bunsell, and A. Massiah, Eds.),
pp. 307-312, Elsevier Applied Science, September 1990.

[120] A. L. Geiger and M. Jackson, “Low-expansion MMCs boost avionics,” Advanced
Materials & Processes, vol. 136, pp. 23-30, July 1989.

[121] W. R. Mohn and D. Vukobratovich, “Recent applications of metal matrix composites
in precisioninstnrments and optical systems,” Optical Engineering, vol. 27, pp. 90—
98, February 1988.

[122] T. G. Nieh and D. J. Chellman, “Modulus measurements in discontinuous reinforced
aluminium composites,” Scripta Metallurgica, vol. 18, no. 9, pp. 925-928, 1984.

[123] Y. Flom and R. J. Arsenault, “Deformation of SiC/Al composites,” J. Metals, vol. 38,
no. 7, pp. 31—34, 1986.

[124] N. Tsangarakis, B. O. Andrews, and C. Cavallaro, “Mechanical properties of some
silicon carbide reinforced aluminum composite,” J. Composite Materials, vol. 21,
no.5, pp. 481-492, 1987.

[125] R. J. Arsenault and S. B. Wu, “A comparison of PM vs. melted SiC/Al composites,”
Scripta Metallurgica, vol. 22, no. 6, pp. 767-772, 1988.

[126] F. M. Hosking, F. F. Portillo, and R. Mehrabian, “Composites of aluminium alloys:
Fabrication and wear behaviour,” J. Materials Science, vol. 17, pp. 477-498, Febru-
ary 1982.

[127] W. A. Logsdon and P. K. Liaw, “Tensile, fracture toughness and fatigue crack grth
rate properties of silicon carbide whisker and particulate reinforced aluminium metal
matrix composites,” Engineering Fracture Mechanics, vol. 25, no. 5, pp. 737—751,
1986. '

[128] T. Christrnan and S. Suresh, “Effects of SiC reinforcement and aging treatment on
fatigue crack growth in an Al-SiC composite,” Materials Science and Engineering,
vol. A102, no. 2, pp. 211-216, 1988.

[129] J. K. Shang, W. Yu, and R. O. Ritchie, “Role of silicon carbide particles in fatigue
crack growth in SiC-particulate-reinforced aluminium alloy composites,” Materials
Science and Engineering, vol. A102, no. 2, pp. 181-192, 1988.

[130] P. M. Singh and J. J. Lewandowski, “Effects of heat treatment on stress corrosion
cracking of a discontinuously reinforced aluminum (DRA) 7XXX alloy during slow

229

strain rate testing,” Scripta Metallurgica et Materialia, vol. 33, no. 9, pp. 1393-1399,
1995.

[131] V. C. Nardone and J. R. Strife, “Analysis of the creep behavior of silicon carbide
reinforced 2124 Al (T 4),” Metallurgical Transactions A, vol. 18A, no. 1, pp. 109-
114, 1987.

[132] K S. Ravichandran and E. S. Dwarakadass, “Advanced aerospace Al alloy,” J. Met-
als, vol. 39, no. 6, pp. 28—32, 1987.

[133] K. Mukherjee, P. A. A. Khan, and M. Tayal, “Laser machining of composites,” in
Laser Materials Processing III (J. Mazumder and K. Mukherjee, Eds.), pp. 217-231,
TMS, 1989.

[134] M. Dell’Erba, L. M. Galantucci, and S. Miglietta, “An experimental study on laser
drilling and cutting of composite materials for the aerospace industry using excimer
and C02 sources,” Composites Manufacturing, vol. 3, no. 1, pp. 14—19, 1992.

[135] V. Tagliaferri, I. C. Visconti, and A. D. Ilio, “Machining of ﬁbre reinforced materials
with laser beam: Cut quality evaluation,” in 6th Int. Conﬁ Composite Materials,
ICCM and ECCM (Second European Conference on Composite Materials), vol. 1,
pp. 1. 190—1. 198, 1987.

[136] V. Tagliaferri, A. D. Ilio, and I. C. Vrsconti, “Laser cutting of ﬁbre-reinforced
polyesters,” Composites, vol. 16, no. 4, pp. 317-325, 1985.

[137] G. L. Vregelahn, S. Kawali, R. J. Scheuerman, and C. W. Browne, “LASER cutting
of ﬁberglass/polyester resin composites,” in Advanced Composites Materials: New
Developments and Applications, Proc. of the Seventh Annual ASM/BSD Advanced
Composites Conference, pp. 143-147, ASM/EDS, ASM International, 1991.

[138] R. J. Wallace and S. M. Copley, “Shaping silicon nitride with a carbon dioxide laser
by overlapping multiple grooves,” J. Engineering for Industry, vol. 111, pp. 315-
321, November 1989.

[139] R. J. Wallace, M. Bass, and S. M. Copley, “Curvature of laser machined grooves in
Si3N4,” J. Applied Physics, vol. 59, no. 10, pp. 3555—3560, 1986.

[140] G. Caprino and V. Tagliaferri, “Maximum cutting speed in laser cutting of ﬁber
reinforced plastics,” Int. J. Mach. Tools. Manufact, vol. 28, no. 4, pp. 389-398,
1988.

230

[141] M. F. Modest and H. Abakians, “Heat conduction in a moving semi-inﬁnite solid
subjected to pulsed laser irradiation,” J. Heat Transfer, vol. 108, no. 3, pp. 597-601,
1986.

[142] M. F. Modest and H. Abakians, “Evaporative cutting of a semi-inﬁnite body with a
moving CW laser,” J. Heat Transfer, vol. 108, no. 3, pp. 601—607, 1986.

[143] G. Chryssolouris, P. Sheng, and W. C. Choi, “Three-dimensional laser cutting of
composite materials,” in Machining of Composites (M. Taya and M. Ramulu, Eds.),
pp. 19—30, ASME Wmter Annual Proc., PED 35, 1988.

[144] R. F. Firestone and E. J. Vesely Jr, “High power laser beam machining of structural
ceramics,” in Symposium on Machining of Advanced Ceramic Materials and Com-
ponents (S. Chandrasekar, R. Komanduri, W. Daniels, and W. Rapp, Eds.), pp. 215-
227, ASME, 1988.

[145] M. A. Clyde, Jr. and G. A. Hardway, “Fundamentals of laser beam machining and
drilling,” IEEE Trans. Ind. and General Appl. , pp. 90—96, 1965.

[146] T. C. Ramaraj, T. D. McCay, and M. H. McCay, “Preliminary investigation of C02
laser glazing of WC-Co tool materials,” in Symposium on Machining of Advanced '
Ceramic Materials and Components (S. Chandrasekar, R. Komanduri, W. Daniels,
and W. Rapp, Eds.), pp. 199—214, ASME, 1988.

[147] G. F. Benedict, “Introduction to machining methods,” in Machining, Metals Hand-
book, 9th ed, vol. 16, pp. 509-510, ASM, 1989.

[148] C. E. Johnston, “Waterjet/Abrasive waterjet machining,” in Machining, Metals
Handbook, 9th ed, vol. 16, pp. 520-527, ASM, 1989.

[149] M. Hashish, “A model for abrasive-waterjet (AWJ) machining,” Engineering Mate-
rials Techonology, vol. 111, pp. 154-162, 1989.

[150] M. D. Mello, “Laser cutting of nonmetallic composites,” in Composites in Manufac-
turing 6, pp. l-6, Society of Manufacturing Engineers, 1987. Paper No. EM87-138.

[151] J. A. McGeough, Principles of Electrochemical Machining. London: Chapman and
Hall, 1981.

[152] I. M. Crichton, J. A. McGeough, W. Munro, and C. White, “Comparative studies of
ECM, EDM, and ECAM,” Precision Engineering, vol. 3, no. 3, pp. 155—398, 1981.

231

[153] J. F. Wilson, Theory and Practice of Electrochemical Machining. London: John
Wiley & Son, 1971.

[154] J. Crookall and T. Shaw, “Why EDM is expanding from the toolroom to the
shopﬁoor,” The Production Engineer, vol. 59, no. 1, pp. 19—23, 1980.

[155] J. R. Davis, Ed, ASM Specialty Handbook - Aluminum and Aluminum Alloys,
ch. Welding, pp. 376-419. ASM International, 1994.

[156] M. M. Schwartz, Joining of Composite Matrix Materials. Materials Park, OH: ASM
International, 1995.

[157] P. L. Threadgill, “The prospects for joining titanium aluminides,” Materials Science
& Engineering A, vol. A 192/193, pp. 640—646, 1995.

[158] G. E. Metzger, “Joining of metal-matrix ﬁber-reinforced composite materials,” Weld-
ing Research Council Bulletin (WRC Bulletin), no. 207, pp. 1—21, 1974.

[159] T. Iseki, T. Kameda, and T. Maruyama, “Interfacial reactions between SiC and alu-
minium during joining,” J. Materials Science, vol. 19, pp. 1692—1698, May 1984.

[160] K. Saganuma, T. Okamoto, and N. Suzuki, “Joining of alumina short-ﬁbre rein-
forced AA6061 alloy to AA6061 alloy and to itself,” .I. Materials Science, vol. 22,
pp. 1580—1584, May 1987.

[161] S. Utsunomiya, M. Okumura, T. Morita, and S. Hiramoto, “Study on fabrication of
MMC with laser beam,” in Proc. 4th Japan-U. S. Conference on Composite Materi-
als, pp. 467-475, 1988.

[162] W. A. Baeslack III and F. H. Froes, “Joining similar and dissimilar advanced engi—
neered materials,” JOM, vol. 47, pp. 13—15, April 1995.

[163] B. Irving, “What’s being done to weld metal-matrix composites ?,” Welding Journal,
vol. 70, pp. 65-67, June 1991.

[164] B. Irving, “Automotive engineers plunge into tomorrow’s joining problems,” Weld-
ing Journal, vol. 73, pp. 47-50, Dec 1994.

[165] A. Hirose, S. Fukumoto, and K. F. Kobayashi, “Joining processes for structural ap-
plications of continuous ﬁber reinforced MMCs,” in Metal Matrix Composites -
Part 2: Microstructural characterization, thermomechanical behavior and model-
ing (G. M. Newaz, H. Neber-Aeschbacher, and F. H. Wb'hlbier, Eds.), vol. 104-107,
pp. 853—872, Trans Tech Pub., 1995.

232

[166] O. T. Midling and O. Grong, “Joining of particle reinforced Al-SiC 'MMCs,” in Key
Engineering Materials, Metal Matrix Composites - Part I : Application and Process-
ing (G. M. Newaz, H. Neber-Aeschbacher, and F. H. thlbier, Eds.), vol. 104-107,
pp. 355-372, Trans Tech Pub., 1995.

[167] M. B. D. Ellis, “Joining of Al-based metal matrix composites - A review,” Materials
and Manufacturing Processes, vol. 11, no. 1, pp. 45-66, 1996.

[168] P. R. Bergquist, S. P. Petrie, and S. E. Wentworth, “An assessment of the adhesive
bondability of silicon carbide reinforced aluminum,” in 22nd Int. SAMPE Tech Conf.
(L. D. Michelove, R. P. Caruso, and P. Adams, Eds.), pp. 815-822, SAMPE, Nov
1990.

[169] J. Doesburg, S. H. J. Lo, and G. J. C. Carpenter, “Adhesive bonding of AlgO3/6061
aluminum metal-matrix composites,” in Proc. of ICCM-10, Metal Matrix Compos-
ites (A. Poursartip, Ed), vol. II, pp. 177-182, The Tenth International Conference on
Composite Materials Society, Whistler, B.C., Canada, Woodhead Pub. Ltd, August
1995.

[170] M. B. D. Ellis, “Joining of metal matrix composites - A review,” tech. rep., The
Welding Institute, 1994. Report 489.

[171] P. G. Partridge and D. V. Dunford, “The role of interlayers in diffusion bonded joints
in metal matrix composites,” J. Materials Science, vol. 26, pp. 2255-2258, 1991.

[172] T. Enjo, K lkeochi, Y. Murakami, and N. Suzuki, “Diffusion bonding of Al-Mg-
Si series 6063 alloy reinforced with alumina ﬁbers,” Transactions of JWRI, vol. 16,
no. 2. PP. 285-292, 1987.

[173] T. Enjo, K lkeochi, Y. Murakami, and N. Suzuki, “Diffusion bonding A6063 alu-
minium alloy reinforced with alumina short ﬁbers,” Quarterly J. Japanese Welding
Society, vol. 5, no. 3, pp. 317—323, 1987.

[174] K Sudhakar, Joining of Aluminum based particulate-reinforced metal-matrix com-
posites. PhD thesis, Ohio State University, 1990.

[175] R. Klehn, “Joining 6061 aluminum matrix ceramic particle reinforced composites,”
Master’s thesis, Massachusetts Institute of Technology, 1991.

[176] V. Sabathier, G. R. Edwards, and C. E. Cross, “Kinetic study of low-temperature
transient liquid phase joining of an aluminum-SiC composite,” Metallurgical and
Materials Transactions A, vol. 25A, no. 10, pp. 2705-2714, 1994.

233

[177] Z. Li, Y. Zhou, and T. H. North, “Counteraction of particulate segregation during
transient liquid-phase bonding of aluminium-based MMC material,” J. Materials
Science, vol. 30, no. 4, pp. 1075-1082, 1995.

[178] S. Fukumoto, K. Imarnura, A. Hirose, and K. F. Kobayashi, “Transient liquid phase
bonding process in SiC ﬁber reinforced Ti-6Al-4V composites,” ISIJ International
(The Iron and Steel Institute of Japan, Special issue on recent advances in welding
metallurgy), vol. 35, no. 10, pp. 1307—1314, 1995.

[179] M. J. Cola, “Inertia-friction welding of particulate-reinforced aluminum matrix com-
posites,” Master’s thesis, Ohio State University, 1992.

[180] O. T. Midling and O. Grong, “A process model for friction welding of Al-Mg-Si
alloys and Al-SiC metal matrix composites-I. HAZ temperature and strain rate dis-
tribution,”Acta. metall. mater, vol. 42, no. 5, pp. 1595-1609, 1994.

[181] O. T. Midling and O. Grong, “A process model for friction welding of Al-Mg-Si al-
loys and Al-SiC metal matrix composites-II. HAZ microstructure and strength evo-
lution,” Acta. metall. mater, vol. 42, no. 5, pp. 1611—1622, 1994.

[182] Y. Zhou, Z. Li, L. Hu, A. Fuji, and T. H. North, “Mechanical properties of particulate
MMC/AISI 304 friction joints,” ISIJ International (The Iron and Steel Institute of
Japan, Special issue on recent advances in welding metallurgy), vol. 35, no. 10,
pp. 1315—1321, 1995.

[183] T. J. Lienert, W. A. Baeslack III, and H. L. Fraser, “Inertia-friction welding of SiC-
reinforced 8009 aluminium,” J. Materials Science, vol. 31, pp. 2149-2157, April
1996.

[184] J. R. Kennedy, “Microstructural observations of are welded boron-aluminum com-
posites,” Welding Journal, vol. 52, no. 3, pp. 1205-1245, 1973.

[185] J. S. Aheam, C. Cooke, and S. G. Fishman, “Fusion welding of SiC-reinforced A1
composites,” Metal Construction, vol. 14, no. 4, pp. 192—197, 1982.

[186] J. H. Devletian, “SiC/Al metal matrix composite welding by a capacitor discharge
process,” Welding Journal, vol. 66, no. 6, pp. 33-39, 1987.

[187] C. D. Lundin, J. C. Danko, and C. J. Swindeman, “Fusion welding of SiC-
reinforced aluminum alloy 2024,” in Recent Trends in Welding Science and Tech-
nology TWR ’89—Proc. of the 2nd International Conference on Trends in Welding

234

Research (S. A. David and J. M. Vrtek, Eds.), pp. 303—307, American Welding Soci-
ety and Welding Research Council, ASM International, Materials Park, Ohio, May
1989.

[188] B. Altshuller, W. Christy, and B. Wiskel, “GMA welding of Al-Al203 metal matrix
composite,” in Weldability of Materials (R. A. Patterson and K W. Mahin, Eds.),
pp. 305—309, The Welding Metallurgy Committee of the Joining Division of ASM,
ASM International, October 1990.

[189] E. I. Kivineva, D. L. Olson, and D. K Matlock, “Particulate-reinforced metal matrix
composite as a weld deposit,” Welding Journal, vol. 74, pp. 833-923, March 1995.

[190] A. Cox, W. A. Baeslack IH, S. Zorko, and C. English, “Capacitor-discharge resis-
tance spot welding of SiC ﬁber-reinforced Ti-6Al-4V,” Welding Journal, vol. 72,
no. 10, pp. 4795—4915, 1993.

[191] T. J. Lienert, E. D. Brandon, and J. C. Lippold, “Laser and electron beam welding
of SiC,, reinforced aluminum A-356 metal matrix composite,” Scripta METALLUR-
GICA et MATERIALIA, vol. 28, no. 11, pp. 1341-1346, 1993.

[192] M. S. Hersh, “Resistance welding of metal matrix composite,” Welding Journal,
vol. 47, no. 9, pp. 4045—4095, 1968;

[193] M. S.Hersh, “Resistance spot welding metal matrix composites,” Welding Journal,
vol. 49, no. 6, pp. 254s-2585, 1970.

[194] M. S. Hersh, “Correlation between boron-aluminum sheet quality and resistance
welding quality and strength,” Welding Journal, vol. 50, no. 12, pp. 5153—521s,
1971.

[195] M. S. Hersh, “The versatility of resistance welding machines for joining alu-
minum/boron composites,” Welding Journal, vol. 51, no. 9, pp. 6268—6328, 1972.

[196] J. E. Gould, “An examination of nugget developments during resistance spot weld-
ing using experimental and analytical techniques,” Welding Journal, vol. 68, no. 1,
pp. ls—lOs, 1987.

[197] N. B. Dahotre, T. D. McCay, and M. H. McCay, “Laser processing of a SiC/Al-alloy
metal matrix composite,” J. Applied Physics, vol. 65, pp. 5072—5077, June 1989.

235

[198] M. J. Cola, T. J. Lienert, J. E. Gould, and J. P. Hurley, “Laser welding of a SiC
particulate reinforced aluminum metal matrix composite,” in Weldability of Materi-
als (R. A. Patterson and K W. Mahin, Eds.), pp. 297—303, The Welding Metallurgy
Committee of the Joining Division of ASM, ASM International, October 1990.

[199] N. B. Dahotre, M. H. McCay, T. D. McCay, C. M. Sharp, and S. Gopinathan, “Laser
welding of SiC/Al-alloy metal matrix composite,” in Laser Materials Processing,
ICALEO’90, vol. SPIE 1601, pp. 343—356, Laser Institute of America and SPIE,
Laser Institute of America, 1990.

[200] N. B. Dahotre, M. H. McCay, T. D. McCay, S. Gopinathan, and L. F. Allard, “Pulse
laser processing of SiC/Al-alloy metal matrix composite,” J. Material Research,
vol. 6, no. 3, pp. 514—529, 1991.

[201] S. Kawali and G. L. Vregelahn, “Attempts at LASER welding of an alumina rein-
forced 6061 aluminum alloy composite,” in Advanced Composites Materials: New
Developments and Applications, Proc. of the Seventh Annual ASM/BSD Advanced
Composites Conference, pp. 283-290, ASM/EDS, ASM International, 1991.

[202] S. M. Kawali, G. L. Vregelahn, and R. Scheucrman, “Laser welding of alumina
reinforced 6061 aluminum alloy composite,” in Proc. Laser Materials Processing
Symp., ICALEO’9I, pp. 156—167, Laser Institute of America, 1991.

[203] K M. Jasim, R. D. Rawlings, R. Sweeney, and D. R. F. West, “Microstructural
changes associated with the laser processing of Al-SiC composites,” J. Materials
Science Letters, vol. 11, pp. 414-417, April 1992.

[204] N. B. Dahotre, T. D. McCay, and M. H. McCay, “Laser induced reaction joining
of metal matrix composite,” in Laser Materials Processing, ICALEO ’93 (P. Denney,
I. Miyamoto, and B. L. Mordike, Eds.), vol. SPIE 2306, pp. 498—506, Laser Institute
of America and SPIE, Laser Institute of America, 1993.

[205] S. Gopinathan, N. B. Dahotre, M. H. McCay, and T. D. McCay, “Carbide formation
during laser welding of SiC/aluminum composites,” in Laser Materials Processing,
ICALEO’93 (P. Denney, I. Miyamoto, and B. L. Mordike, Eds.), vol. SPIE 2306,
pp. 783-793, Laser Institute of America and SPIE, Laser Institute of America, 1993.

[206] A. Hirose, Y. Matsuhiro, M. Kotoh, S. Fukumoto, and K. F. Kobayashi, “Laser-
beam welding of SiC ﬁbre-reinforced Ti-6Al-4V composites,” J. Materials Scienc‘e,
vol. 28, no. 2, pp. 349-355, 1993.

 

 

236

[207] P. A. A. Khan and A. J. Paul, “High speed joining of aluminum metal matrix com-
posites using continuous wave and pulsed lasers,” in Joining and Adhesion of Ad-
vanced Inorganic Materials (A. H.Carim, D. S. Schwartz, and R. S. Silberglitt, Eds.),
vol. 314, pp. 137-142, MRS Symposium Proc., 1993.

[208] N. B. Dahotre, T. D. McCay, and M. H. McCay, “Laser induced liquid phase reaction
synthesis assisted joining of metal matrix composites,” Materials and Manufactur-
ing Processes, vol. 9, no. 3, pp. 447-466, 1994.

[209] P. W. Fucrschbach and M. J. Cieslak, “Restraint effects in laser welding of an alu-
nrinum MMC,” IEEE Transactions on Components, Packaging, and Manufacturing
Technology-Part B: Advanced Packaging, vol. 17, pp. 108—114, February 1994.

[210] T. M. Yue, J. H. Xu, and H. C. Man, “Pulsed Nd-YAG laser welding of a SiC par-
ticulate reinforced aluminium alloy composite,” in Proc. of ICCM-I 0, Metal Matrix
Composites (A. Poursartip, Ed), vol. II, pp. 169—176, The Tenth International Con-
ference on Composite Materials Society, Whistler, B.C., Canada, Woodhead Pub.
Ltd, August 1995.

[211] J. Howard, “Laser machining of polymeric composites for joint preparation,” Mas-
ter’s thesis, Michigan State University, 1993.

[212] C. L. Cline, “An analytical and experimental study of diffusion bonding,” Welding
Journal, vol. 45, pp. 4818-4898, Nov 1966.

[213] M. F. Gittos and P. L. Threadgill, “Preliminary studies of joining particulate rein-
forced aluminium alloy metal matrix composites,” in Metal Matrix Composites III:
Exploiting the Investment, Institute of Materials, London, 1991.

[214] J. R. Kennedy, “Fusion welding of titanium-tungsten and titanium-graphite compos-
ites,” Welding Journal, vol. 51, pp. 250s-259s, May 1972.

[215] D. T. Swift-Hook and A. E. F.Gick, “Penetration welding with lasers,” Welding Jour-
nal, vol. 52, no. 11, pp. 492s—499s, 1973.

[216] P. A. Molian and T. S. Srivatsan, “Laser welding of high modulus low density Al-Li
alloys,” in Laser Materials Processing, Proc. of the 7th International Congress on
Applications of Lasers and Electrooptics ICALEO ‘88 (G. Bruck, Ed), pp. 238-248,
1988.

 

237

[217] S. Katayama, C. D. Lundin, J. C. Danko, and T. D. McKay, “Laser weldability of
aluminum alloys,” in Recent Trends in Welding Science and Technology TWR '89-
Proc. of the 2nd International Conference on Trends in Welding Research (S. A.
David and J. M. Vrtck, Eds.), pp. 687—691, American Welding Society and Welding
Research Council, ASM international, Materials park, Ohio, May 1989.

[218] E. Kannatey-Asibu, Jr, “Thermal aspects of the split-beam laser welding concept,”
in Modeling of Machine Tools: Accuracy, Dynamics, and Control (P. M. Ferreia,
S. G. Kappor, and A. C. Y. Wang, Eds.), vol. PED 45, pp. 175-190, ASME, 1990.

[219] J. R. Davis, Ed, ASM Specialty Handbook - Aluminum and Aluminum Alloys,
ch. Aluminum-Matrix Composites, pp. 160—179. ASM International, 1994.

[220] ASM International Handbook Committee, ASM Handbook - Heat Treating of Alu-
minum Alloys, vol. 4, pp. 841—879. ASM International, 1991.

[221] R. B. C. Cayless, Metals Handbook: Properties and selection - nonferrous alloys
and special-purpose materials, vol. 2, ch. Alloy and temper designation system for
Aluminum and Aluminum alloys, pp. 15-28. ASM International, tenth ed, 1991.

[222] Duralcan, San Diego, CA, Duralcan Metal-Matrix Composites Data Package, 1993. '

[223] V. P. Garashchuk, V. I. Kirscy, and E. O. Paton, “Effect of polarizatiOn of continuous
C02 laser radiation on the weld formation,” in Power Beam Processing - Electron,
Laser, Plasma-arc (E. A. Metzbower and D. Hauscr, Eds.), pp. 185—189, May 1988.

[224] E. J. Daniclewicz, G. H. Sherman, and P. Muys, “High-power C02 lasers drive new
optics solutions,” Laser Focus World, pp. 113—119, January 1992.

[225] J. Kim, “Parametric evaluations of C02 laser beam welding,” in Welding and Joining
Processes (Elijah Kannatey-Asibu, Jr, H. S. Cho, and S. Fukuda, Eds.), vol. PED-51,
pp. 195-201, ASME, December 1991.

[226] O. Kubaschewski and ELL. Evans, Metallurgical Thermochemistry. New York:
John Wiley & Sons, 1956.

[227] R. Hultgrcn, R. L. Orr, P. D. Anderson, and K. K. Kelley, Selected values of thermo-
dynamic properties of metals and alloys. New York: John Wiley & Sons, 1963.

[228] J. E. Harry and F. W. Lunau, “Machining and welding,” IEEE J. Quantum Electron-
ics, vol. 7, no. 6, pp. 276—280, 1971.

 

 

238

[229] W. M. Steen and J. N. Kamalu, “Laser cutting,” in Laser Materials Processing
(M. Bass, Ed), ch. 2, pp. 15—111, North-Holland Pub. Co., 1983.

[230] Y. Arata, Plasma, Electron & Laser Beam Technology: Development and use in
Materials Processing, ch. Fundamental characteristics of ultra high energy density
heat source, pp. 234-244. ASM, 1986.

[231] P. A. Molian and M. Baldwin, “Combustion-assisted laser cutting of a difﬁcult-to-
machinc superalloy,” J. Laser Applications, vol. 4, no. 1, pp. 9—14, 1992.

[232] N. Taniguchi, Energy-Beam Processing of Materials. Oxford: Clarendon, 1989.

[233] G. Chryssolouris, J. Brcdt, S. Kordas, and E. Wilson, “Theoretical aspects of a laser
machine tool,” J. Engineering Industry, vol. 110, pp. 65-70, February 1988.

[234] G. Simon and U. Gratzke, “Theoretical investigations of instabilities in laser gas
cutting,” in High Power Lasers and Laser Machining Technology (M. L. Gaillard
and A. Quenzcr, Eds.), pp. 204—210, Proc. SPIE 1132, The International Society for
Optical Engineering, 1989.

[235] T. H. Kim, “Porosity formation in laser-beam materials processing,” J. Materials
Science Letters, vol. 10, pp. 400—402, April 1991.

[236] W. Tuttlc, “Understanding aluminum welding,” Welding Journal, vol. 70, pp. 43-46,
February 1991.

[237] M. J. Dvomak, R. H. Frost, and D. L. Olson, “Inﬂuence of solidiﬁcation kinetics on
aluminum weld grain reﬁnement,” Welding Journal, vol. 70, pp. 2718-2768, October
1991.

[238] P. J. Schneider, Conduction Heat Transfer. Reading: Addison-Wesley Pub. Co, 2 ed,
1982.

[239] M. N. Ozisik, Heat Conduction. New York: John Wiley & Sons, 2 ed, 1993.

[240] J. Mazumder, “Laser welding,” in Laser Materials Processing (M. Bass, Ed), ch. 3,
pp. 113—200, North-Holland Pub. Co., 1983.