Laser Forming and Welding Processes

Materials Forming, Machining and Tribology
Bekir Sami Yilbas

Sohail Akhtar
Shahzada Zaman Shuja
Laser Forming and

Welding Processes
Materials Forming, Machining and Tribology
Series Editor
J. Paulo Davim
For further volumes:

http://www.springer.com/series/11181
Bekir Sami Yilbas Sohail Akhtar
•
Laser Forming and

Welding Processes
123
Bekir Sami Yilbas
Sohail Akhtar
Mechanical Engineering Department
King Fahd University of Petroleum and Minerals
Dhahran
Saudi Arabia
ISSN 2195-0911 ISSN 2195-092X (electronic)

ISBN 978-3-319-00980-3 ISBN 978-3-319-00981-0 (eBook)
DOI 10.1007/978-3-319-00981-0
Springer Cham Heidelberg New York Dordrecht London
Library of Congress Control Number: 2013940956
Ó Springer International Publishing Switzerland 2013

This work is subject to copyright. All rights are reserved by the Publisher, whether the whole or part of
the material is concerned, specifically the rights of translation, reprinting, reuse of illustrations,
recitation, broadcasting, reproduction on microfilms or in any other physical way, and transmission or
information storage and retrieval, electronic adaptation, computer software, or by similar or dissimilar
methodology now known or hereafter developed. Exempted from this legal reservation are brief
excerpts in connection with reviews or scholarly analysis or material supplied specifically for the
purpose of being entered and executed on a computer system, for exclusive use by the purchaser of the
work. Duplication of this publication or parts thereof is permitted only under the provisions of
the Copyright Law of the Publisher’s location, in its current version, and permission for use must
always be obtained from Springer. Permissions for use may be obtained through RightsLink at the
Copyright Clearance Center. Violations are liable to prosecution under the respective Copyright Law.
The use of general descriptive names, registered names, trademarks, service marks, etc. in this
publication does not imply, even in the absence of a specific statement, that such names are exempt
from the relevant protective laws and regulations and therefore free for general use.
While the advice and information in this book are believed to be true and accurate at the date of
publication, neither the authors nor the editors nor the publisher can accept any legal responsibility for
any errors or omissions that may be made. The publisher makes no warranty, express or implied, with
respect to the material contained herein.
Printed on acid-free paper
Springer is part of Springer Science+Business Media (www.springer.com)

Preface
Laser forming and welding processes can meet the industry demands associated
with advantages such as precision of operation, low cost, and local treatment.
Although the process is involved with multi-physics phenomena, which is com-
plicated in nature, the proper selection of the process parameters minimizes the
process complications and provides high end product quality. High laser power
requirements of the process trigger the phase change and high temperature gra-
dients forming in the irradiated region. This is unavoidable because of the
requirement for deep penetration of the laser beam into the substrate material to
achieve improved case depths during the process. High power laser heating
involves solid phase heating, melting, and subsequent evaporation at the surface.
Since the laser heating process is fast, the formation of high temperature gradients
in the neighborhood of the heated region is inescapable and it becomes crucial for
high power laser processing. The formation of high stress levels due to the pres-
ence of high temperature gradients causes the failure of the process through crack
formation in the irradiated region. In addition, the defect sites formed in the
workpiece, due to excessive heating, limit the practical application of the laser
processing due to resulting low end product quality. To achieve the required case
depth in the substrate material while avoiding the defect sites due to excessive
heating, the model study comprehending the heating process becomes necessary.
In addition, the model study provides physical insight into the process and gives
information about the influence of process parameters on the end product quality.
Although the analytical treatment of the heating process involves some useful
assumptions to minimize the mathematical complications, the findings are
important to correlate the dependent variables with the independent process
parameters. However, the number of assumptions may reduce for the numerical
simulation of the heating process at the expense of considerable computational
efforts. On the other hand, the predictions of numerical simulations and the result
of analytical solutions are not sufficient to make a precise assessment of the
process. Consequently, experimental investigation of the laser heating process
becomes essential to validate the predictions for sound assessments, despite the
high cost and difficulties faced during experiments.
In this book, laser forming and welding processes are presented in detail. The
analytical and numerical approaches for the predictions of temperature and stress
v
vi Preface
fields in the irradiated region are presented during the heating and cooling cycles
of the process. The thermal efficiency of the welding process is also included to
extend the thermodynamics analysis associated with the process. Experimental
findings and observations are included to explore the metallurgical and mechanical
characterizations of the regions treated by using laser processing pertinent to
forming and welding of the metallic substrates. Although this book covers most of
the aspects of forming and welding processes, it is not claimed to be complete and
is open for further improvements in the next editions.
Acknowledgments
We would like to acknowledge to role of King Fahd University of Petroleum and

Minerals in extending strong support from beginning to end facilitating every
means during the preparation of the book. The author wishes to thank the
colleagues who contributed to the work presented in the book through previous
cooperation of the authors. In particular, thanks to Dr. Muammer Kalyon,
Dr. Nasser Al-Aqeeli, Dr. Saad Bin Mansoor, Dr. A. F. M. Arif, and all our
graduate students.
vii
Contents
1 Introduction to Laser Forming and Welding Processes . . . . . . . . . 1

References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 4
2 Analytical Treatment of Laser Forming

and Welding Processes . . . . . . . . . . . . . . . . . . . . . . . . ......... 5
2.1 Analytical Formulation of Heating and Stress Field
Involving Melting and Solid Heating . . . . . . . . . . . ......... 5
Involving Evaporation and Solidification . . . . . . . . . . . . . . . . . 10
2.3 Findings and Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 32
2.3.1 Stationary Source Consideration . . . . . . . . . . . . . . . . . . 33
2.3.2 Moving Source Consideration. . . . . . . . . . . . . . . . . . . . 36
2.4 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 37
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 38
3 Numerical Analysis for Laser Forming and Welding . . . . . . . . . . . 39

3.1 Numerical Analysis for Laser Forming Process . . . . . . . . . . . . . 39
3.1.1 Solid Heating and Stress Field Formation . . . . . . . . . . . 40
3.1.2 Phase Change and Stress Field Formation . . . . . . . . . . . 47
3.2 Numerical Analysis for Laser Welding Process . . . . . . . . . . . . . 49
3.2.1 Phase Change Including Melting and Evaporation. . . . . . 49
3.2.2 Transiently Developing Vapor Jet . . . . . . . . . . . . . . . . . 52
3.2.3 Numerical Solution . . . . . . . . . . . . . . . . . . . . . . . . . . . 58
3.2.4 Analysis for Laser Welding . . . . . . . . . . . . . . . . . . . . . 59
3.3 Efficiency Analysis of Welding. . . . . . . . . . . . . . . . . . . . . . . . 64
3.3.1 First Law (Energy) Efficiency . . . . . . . . . . . . . . . . . . . 66
3.3.2 Second Law (Exergy) Efficiency . . . . . . . . . . . . . . . . . 67
3.4.1 Phase Change Process Including Surface Evaporation . . . 68
3.4.2 Laser Multi Spot Heating of Thin Metal Sheet . . . . . . . . 71
3.4.3 Three-Dimensional Heating and Stress Fields . . . . . . . . . 72
ix
x Contents
3.4.4 Temperature and Stress Fields for Welding

and Forming Process . . .............. . . . . . . . . . . 75
3.4.5 Welding Efficiency . . . .............. . . . . . . . . . . 78
3.5 Concluding Remarks . . . . . . . .............. . . . . . . . . . . 81
References . . . . . . . . . . . . . . . . . . .............. . . . . . . . . . . 83
4 Experimental Analysis for Laser Forming and Welding . . . . . . . . 85

4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 85
4.2 Lasers Used for Welding and Forming . . . . . . . . . . . . . . . . . . . 85
4.3 Analytical Tools for Microstructural Analysis . . . . . . . . . . . . . . 87
4.4 Mechanical Characterisation of Laser Welded
and Bended Sections . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 88
4.5.1 Bending Angle Measurements, Morphology
and Microstructure for Laser Forming . . . . . . . . . . . . . . 90
4.5.2 Morphology, Microstructure and Mechanical Properties
for Laser Welding Process . . . . . . . . . . . . . . . . . . . . . . 95
4.5.3 Laser Welding of Hayness 188 Alloy . . . . . . . . . . . . . . 98
4.5.4 Welding of AISI 316 Steel. . . . . . . . . . . . . . . . . . . . . . 100
4.6 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 104
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 105
Chapter 1
Introduction to Laser Forming
and Welding Processes
Lasers are widely used in industry as thermal processing tool due to its precision of
operation, local treatment, fast processing, and low cost. One of the laser thermal
processing methods is the laser forming and welding of sheet metals. The process
is involved with high intensity beam interaction with the solid surface while
resulting in phase change in the irradiated region. Although laser forming differs
slightly from the welding, the process is almost the same in terms of heating and
microstructure formation. However, laser forming requires shallow depth of phase
change at the surface while laser welding requires deep penetration of the laser
beam into the substrate material. From the heating point of view, the irradiated
surface undergoes solid heating and melting following the solidification for
forming process. In the case of welding, the phase change may extend to include
the evaporation at the surface. This is particular important for deep penetration
welding of thick substrate materials. As the laser power intensity increases, the
laser power absorbed in the skin of the surface acts like a heat source resulting in
extensive temperature rise at the surface. In some cases, temperature rise exceeds
the evaporation temperature of the substrate material causing the evaporation of
the surface. As the surface evaporates, the melted substrate surface recesses
towards the solid bulk. This, in turn, results in cavity formation at the surface.
Although the depth of cavity, in general, is shallow, it increases the surface
roughness and reduces the weld cross-section. However, the surface evaporation is
unavoidable for deep penetration welding of metallic parts. On the other hand,
since laser heating and phase change process is almost rapid, the high cooling rates
is resulted during the solidification of the melted regions. This causes attainment of
high temperature gradients in the phase change zones increasing the thermal strain
in these regions. Hence, thermal stress formation becomes unavoidable in the
welded region. Depending upon the magnitude and the form of the residual stress,
compressive or tensile, generated in the irradiated region, thermally induced
bending takes place in the substrate. In the case of welding, high magnitude of
residual stress, in some cases, forms cracks in the welded region. Consequently,
a care must e taken to reduce the residual stress levels in the weld section. Since
B. S. Yilbas et al., Laser Forming and Welding Processes, 1

Materials Forming, Machining and Tribology, DOI: 10.1007/978-3-319-00981-0_1,
2 1 Introduction to Laser Forming and Welding Processes
the laser heating and cooling processes are rapid, the size of the heat affected zone
becomes narrow. This provides advantages of the local treatment by a laser beam.
Laser-induced forming is a non-contact technique of forming metallic com-
ponents. The process is used for producing bending, spatial forming and alignment
of metallic and nonmetallic components through the controlled application of laser
energy [1]. In general, there is no mechanical external loading involved in the laser
forming process. Hence, it is only the temperature distribution that controls the
deformation of the plate. Therefore, it is very important to understand how various
processing parameters affect the transfer of heat generated by the laser beam. The
laser forming process is accomplished by setting up thermal stresses into a
work-piece by irradiation with a laser beam, thus inducing rapid localized heating
followed by cooling as the laser energy is switched. During the heating cycle, if
the thermal strains in the irradiated zone exceed the elastic strains that can be
carried by the material, which are temperature and geometry dependent, then these
will be converted into plastic compressive strains. On cooling, the irradiated
material will undergo shrinkage, leading to the development of bending or a
change of shape of the work-piece at the irradiated region. Laser forming has
potential applications in rapid prototyping and shape alteration in the aerospace,
marine and automotive industries. Also several micro-scale adjustment techniques
are under development, establishing a fresh approach to the design of products so
that accuracy adjustments can be made at the final stages of assembly [1].
Laser-induced forming involves a complex interaction of many process
parameters, ranging from those connected with the laser beam irradiation (such as
power density, wavelength, temporal energy variations, speed of motion relative to
the work-piece) to those thermal and mechanical properties of the processed
material (such as absorption coefficient, thermal conductivity etc.). Geiger and
Vollertsen [2] explained three important mechanisms of thermo-mechanical
behavior related to laser forming, each associated with specific combinations of
component geometries and laser process parameters. These include the tempera-
ture gradient mechanism (TGM), the buckling mechanism (BM) and the short-
ening or upsetting mechanism (UM). The TGM is dominated if the laser beam
diameter is on the order of the sheet metal thickness. The temperature generated by
the laser beam is very high in the surface being scanned and the temperature decay
is very quick with the distance from the scanning point. Consequently, at a place
far away from the scan point the temperature is very low, which, in most cases, is
almost not affected. Thus, the temperature gradient is usually not a constant along
the thickness. This means that, during the process of laser forming, stresses and the
corresponding stress-induced strains will be generated. Therefore the deformation
of the sheet is the result of both the thermal and stress-induced strains. During
heating, the sheet is bending away from the laser beam because of the high thermal
expansion on the top surface where the sheet is scanned. The non-uniform dis-
tribution of the thermal strain gradient creates a complicated distribution of stress,
which bends the sheet in an opposite direction. With the increase of temperature,
the initially elastic bending induced by the stress becomes plastic. While during
cooling, the sheet is bending towards the laser beam because of the thermal
1 Introduction to Laser Forming and Welding Processes 3
contraction at the top surface. Similar to the heating, stress is generated during
cooling. However, because the stress is decreased during cooling, the plastic
deformation generated during cooling is less than that generated during the
heating. Consequently, the final bending of the sheet is towards to the laser beam
[3]. The BM allows both bending toward and opposite to the incident laser beam
and acts when the irradiated surface is large in comparison to the material
thickness. Thereby, the whole cross-section of the sheet will be heated so that the
temperature gradient between top surface and bottom surface is small. The induced
thermal expansion will be hindered by the surrounding material which causes the
buckling of the material in the irradiated area. In the transition region between the
buckled material and the flat material outside of the irradiated area, there is a larger
plastic deformation, because of high temperature and low yield stress. After
irradiation the material is cooling down associated with relaxation of the elastic
deformation and generation of bending. The UM involves similar parameters to
that in the BM, but occurs when the geometry of the part is structurally ‘stiffer’ and
thus buckling cannot take place. Instead, a shortening and simultaneous upsetting
(or thickening) of the irradiated zone occurs [3].
Yu et al. [4] presented a thermo-mechanical model for the prediction of angular
deformations of metal plates due to laser line heating. Their model employed a
semi-analytically determined temperature distribution, which incorporated the
effects of heat loss and a distributed moving heat source, to calculate the
dimensions of a critical heat-affected region. They used dimensions of this region
to find the angular deformation by an analytic solution method. Shen et al. [5]
derived a formula of bending angle in laser forming based on the assumptions that
the plastic deformation is generated only during heating, and during cooling only
the elastic deformation occurs. In this formula, the yield strength reduction factor
due to the temperature increase and the characteristic depth of plastic zone are
involved. Mucha [6] presented an analytical model of laser plate bending based on
temperature gradient, in which the restrain rigidity coefficient was defined. The
model gave the solutions for longitudinal and transversal angle deformation.
Shidid et al. [7] studied the effects of inert gas shielding on laser bending of
titanium sheets. They used different gas flow conditions, nozzle positions and inert
gas combinations to enhance the bend quality and bend angle. They found that
these process changes lead to final bending angle increase and decrease in width of
Heat Affected Zone (HAZ), section thickness of sheet at HAZ and surface hard-
ness. They also used different coatings on Grade-2 Titanium, which is highly
reflective, study the improvement in the absorption of laser beam, which resulted
in further increase in bending angle.
Due to the complexity of analytically modeling laser forming processes where
the work-piece temperatures, dimensions and properties are varying both in time
and space and which depend on many variables, the numerical approach is often
more beneficial for modeling these situations. In recent years, the computational
efficiency has made thermo-mechanical studies more practical. Shen et al. [8]
presented a finite element model for the heat transfer problem of laser forming of
plates. Their results showed that the influence of the heat exchange through
4 1 Introduction to Laser Forming and Welding Processes
radiation on boundaries on the temperature distribution can be insignificant and a

uniform temperature gradient can be accomplished by using the acceleration
scanning scheme. Hsiao et al. [9] employed ABAQUS code to model the laser
bending process. They used the model in their work to emphasize the importance
of the specimen size. Their results showed that the angular distortion obtained on a
short specimen is much smaller than for a long specimen. Shen et al. [10] studied
the edge effects of laser bending using varying scan speed by incorporating a
numerical model. Since the laser forming process is time dependent, the finite
element analysis model involves the discretisation in space and in time. The
variation of the scan velocity thus was implemented in the model by specifying the
dwell times on corresponding elements. Later, they [11] proposed a total of seven
varying velocity scanning schemes to reduce the edge effects, and the results
showed that the combination of acceleration and deceleration scanning scheme can
minimize the edge effects. Cheng et al. [12] developed a process design approach
for laser forming of doubly curved thin plates. They showed that strain field
determination via large-deformation elastic FEM is valid and effective. They also
shown theoretically and validated FEM that the strain field determination for a
desired shape can be achieved independent of any material properties.
References
1. Dearden G, Edwardson SP (2003) Some recent developments in two-and three-dimensional

laser forming for ‘macro’ and ‘micro’ applications. J Opt A: Pure Appl Opt 5:S8–S15
2. Geiger M, Vollertsen F (1993) The mechanism of laser forming. CIRP Ann 42:301–304
3. Shen H, Vollertsen F (2009) Modeling of laser forming—an review. Comput Mater Sci
46:834–840
4. Yu G, Anderson RJ, Maekawa T, Patrikalakis NM (2001) Efficient simulation of shell
forming by line heating. Int J Mech Sci 43:2349–2370
5. Shen H, Yao ZQ, Shi YJ, Hu J (2006) An analytical formula for estimating the bending angle
by laser forming. Proceedings of the institution of mechanical engineers, Part C. J Mech Eng
Sci 220(2):243–247
6. Mucha Z (2007) In: Proceedings of SPIE 6598, 65980 M.1–65980 M.9
7. Shidid DP, Gollo MH, Brandt M, Mahdavian M (2013) Study of effect of process parameters
on titanium sheet metal bending using Nd: YAG laser. Opt Laser Technol 47:242–247
8. Shen H, Yao ZQ, Shi YJ, Hu J (2007) The simulation of temperature field in the laser forming
of steel plates. Int J Model Identif Control 2(3):241–249
9. Hsiao YC, Shimizu H, Firth L, Maher W, Masabuchi K (1997) Finite element modeling of
laser forming. In: Proceedings of the international congress on applications of lasers and
electro-optics (ICALEO’97), section A, pp 31–40
10. Shen H, Zhou J, Shi YJ, Yao ZQ, Hu J (2007) Varying velocity scan in laser forming of
plates. Mater Sci Technol 23(4):483–486
11. Shen H, Yao ZQ (2008) Analysis of varying velocity scanning schemes on bending angle in
laser forming. In: International workshop on thermal forming and welding distortion,
pp 215–227
12. Cheng J, Yao YL (2004) Process design of laser forming for three-dimensional thin plates.
ASME J Manuf Sci Eng 126:217–225
Chapter 2
Analytical Treatment of Laser Forming
and Welding Processes
The analytical treatment of laser forming and welding is categorized into two
groups. In the first group, laser heating and melting is presented and in the second
group laser solid phase heating, melting, and phase change is included. Due to the
difficulties associated with the surface evaporation, stationary source consideration
will be incorporated in the second group. The study is extended to include the
analytical treatment of the thermal stress development in the laser irradiated
region.

Involving Melting and Solid Heating
The analytical treatment of laser heating of sheet metal surfaces can extend to include
the moving heat source consideration. Since the heating problem involves with
melting and the heating time is much longer than the thermalization time of the
substrate material, the Fourier heating law can be used to formulate the heating
situation. The heating process is formulated below in line with the previous studies [1].
The transient heat transfer equation, which is appropriate to a line heat source
moving at a constant speed v along x-direction on a thin plate surface (Fig. 2.1),
can be written as:
o2 T o2 T o2 T oT
þ þ ¼ 2k ð2:1Þ
ox2 oy2 oz2 ot
where
qCp
2k ¼ ð2:2Þ
k

6 2 Analytical Treatment of Laser Forming
Fig. 2.1 Schematic view of

laser heat source
Consider the variable n ¼ x vt and introducing into Eq. 2.1 results in:
o2 T o2 T o2 T oT oT
2
þ 2 þ 2 2kv þ 2k ¼0 ð2:3Þ
on oy oz on ot
If a solid is of sufficient length, temperature distribution around the heat source
becomes constant with time, i.e. oT ot 0. Equation 2.3, then, becomes:
o2 T o 2 T o2 T oT
2
þ 2 þ 2 ¼ 2kv ð2:4Þ
on oy oz on
Let the solution of Eq. 2.4 is:
T ¼ eðkvnÞ Fðn; y; zÞ ð2:5Þ

Substituting Eq. 2.5 into Eq. 2.4 results in:
r2 F ðkvÞ2 F ¼ 0 ð2:6Þ
Since the thickness of the plate is small and it is, therefore, acceptable to
consider no heat flow in the z-direction (i.e. temperature in the z-direction does not
alter), then:
oT
¼0 ð2:7Þ
oz
The relevant boundary conditions for Eq. (2.4) are as follows.
At infinitely long distance away from the laser source, no heat transfer takes
place, i.e. temperature gradient in both n and y directions become zero, i.e.:
oT
¼0 at n ¼ 1 ð2:8Þ
on
and
oT
¼0 at y ¼ 1 ð2:9Þ
oy
2.1 Analytical Formulation of Heating and Stress 7
Now consider the cylindrical coordinates. The heat source with its radius r can
be written as:
oT
Q ¼ k 2pr ð2:10Þ
or
where Q is the rate of heat per unit length of source.
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
Moreover, r ¼ n2 þ y2 , then Eq. 2.6 becomes:
o2 F 1 oF
þ ðkvÞ2 F ¼ 0 ð2:11Þ
or 2 r or
The solution of Eq. 2.11 satisfies the Bessel’s function of the second kind, zero
order i.e.:
T ¼ AeðkvnÞ Ko ½kvr ð2:12Þ

when r approaches to 0 then:
oKo ½kvr 1
¼ ð2:13Þ
or r
Hence the solution with condition at Q ¼ k oT
or 2pr is satisfied for:
Q ðkvnÞ
T¼ e Ko ½kvr ð2:14Þ
2pk
where k is the thermal conductivity.
Equation 2.14 is the solution of the temperature distribution in an infinite solid
due to quantity of heat Q instantaneously generated at t = 0 at a point (x0 , y0 , z0 ).
However, Eq. 2.14 can be reduced to an instantaneous source of heat in a sub-
stance lying in the plane z = 0. In this case, in the absence of radiation heat loss
from the surface, Eq. 2.14 yields:
2
Q x þ y2
T¼ exp ð2:15Þ
4pkDt 4at
where D is the melt depth, which is equal to the workpiece thickness.
When the substance lies in the x–y plane (z = 0) and moves with a velocity u in
the direction of x axis, then, the steady temperature at the point x–y due to heat flux
q per unit time at the region of origin can be written as:
q ux h ur i
T¼ exp Ko ð2:16Þ
2pkD 2a 2a
where r is the radial distance to the point being considered (satisfying
r2 = x2 ? y2), k is the thermal conductivity, D is the thickness of the workpiece,
and Ko is the modified Bessel function of second kind zero order. When laser
heating is considered, the heat flux per unit time becomes:
q ¼ ð1 Rf ÞIo ð2:17Þ
where Rf is the surface reflectivity and Io is the incident laser power .
As temperature reaches the melting temperature, melting isotherm develops at
the surface. The temperature of the melting isotherm can be expressed as:
ð1 Rf ÞIo ux h ur i
Tm ¼ exp Ko ð2:18Þ
2pkD 2a 2a
Introducing dimensionless coordinates as:
ux uy ur
X¼ ; Y¼ ; R¼ ð2:19Þ
2a 2a 2a
Equation 2.18 becomes:
ð1 Rf ÞIo
Tm ¼ expð X ÞKo ½R ð2:20Þ
2pkD
To determine the maximum melt width, re-arrangement of Eq. 2.20 is neces-
sary, i.e. it yields:
2pkDTm 1
expðXÞ ¼ ð2:21Þ
ð1 Rf ÞIo Ko ½R
or
hpffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffii 2pkDTm
Ko X2 þ Y 2 ¼ expðXÞ ð2:22Þ
ð1 Rf ÞIo
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
where R ¼ ðX 2 þ Y 2 Þ.
Differentiating with respect to X gives:
hpffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffii dY 2

1=2 2pkDTm
K1 X2 þ Y 2 X þ Y X þ Y2 ¼ expðXÞ ð2:23Þ
dX ð1 Rf ÞIo
where dY/dX = 0 then:
XK1 ½R 2pkDTm
¼ expðXÞ ð2:24Þ
R ð1 Rf ÞIo
where K1[R] is the modified Bessel function of second kind first order. Using
Eq. 2.24, it results:
XK1 ½R
¼ Ko ½R ð2:25Þ
R
Therefore, for maximum Y:
RKo ½R
X¼ ð2:26Þ
K1 ½R
and

2pkDTm RKo ½R
Ko ½R ¼ exp ð2:27Þ
ð1 Rf ÞIo K1 ½R
Solution of Eq. 2.27 gives the values of R for melt at the point of maximum a
width. Y can be found from:
Y ¼ R sin u ð2:28Þ
where

Ko ½R
1
u ¼ cos ð2:29Þ
K1 ½R
A computer program was developed to compute the relevant melting
parameters.
The stress analysis associated with the laser forming process can be briskly
described in line with the previous study [2]. Laser bending angle can be for-
mulated for the plane sheet metal after considering the equivalent nodal forces and
the change of the momentum of the plate during the bending process due to laser
heating. Consider the plane sheet, which undergoes bending due to the thermal
load. In line with the previous study [2], the transverse bending moment per unit
length can be written as:
Z
dh=2
My ¼ ry zdz ð2:30Þ
h=2
where ry ¼ EeðzÞ. The solution of Eq. 2.30 is straight forward and it yields:
!
ðxÞhðxÞ dðxÞ2
My ¼ E emax ð2:31Þ
8 3
The transverse shrinkage force per unit length is:
Z
dh=2
fy ¼ ry dz ð2:32Þ
h=2
The solution of Eq. 2.32 yields:

p E emax
fy ¼ dð xÞ ð2:33Þ
4
The variables in Eqs. 2.31–2.33 are d is the depth of heat affected zone (HAZ).
emax is the maximum plastic strain occurring at the heated surface
(emax ¼ ath Tmax ry =E, where ath is thermal expansion coefficient, Tmax is the
maximum temperature increase, E is the elastic modules, and ry is the yield stress).
Since the bending angle varies along the x-axis, the bending curvature at a
location with h(x) thickness can be used to calculate the bending angle, i.e.:
1 My 12ð1 m2 Þ
¼ ¼ My ð2:34Þ
q D EhðxÞ3
where the bending rigidity per unit length (D) is:
ZhðxÞ
E EhðxÞ3
D¼ z2 dz ¼ ð2:35Þ
1 m2 12ð1 m2 Þ
0
Therefore, the bending angle (aB) of the plate subjected to a laser heating can be
written as [2]:

3dp 4d2
aB ¼ bð1 m2 Þemax ð2:36Þ
2h2 h3
where b is the width of laser scanned area at the workpiece surface, m is the
Poisson’s ratio, emax is the maximum plastic strain (emax ¼ ath Tmax r=E, where
ath is the thermal expansion coefficient, Tmax is the maximum temperature, which
can be obtained from Eq. 2.16, r is the yield stress, and E is the elastic modulus),
d is the depth of heat affected zone, h is the thickness of the plate. Equation 2.36
can be used to predict the bending angle.

Involving Evaporation and Solidification
Laser deep penetration welding involves with surface evaporation. In this case,
material subjected to the laser irradiation evaporates at velocity equals to the vapor
front velocity. However, the melt surface recesses towards the solid bulk with a
recession velocity, which is much lower than the vapor front velocity. Since the
heating situation is complex, the analytical solution to evaporation process with a
moving heat source becomes difficult. Consequently, surface evaporation is treated
here after considering the stationary volumetric heat source in line with the
previous study [3].
The transient heat transfer equation pertinent to the evaporation at the surface
including the volumetric heat source term can be written as:
o2 T oT oT
k 2
þ qCp V þ I0 d expðdxÞ ¼ qCp ð2:37Þ
ox ox ot
where V is the vapor front velocity and can be written as [4]:

kB Ts 1=2 L
V¼ exp ð2:38Þ
2pm kB Ts
It can be observed that the transient heating problem is non-linear, since the
velocity is surface temperature (Ts) dependent. Consequently, closed form solution
for Eq. 2.37 is extremely difficult; however, the quasi-steady solution can be
possible. Since the evaporation takes place at the surface, energy dissipated via
evaporation can be considered as the boundary condition at the surface. As the
depth below the surface increases, temperature is assumed to reduce the initial
temperature, which is 0. In addition, initially the substrate material is assumed to
be at uniform constant temperature. Therefore, the boundary and initial conditions
for Eq. 2.37 are:

oT
k ¼ qVL; Tð1; tÞ ¼ 0 and Tðx; 0Þ ¼ 0: ð2:39Þ
ox x¼0
The solution of Eq. 2.37 with the appropriate boundary conditions can be obtain
using a Laplace transformation method. In this case, the Laplace transformation of
Eq. 2.37 yields:
d 2 Tðx; pÞ V dTðx; pÞ p I0 d
2
þ Tðx; pÞ ¼ expðdxÞ ð2:40Þ
dx a dx a kp
The Laplace transformation of boundary and initial conditions becomes:

dTðx; pÞ qVL
Tðx; 0Þ ¼ 0; ¼ ; Tð1; pÞ ¼ 0 ð2:41Þ
dx x¼0 kp
where p is the Laplace transform variable.
The solution to Eq. 2.40 yields the result in the Laplace domain:
pffiffiffiffiffiffiffiffiffiffiffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffi
x x
Tðx; pÞ ¼A exp pffiffiffi ðb b2 þ pÞ þ B exp pffiffiffi ðb þ b2 þ pÞ
a a

I0 da expðdxÞ
ð2:42Þ
kp c2 ðb2 þ pÞ
pffiffiffi
where b ¼ 2pV ffiffia, c ¼ b d a and A and B are constants of integration. Using the
boundary condition (Eq. 2.41), it becomes:
A¼0 ð2:43Þ
The substitution of boundary condition into Eq. 2.42 results:
pffiffiffi
a I0 ad2 qVL
B¼ pffiffiffiffiffiffiffiffiffiffiffiffiffi ð2:44Þ
b þ b2 þ p kpðc2 b2 pÞ kp
Therefore, the complete solution to the transformed equation (Eq. 2.42)

becomes:
pffiffiffi
a I0 ad2 qVL
Tðx; pÞ ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffi
b þ b2 þ p kpðc2 b2 pÞ kp
p ffiffiffiffiffiffiffiffiffiffiffiffiffi
x I0 da expðdxÞ
exp pffiffiffi ðb þ b2 þ pÞ ð2:45Þ
a kp c2 ðb2 þ pÞ
The difficulty arises due to the first term in Eq. 2.45 during the inversion of the
solution in the transformed plane. However, the problem may be tacked after
writing the first term in the indefinite integral form, hence:
Zx pffiffiffiffiffiffiffiffiffiffiffiffiffi
I0 ad2 qVL x
f ðx; pÞ ¼ exp pffiffiffi ðb þ b2 þ pÞ dx ð2:46Þ
kpðc2 b2 pÞ kp a
0
or
Z x
f ðx; pÞ ¼ gðx; pÞdx ð2:47Þ
0
The inverse transformation of f(x, p) function may be carried out in the fol-
lowing manner:
Z x Z x
L1 ½f ðx; pÞ ¼ L1 gðx; pÞdx ¼ L1 ½gðx; pÞdx ð2:48Þ
0 0
1
where L is the inverse Laplace transformation. The result can be obtained
after introducing the method of expansion into partial fractions. Using the
relationship:
L1 ½uðp þ aÞ ¼ eat L1 ½uðpÞ ð2:49Þ

it yields:
"pffiffiffi #
a I0 ad2 x pffiffiffiffiffiffiffiffiffiffiffiffiffi
1
L pffiffiffiffiffiffiffiffiffiffiffiffiffi exp pffiffiffi ðb þ b þ pÞ
2
b þ b2 þ p kpðc2 b2 pÞ a
2
bx I0 ad expðqxÞ
¼ exp pffiffiffi þ b2 t L1 p ffiffi
ffi ð2:50Þ
a kðp b2 Þðq c2 Þðq þ b= aÞ
pffiffiffi pffiffiffi
where q2 ¼ p=a, b ¼ V=ð2 aÞ and c ¼ b d a.
This expression may be expanded into partial fractions using the residual
theorem:
8 pffiffiffi 9
>
> a a expðqxÞ a2 ð5b2 c2 Þ expðqxÞ >
>
>
> p ffiffi
ffi p ffiffi
ffi >
>
>
> 2bðb2 c2 Þ ðq þ b= aÞ2 4b2 ðb2 c2 Þ ðq þ b= aÞ >
>
>
> >
>
2 >
< >
=
I0 d bx 2 2
2 1 a expðqzÞ a expðqzÞ
exp p ffiffi
ffi þb t L þ 2 2 p ffiffiffi þ pffiffi
ffi
ka a >
> 4b ðb c2 Þðq þ b= aÞ 4b2 ðb2 c2 Þðq b= aÞ >
>
>
> >
>
>
> >
>
>
> 2
a expðqxÞ 2
a expðqxÞ >
>
>
:þ pffiffi
ffi þ p ffiffi
ffi >
;
2cðb þ cÞðb2 c2 Þðq c= aÞ 2cðb cÞðb2 c2 Þðq þ c= aÞ
ð2:51Þ
which gives on inversion and after much algebraic manipulations:
8 9
> x pffi 3b2 þ c2 x pffi >
>
> 4t ierfc p ffiffiffiffi þ b t þ erfc p ffiffiffiffi þ b t >
>
>
> 2 at 2bðb2 c2 Þ 2 at >
>
>
> >
>
>
> p ffi >
>
>
> 1 2bx x >
>
pffiffiffi >
< þ exp p ffiffi
ffi erfc p ffiffiffiffi b t >
=
I0 d a 2b a 2 at

2qCp ðad VÞ >
> 1 x pffi >
>
>
> exp ðdx þ b2 c2 Þt erfc pffiffiffiffi þ c t >
>
>
> b c 2 at >
>
>
> >
>
>
> p ffi >
>
>
> 1 x x >
: exp pffiffiffi ðb þ cÞ þ ðb c Þt erfc pffiffiffiffi c t >
2 2 ;
bþc a 2 at
ð2:52Þ
The second part of the term in the transformed solution may be inverted in a
similar manner:
8 h pffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffi i9
<qVLpffiffiaffi exp ðx= aÞðb þ p þ b2 Þ =
L1 pffiffiffiffiffiffiffiffiffiffiffiffiffi
: kp b þ p þ b2 ;
( )
qVL 2 bx 1 expðqxÞ expðqxÞ 2b expðqxÞ
¼ exp b t þ pffiffiffi L pffiffiffi pffiffiffi pffiffiffi pffiffiffi
4kb2 a q b= a q þ b= a a ðq þ b= aÞ2
ð2:53Þ
which after transformation gives:
2 3
pffiffiffiffi x pffi pffiffiffi x pffi
4b at ierfc pffiffiffiffi þ b t a pffiffiffiffi þ b t
qVL 66 2 at 2 at 7
7
6 7 ð2:54Þ
4bk 4 pffiffiffi 2bx x pffi 5
þ a exp pffiffiffi erfc pffiffiffiffi b t
a 2 at
Finally the term:

pffiffiffi
1 I0 ad expðdxÞ I0 d a exp½ðdx þ ðb2 c2 ÞtÞ expðdxÞ
L ¼
k pðp þ b2 c2 Þ qCp ðad VÞ bc bc
ð2:55Þ
pffiffiffi
where b c ¼ d a.
Substitution of all the terms above into Eq. 2.45 gives the analytic solution to
the heating problem associated with the evaporation at the surface. Therefore, the
solution becomes:
8 pffi pffi 9
>
> 4 ierfc
2 2
px ffiffiffi þ b t þ 3b 2þc 2 erfc px ffiffiffi þ b t >
>
>
> 2 at 2bðb c Þ 2 at >
>
>
>
p ffi >
>
>
> >
>
> þ 1
exp 2bx
erfc pffiffiffi b t
x >
pffiffiffi >
> 2b a 2 at >
>
I0 d a < h i =
pffi
Tðx; tÞ ¼ 1 2 2
þ ðbcÞ exp½½dx þ ðb c Þt erfc 2pffi þ c t x
2qCp ðad VÞ >> >
>
> h i p ffi >
>
>
>
> >
>
>
>
1
exp pffiffia ðb þ cÞ þ ðb c Þt erfc 2pffiffiffi
x 2 2 x
c t > >
>
>
ðbþcÞ at >
>
>
: 2 expðdxÞ >
;
ðbcÞ
8 pffiffiffiffi pffi pffi > 9
> pffiffiffi
px ffiffiffi px ffiffiffi
qVL < 4b at ierfc 2 at þ b t a erfc 2 at þ b t =
pffiffiffi pffi
4bk >
: þ a exp 2bx pffiffi erfc p xffi
b t >
;
a 2
ð2:56Þ
Setting x = 0 in Eq. 2.56 results in the surface temperature, i.e.:
2 pffi pffi
ðb2 þc2 Þ pffi
3
4 t ierfc b t þ bðb2 c2 Þ erfc b t þ 1b
pffiffiffi 6 7
I0 d a 6 exp½ðb2 c2 Þt pffi
7
Tð0; tÞ ¼ 6þ erfc c t 7
2qCp ðad V Þ 46 ð bc Þ 7
5
exp½ðb2 c2 Þt pffi
2
ðbþcÞ erfc c t bc
qVL pffiffiffiffi pffi
pffiffiffi pffi
pffiffiffi pffi

4b at ierfc b t a erfc b t þ a 2 erfc b t ð2:57Þ
4bk
After rearrangement it becomes:
2 pffi pffi
ðb2 þc2 Þ pffi
3
pffiffiffi 4 t ierfc b t þ bðb
I0 d a 2 c2 Þ erfc b t
Tð0; tÞ ¼ 4
5
2qCp ðad V Þ þ 22c 2 exp½ðb2 c2 Þt erfc c ffit
p ðbþcÞ
b c bðbcÞ
qVL pffiffiffiffi pffi
pffiffiffi pffi

2b at ierfc b t þ a erfc b t
2bk
ð2:58Þ
Equations 2.56 and 2.58 give the complete quasi-steady solution of the heat
transfer equation. However, Eqs. 2.56 and 2.58 can be used to acquire a more
accurate solution which can be obtained by an iterative procedure. In this case, the
values for the velocity and the surface temperature can be obtained by stepping
forward in time using time steps which are small enough so that the change in the
surface velocity between steps is small and therefore the velocity derived in the
previous step can be used directly in Eqs. 2.56 and 2.58. With this new value of
the surface temperature, an improved estimate of the surface velocity (V) can be
obtained and the iteration repeated to give a convergent solution.
Equation 2.58 can be simplified assuming ad V, which is true for high
power intensities, it may be further assumed that t I/(ad2), which in practice for
metals means well in excess of 1 ps, then Eq. 2.58 reduces to:
rffiffiffi rffiffiffi
Io qVL pffiffiffiffi V a a V t
T ð0; tÞ ¼ at ierfc þ erfc ð2:59Þ
k 2 t V 2 a
After re-arranging Eqs. 2.58 and 2.59, it results in:
h i
T ð0; tÞ V ðtÞ 1 VL
Io
h i ¼ 2x ierfcð xÞ þ erf ð xÞ ð2:60Þ
T ð0; 1Þ V ð1Þ 1 VL
Io t¼1
where
rffiffiffi
V t
x¼ ð2:61Þ
2 a
Introducing of the fraction, r, of the steady temperature which has been reached,
allows Eq. 2.60 to be written as:
r 2=3 expðqÞ
2x iefcð xÞ þ erf ð xÞ ¼
Lo
1 þ CpT ½ 1 Tr 2 ð1Þ1=2 r 2=3 1 Tr 2 ð1Þ 1=2 expðqÞ
r 2
c
ð2:62Þ
where
1=2
mLo 1 2
1=2 1 2
q¼ 1 Tr ð1Þ 2 Tr ð1Þ ð2:63Þ
kTc Tr ð1Þ r
and
T ð0; tÞ
Tr ðtÞ ¼ ð2:64Þ
Tc
Equation 2.64 can be used to estimate the times for the surface to reach 90 %
(r = 0.90) of its steady state temperature.
The thermal stress analysis for the heating situation associated with the evap-
oration surface can be reduced to the stress analysis in the solid phase. This is
because of the fact that thermal stress developed in the liquid phase is almost zero.
Since the liquid layer thickness is small, the hydrostatic pressure generated in the
liquid phase is negligibly small. Moreover, the surface evaporation results in
convection boundary at the surface. In this case, the analytic solution for the heat
transfer equation in the solid phase needs to be obtained after incorporating the
convective boundary at the surface. The resulting equation is, then, used to
determine the stress filed due to laser heating and surface evaporation situation.
The thermal stress analysis is given below in the line of the previous study [5].
Since the solution of the thermal stress equation requires the temperature dis-
tribution in the solid, the solution of the Fourier heat transfer equation is provided
below briskly for a laser pulse intensity decaying exponentially with time. The
governing equation of heat transfer can be written as:
o2 T I 1 d 1 oT
2
þ expðbtÞ expðdxÞ ¼ ð2:65Þ
ox k a ot
where:

I1 ¼ 1 rf I o ð2:66Þ
The substrate material is considered as a semi-infinite body and heated by a
laser beam on the surface. The convective boundary condition is assumed be on
the substrate surface. In addition, as the depth is considered to extend to infinity
and the temperature to go down to zero. Heating occurs in the surface region
during the laser pulse. Therefore, the corresponding boundary conditions are:

At x ¼ 0 ) oT h
ox x¼0 ¼ k ðTð0; tÞ T0 Þ ð2:67Þ
At x ¼ 1 ) Tð1; tÞ ¼ 0
Initially substrate material is assumed to be at uniform temperature. Therefore,
the initial condition is:
At t ¼ 0 ) Tðx; 0Þ ¼ 0 ð2:68Þ
The Laplace transformation of Eq. (2.65) with respect to t, results in:
o2 T I1 d expðdxÞ 1
þ ¼ ½sT ðx; sÞ T ðx; 0Þ ð2:69Þ
ox2 k ðs þ b Þ a
Introducing the initial condition and rearranging the Eq. 2.69 yields:
o2 T I1 d expðdxÞ
2
g2 T ¼ ð2:70Þ
ox k ðs þ bÞ
where g2 = s/a and s is the Laplace transform variable. Equation 2.70 has the
solution:
sÞ ¼ A1 expðgxÞ þ A2 expðgxÞ I1 d expðdxÞ
Tðx; ð2:71Þ
kðs þ bÞðd2 g2 Þ
where A1 and A2 are constants. Introducing the boundary conditions will enable
calculation of the constants A1 and A2, i.e.:
I1 dðh þ dkÞ hT0
A1 ¼ 0 and A2 ¼ 2
þ ð2:72Þ
kðs þ bÞðd g Þðh þ kgÞ sðh þ kgÞ
2
After substituting the values of A1 and A2 in Eq. 2.71, it yields:
sÞ ¼ I1 dðh þ dkÞegx hT0 egx I1 dedx

Tðx; 2
þ 2
ð2:73Þ
kðs þ bÞðd g2 Þðh þ kgÞ sðh þ kgÞ kðd g2 Þðs þ bÞ
which gives the solution for the temperature distribution in the Laplace domain.
The inverse Laplace Transform of Eq. 2.73 provides the temperature distribu-
tion within the substrate material in space (x) and time (t). The mathematical
arrangements of the Laplace inversion of Eq. 2.73 are given in the previous study
[6]. Therefore, the equation after the Laplace inversion is given below as:
8 9
> pffiffiffi >
> e ðad2 tdxÞ
Erfc px ffiffid at >
>
> 2 at

>
>
> pffi pffiffi >
>
> 2 h a >
>
>
> 2ðbþad Þ k þd a >
>
>
> >
>
>
> p ffiffiffi >
>
> e ðad2 tþdxÞ
Erfc 2pffiffi x
d at >
>
> at

>
>
> pffi ffiffi >
>
> 2 h a p >
>
>
> 2ðbþad Þ k þd a >
>
>
> pffiffiffiffiffiffiffiffi
p ffiffiffi >
>
>
> x ffiffi
>
>
>
> e bt
e ðb=aÞxi
Erfc p þ bt i >
>
I1 da ðh þ kdÞ þ
3=2 < 2 at
pffi pffiffi
=
2 h a
Tðx; tÞ ¼ 2ðbþad Þ k b i
k2 >
> pffiffiffiffiffiffiffiffi
pffiffiffi > >
>
> ebt e ðb=aÞxi Erfc 2pxffiffiat bti >
>
>
> >
>
>
>þ ffi p ffiffi
>
>
>
> 2
Þ
p
h a
þ >
>
>
>
2ðbþad k b i >
>
>
> >
>
>
> p ffiffiffi >
pffiffi e e Erfc 2pffiþk at >
2 =k2 Þat
>
>
hx=k ðh x h
>
>
>þ
>
h a

>
>
>
> k h 2a
þb h 2a
ad 2 >
>
>
> k 2 k 2 >
>
>
> >
>
>
:p dx ad2 t
ke ðe
ffiffi e Þ bt >
;
2
aðbþad ÞðhþkdÞ
" !#
x 2 2 x h p ffiffiffiffi
þ T0 Erfc pffiffiffiffi ehx=k eðh =k Þat Erfc pffi þ at ð2:74Þ
2 at 2 k
where Erfc is the complementary error function. Equation 2.74 is the closed form
solution for temperature distribution. The temperature distribution can be
expressed in a non-dimensional form by introducing dimensionless quantities and
substituting in Eq. 2.74. The dimensionless quantities are:
Tkd h
x ¼ xd; t ¼ ad2 t; T ¼ ; h ¼ ; b ¼ bt ð2:75Þ
I1 dk
The dimensionless temperature distribution for a full pulse is then:

82
pffiffiffi

pffiffiffi
39
>
>

ex et Erfc p x ffiffiffi
t

t >
>
>6

7>>
7>
2 t 2 t
>
>
> 6 2ðb þ1Þðh þ1Þ 2ðb þ1Þðh 1Þ >
>
>
>
> 6 7 >
>
>
> 6 p ffiffiffi p ffiffiffiffiffiffi p ffiffiffi p ffiffiffiffiffiffi 7 >
>
> 6
> 6 eb e b x i Erfc px ffiffi ffi b t i

eb t e b x i Erfc p
x ffiffiffi

b i 7 >
>
>
> 2 t
t 7 >
>
>
> 6 þ 2 t
p ffiffiffi ffi
þ p ffiffiffiffi
7 >
>
>6
> 6
2ðb þ1Þ h b i
2ðb þ1Þ h þ b i 7 >
>
>
> 7 >
>
>
> 6 7 >
>
>
> 6 h x ðh Þ2 t x ffiffiffi
p ffiffiffi 7 >
>
>
> 4 h e e Erfc p
h t
e x
ð e t
e

b t
Þ 5 >
>
>
< þ
2 t

>
=
2ð ðh Þ 2
þb Þ ð ðh Þ 2
1 Þ ðb þ1Þðh þ1Þ
T ðx ; t Þ ¼ ðh þ 1Þ 2 3
>
> pffiffiffi pffiffiffi >
>
>
>
t

t >
>
>
> 6 2 t 2 t
7 >
>
> 6
> 7 >
>
>
> 6

2ðc þ1Þðh þ1Þ
2ðc þ1Þðh 1Þ
7 >
>
>
> 6 pffiffiffi
pffiffiffi
7 >
>
>
> 6 p ffiffiffiffiffi ffi p ffiffiffiffiffi ffi 7 >
>
>
> 6 e c
e

c x i
Erfc px ffiffiffi

c t i e

c t c x i
e

Erfc px ffiffiffi

c t i
7 >
>
>
> 6 þ
2 t
p ffiffiffi þ
2 t
p ffiffiffi 7 >
>
>
> 6 2 ð c þ1Þðh c iÞ 2 ð c þ1Þðh þ c iÞ
7 >
>
>
> 6 7 >
>
>
> 6 7 >
>
>
> p ffiffi
ffi >
>
Erfc pffiffi ffih t
x ðh Þ2 t
>
> 4 h e h
e x
e x
ð e t
e

c t
Þ 5 >
>
: þ
2 t

;
2ððh Þ þc Þððh Þ 1Þ
2 2 ðc þ1Þðh þ1Þ
ð2:76Þ
In order to solve for the stress distribution within the substrate it is possible to
consider the equation governing the momentum in a one-dimensional solid for the
linear elastic case, i.e.:
o2 rx 1 o2 rx o2 T
2
2 2 ¼ c2 2 ð2:77Þ
ox c1 ot ot
pffiffiffiffiffiffiffiffiffi
where c1 is the wave speed in the solid c1 ¼ E=q and
1þt
c2 ¼ qaT ð2:78Þ
1t
where t is Poisson’s ratio, q is the density of the solid and aT is the thermal
expansion coefficient of the solid.
In order to solve the momentum equation (Eq. 2.76), it is necessary to establish
the initial conditions for stress and temperature fields. In this case, the substrate
material is assumed to be free from stresses initially (at time = 0) and as the time
extends to infinity, the stress-free state must apply in the substrate. The same initial
condition for the temperature is applied as in Eq. 2.77, provided that as time
approaches infinity, the temperature in the substrate material reduces to zero. This
is due to the fact that the laser pulse decays exponentially with time; therefore, as
time approaches infinity, the laser pulse intensity becomes zero. Therefore, the
initial and boundary conditions for the stress field are:
At t ¼ 0 ) rx ¼0
At t ¼ 1 ) rx ¼0
ð2:79Þ
At x ¼ 0 ) rx ¼0
At x ¼ 1 ) rx ¼0
Taking the Laplace Transformation of Eq. 2.76 with respect to time yields:
o2 r
x 1 2 h : i
2
s r
x ð x; s Þ srx ð x; 0 Þ rx ð x; 0 Þ ¼ c 2 s T ð x; s Þ sT ð x; 0 Þ T ð x; 0 Þ
ox2 c21
ð2:80Þ
where rx ðx; sÞ and T ðx; sÞ are the Laplace transforms of thermal stress and tem-
perature respectively in the x and s domains.
By substituting the initial conditions, Eq. 2.80 reduces to:
o2 r
x s2
r x ðx; sÞ ¼ c2 s2 T ðx; sÞ ð2:81Þ
ox2 c21
Considering the temperature distribution in a Laplace domain for an expo-
nentially decaying pulse with time, Eq. 2.73, and substituting it into Eq. 2.81, and
solving for the stress field, yields:

o2 r
x s2 2 I1 dðh þ dkÞegx hT0 egx I1 dedx
r
x ðx; sÞ ¼ c 2 s þ
ox2 c21 kðs þ bÞðd2 g2 Þðh þ kgÞ sðh þ kgÞ kðd2 g2 Þðs þ bÞ
ð2:82Þ
Now let M1 and M2 be defined as:
I1 dðh þ dkÞc2 s2 hT0 c2 s

M1 ¼ þ ð2:83Þ
kðs þ bÞðd2 g2 Þðh þ kgÞ ðh þ kgÞ
and
I1 dc2 s2
M2 ¼ ð2:84Þ
kðd2 g2 Þðs þ bÞ
Then Eq. 2.82 becomes:
o2 r
x s2
r x ðx; sÞ ¼ M1 egx þ M2 edx ð2:85Þ
ox2 c21
The complementary and the particular solutions of Eq. 2.85 are:
sx sx
ðr
x Þh ¼ A3 ec1 þ A4 e c1 ð2:86Þ
While the particular solution has two parts, the first part is:
pffis
x Þp1 ¼ G1 e ax
ðr ð2:87Þ
Substituting Eq. 2.87 in Eq. 2.85 yields:

M1
G1 ¼ 2 ð2:88Þ
g2 cs2
1
The second part of the particular solution is:
x Þp2 ¼ G2 edx
ðr ð2:89Þ
Substituting Eq. 2.89 in Eq. 2.85 yields:

M2
G2 ¼ 22 ð2:90Þ
d cs2
1
So, the general solution for the stress field becomes:

sx sx
x Þg ¼ A3 ec1 þ A4 e c1 þ G1 egx þ G2 edx
ðr ð2:91Þ
From the boundary condition ðx ¼ 1 ) rx ¼ 0Þ, this yields A3 ¼ 0.

Then, Eq. 2.91 reduces to:
sx
pffis
x Þg ¼ A4 e c1 þ G1 e ax þ G2 edx
ðr ð2:92Þ
Consider the boundary condition of the stress field at the surface, where at
x ¼ 0 ) orx ¼ 0, the constant in Eq. 2.92 becomes:
rffiffiffi
c1 s
A4 ¼ G1 ðsÞ þ dG2 ðsÞ ð2:93Þ
s a
Therefore, Eq. 2.92 becomes:
pffis c1 sx c1 d sx
x ðx; sÞ ¼ G1 ðsÞe
r ax G1 ðsÞ pffiffiffiffiffi e c1 þ G2 ðsÞedx G2 ðsÞ e c1 ð2:94Þ
sa s
Finding the solution for rx in the x and t domain, one should take the inverse
Laplace Transform for each term in Eq. 2.94. To do this, the following terms are
introduced:
pffis sx
x Þ1 ¼ G1 ðsÞe ax ðr
ðr x Þ2 ¼ G1 ðsÞ pcffiffiffi
1
sa
e c1
sx ð2:95Þ
x Þ3 ¼ G2 ðsÞedx
ðr x Þ4 ¼ G2 ðsÞ c1sd e c1
ðr
Consequently, the solution for the stress distribution is the summation of the
inverse Laplace Transforms of the above terms.
Therefore, the Laplace inversion of Terms ðr

x Þ1 ; ðr
x Þ2 ; ðr
x Þ3 ; ðr
x Þ4 can be
stated as follows.
ðr
x Þ1 is composed of the terms:
ðr
x Þ1 ¼ ðr
x Þ11 þðr
x Þ12 ð2:96Þ
where
" pffis #
I1 dðh þ dkÞc2 se ax
ðr
x Þ11 ¼ pffiffiffiffiffiffiffi ð2:97Þ
k ðd2 s=aÞðs þ bÞðh þ k s=aÞð1=a s=c21 Þ
and
" pffis #
e ax
ðr
x Þ12 ¼ hT0 c2 pffiffiffiffiffiffiffi ð2:98Þ
ðh þ k s=aÞð1=a s=c21 Þ
Let
I1 dðh þ dkÞc2
C10 ¼ ð2:99Þ
k
and
C20 ¼ hT0 c2 ð2:100Þ
Then, the inverse Laplace transformation of ðrx Þ1 can be written as:

L1 ðrx Þ1 ¼ ðrx Þ11 þðrx Þ12 ð2:101Þ
where
ðrx Þ11 ¼ ðrx Þ111 þðrx Þ112 þðrx Þ113 þðrx Þ114 þðrx Þ115 þðrx Þ116 þðrx Þ117 þðrx Þ118
ð2:102Þ
These terms are:
2 pffiffiffi 2 pffiffiffiffi 3
h a h 2athxk h at x
p ffiffi
ffi 6 k e k Erfc þ p ffiffiffiffi
c21 h2 k3 a2 aC10 6 k 2 at 7 7
ðrx Þ111 ¼ 2 6 p ffiffi
ffi p ffiffiffiffi 7
2ða2 h2 c1 k2 Þðh2 a þ k2 bÞðh2 k2 d Þ 4 h a h2 atþhx
2
h at x 5
þ e k2 k Erfc þ pffiffiffiffi
k k 2 at
ð2:103Þ
and
2 pffiffiffi pffi 3
bt bax
pffiffiffiffiffi x
pffiffiffi 6 be Erfc bt þ pffiffiffiffi 7
c21 ha3 bC10 6 2 at 7
ðrx Þ112 ¼ 6 pffi 7
2ðb þ ad2 Þðh2 a þ k2 bÞðc21 þ abÞ 4 pffiffiffi bxþbt pffiffiffiffiffi x 5
þ be a Erfc bt þ pffiffiffiffi
2 at
ð2:104Þ
2 rffiffiffi 3
c1 c21 tc1 x t x
pffiffiffi 6 pffiffiffi e a a Erfc c1 þ pffiffiffiffi 7
c31 ha4 aC10 6 a a 2 at 7
ðrx Þ113 ¼ 6 7
rffiffiffi 7
2ða2 h2 c21 k2 Þðc21 þ abÞðd2 a2 c21 Þ 6
4 c1 c21 tþc1 x t x 5
þ pffiffiffi e a a Erfc c1 þ pffiffiffiffi
a a 2 at
ð2:105Þ
and
2 3
pffiffiffi ad2 tdx pffiffiffiffi x
p ffiffi
ffi 6 a de Erfc d at þ p ffiffiffiffi
c21 ha2 adC10 6 2 at 7 7
ðrx Þ114 ¼ 6 7
2ðb þ ad2 Þðd2 a2 c21 Þðh2 k2 d2 Þ 4 pffiffiffi ad2 tþdx pffiffiffiffi x 5
þ ade Erfc d at þ pffiffiffiffi
2 at
ð2:106Þ
and
2 2 3
2 x
6 pffiffiffiffiffi exp 7
6 pt 4at 7
p ffiffi
ffi 6 pffiffiffi pffiffiffiffi 7
2 2 3 2
c1 h k a aC10 6 h a h 2athxk h at x 7
6 2
7
ðrx Þ115 ¼ 2 6 þ ek Erfc þ pffiffiffiffi 7
2ða2 h2 c1 k2 Þðh2 a þ k2 bÞðh2 k2 d Þ 6
2
6
k k 2 at 7 7
6 pffiffiffi pffiffiffiffi 7
4 h a h22atþhxk h at x 5
ek Erfc þ pffiffiffiffi
k k 2 at
ð2:107Þ
and
2 2
3
p2ffiffiffi exp x
6 pt 4at
pffiffiffi
c21 ka2 abC10 6 pffiffiffi btxpffib pffiffiffiffiffi 7
7
ðrx Þ116 ¼ 6 þ be a Erfc bt þ 2pffiffiffi
x 7
2 2 2 2 6 at 7
2ðc1 þ abÞðh a þ k bÞðb þ ad Þ 4 pffib 5
pffiffiffi pffiffiffiffiffi
bebtþx a Erfc bt þ 2pxffiffiffi
at
ð2:108Þ
and
2 2
3
p2ffiffiffi exp x
4at
pffiffiffi 6 pt 7
c21 ka3 aC10 6 c c21 tc1 x pffiffit 7
ðrx Þ117 ¼ 6 þ ae
p 1ffiffi a a x ffiffiffi 7
Erfc c1 a þ 2 at 7
p
2 2 2 2 2 2 2 2 6
2ða h c1 k Þðc1 þ abÞðd a c1 Þ 4 c2 t c x pffiffi 5
1 1
pc1ffiffia e a þ a Erfc c1 at þ 2pxffiffiffi
at
ð2:109Þ
2 2
3
p2ffiffiffi exp x
pffiffiffi 2 6 pt 4at
pffiffiffiffi 7
c21 ka2
ad C10 6 pffiffiffi ad2 tdx 7
ðrx Þ118 ¼ 6 þ a de Erfc d at þ px ffiffiffi 7
2 2 2
2 2 4 2 at 5
2 2
2ðb þ ad Þðd a c1 Þðk d h Þ pffiffiffi 2
pffiffiffiffi
adead tþdx Erfc d at þ 2pxffiffiffi at
ð2:110Þ
Also ðrx Þ12 consists of the terms:
ðrx Þ12 ¼ ðrx Þ121 þðrx Þ122 þðrx Þ123 þðrx Þ124 ð2:111Þ
These terms are:
2 pffiffiffi pffiffiffiffi 3
h a h22athxk h at x
p ffiffi
ffi e Erfc þ p ffiffiffiffi
c2 hka aC20 6 2 at 7
k
6 k k 7
ðrx Þ121 ¼ 1 2 2 6 p ffiffi
ffi p ffiffiffiffi 7 ð2:112Þ
2hða h c21 k2 Þ 4 h a h2 atþhx h at x 5
þ e k2 k Erfc þ pffiffiffiffi
k k 2 at
and
2 rffiffiffi 3
c1 c21 tc1 x t x
pffiffiffi pffiffiffi e a a Erfc c1 þ pffiffiffiffi
c1 ha2 aC20 6 6 a a 2 at 7 7
ðrx Þ122 ¼ 6 rffiffiffi 7 ð2:113Þ
2ða2 h2 c21 k2 Þ 4 c1 c21 tþc1 x t x 5
þ pffiffiffi e a a Erfc c1 þ pffiffiffiffi
a a 2 at
and
2 3
2
p2ffiffiffi exp x
p ffiffi
ffi 6 pt
pffiffi 2
4at
pffiffiffi 7
c21 ka aC20 6 6 h a h 2athx h at
7
7
ðrx Þ123 ¼ 2 2 2 2 6 þ k ek k Erfc k þ 2pffiffiffi x
at 7 ð2:114Þ
2ða h c1 k Þ 4 pffiffi 2 pffiffiffi 5
h a h 2atþhx h at x ffiffiffi
k e k k Erfc k þ 2 at p
and
2 2
3
p2ffiffiffi exp x
4at
pffiffiffi 6 pt 7
c21 ka aC20 6 6
c2 t
c1ffiffi a1 c1a x
pffiffit 7
x ffiffiffi 7
ðrx Þ124 ¼ þ ae
p Erfc c1 a þ 2 at 7
p ð2:115Þ
2ða2 h2 c21 k2 Þ 6
4 pffiffi 5
c2 t c x
1 1
pc1ffiffia e a þ a Erfc c1 at þ 2pxffiffiffi
at
The Laplace transform of ðr

x Þ2 can be written as:

L1 ðrx Þ2 ¼ ðrx Þ21 þðrx Þ22 ð2:116Þ
where ðrx Þ21 consists of:
ðrx Þ21 ¼ ðrx Þ211 þðrx Þ212 þðrx Þ213 þðrx Þ214 þðrx Þ215 þðrx Þ216 þðrx Þ217 þðrx Þ218
ð2:117Þ
Now let:
I1 dðh þ dkÞc2 c1
C30 ¼ pffiffiffi ð2:118Þ
ak
and
hT0 c2 c1 hT0 c2 c1
C40 ¼ pffiffiffi C40 ¼ pffiffiffi ð2:119Þ
a a
then
2 3
pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 ffiþ
c21 hk6 a2 C30 6 pðtx=c 1 Þ 7 x
ðrx Þ211 ¼ 4 p ffiffi 2 p ffiffi p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 5 U t
k2 ðc21 k2 h a Þðh2 a þ k2 bÞðh2
2 2 k2 d2 Þ h a ehk2at Erf h a t x=c1 c1
k k
ð2:120Þ
and
" #
c21 ha3 C30 pffiffiffiffiffiffiffiffiffiffiffiffiffiffi
1 ffiþ
pðtx=c1 Þ
ðrx Þ212 ¼ pffiffiffiffiffiffiffi bt pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ðb þ ad2 Þðh2 a þ k2 bÞðc21 þ abÞ be Erf bðt x=c1 Þ

x
U t
c1
ð2:121Þ
and
2 3
1 ffiþ
c21 ha5 C30 4
pðtx=c1 Þ
ðrx Þ213 ¼ c2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 5
ða2 h2 c21 k2 Þðc21 þ abÞðd2 a2 c21 Þ pc1ffiffi e a1 t Erf pc1ffiffi t x=c1
a a

x
U t
c1
ð2:122Þ
and
2 3
1 ffiþ
c21 ha2 C30 4 pffiffipðtx=c 1Þ
pffiffiffi pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi 5
ðrx Þ214 ¼ ffi
ðb þ ad2 Þðd2 a2 c21 Þðh2 k2 d2 Þ
2
adead t Erf ad t x=c1

x
U t
c1
ð2:123Þ
and

c21 k5 a2 C30 x h2 a h22aðtx=c1 Þ
ðrx Þ215 ¼ D t þ 2 ek
h2 c21 k2 Þðh2 a þ k2 bÞðh2 k2 d2 Þ
ða2 c1 k
x
U t
c1
ð2:124Þ
and
pffiffiffi
c21 ka2 aC30 x bðtx=c1 Þ
ðrx Þ216 ¼ 2 2
D t be
ðc 2 2 c1
1 þ abÞðh
a þ k bÞðb þ ad Þ
x
U t ð2:125Þ
c1
and
pffiffiffi
c21 ka4 aC30 x c21 c21 ðtx=c1 Þ
ðrx Þ217 ¼ 2 2 2 2 2 2
D t þ e a
ða 2 2 c1 a
h c1 k Þðc1 þ abÞðd a c1 Þ
x
U t
c1
ð2:126Þ
and
pffiffiffi
c21 ka aC30 x 2 ad2 ðtx=c1 Þ
ðrx Þ218 ¼ D t þ ad e
ðb þ ad2 Þðd2 a2 c21 Þðk2 d2 h2 Þ c1

x
U t ð2:127Þ
c1
Also ðrx Þ22 consists of the terms:
ðrx Þ22 ¼ ðrx Þ221 þðrx Þ222 þðrx Þ223 þðrx Þ224 þðrx Þ225 þðrx Þ226 ð2:128Þ
These terms are:

aC40 x
ðrx Þ221 ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi U t ð2:129Þ
h pðt x=c1 Þ c1
and
" pffiffiffi pffiffiffi rffiffiffiffiffiffiffiffiffiffiffi#
c21 k2 aC40 1 h a h22at h a x
ðrx Þ222 ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ e k Erf t
hða2 h2 c21 k2 Þ pðt x=c1 Þ k k c 1

x
U t ð2:130Þ
c1
and
" rffiffiffiffiffiffiffiffiffiffiffiffi#
ha3 C40 1 c1 c21 t c1 x
ðrx Þ223 ¼ 2 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi þ pffiffiffi e a Erf pffiffiffi t
ða h c21 k2 Þ pðt x=c1 Þ a a c1

x
U t ð2:131Þ
c1
and
pffiffiffi
k aC40 x x
ðrx Þ224 ¼ D t U t ð2:132Þ
h2 c1 c1
and
pffiffiffi
c21 k3 aC40 x h2 a h22aðtx=c1 Þ x
ðrx Þ225 ¼ 2 2 2 D t þ 2 e k U t ð2:133Þ
h ðc1 k h2 a2 Þ c1 k c1
and
pffiffiffi
ka2 aC40 x c21 c21 ðtx=c1 Þ x
ðrx Þ226 ¼ D t þ e a U t ð2:134Þ
ðc21 k2 h2 a2 Þ c1 a c1
Further
" #
I1 dc2 s2 edx
ðr
x Þ3 ¼ ð2:135Þ
k ðd2 s=aÞðs þ bÞðd2 s2 =c21 Þ
and
2 3
c21 aec1 dtdx c31 aeðc1 dtþdxÞ

I1 dc2 6
6 2ðb þ c1 dÞðc1 adÞ 2ðc1 d bÞðc1 þ adÞ 7 7
L1 ðr
x Þ3 ¼ ðrx Þ3 ¼ 6 2 3 2 ad2 tdx
7
k 4 2 2 ðbtþdxÞ
c ab e c1 a d e 5
2 1 2 2 2
þ 2 2
ðb c1 d Þðb þ ad Þ ða2 d c21 Þðb þ ad Þ
ð2:136Þ
also
" #
I 1 d2 c 2 c 1 seðs=c1 Þx
ðr
x Þ4 ¼ ð2:137Þ
k ðd2 s=aÞðs þ bÞðd2 s2 =c21 Þ
and
2 c2 aec1 dðtx=c1 Þ 3
2 1
þ
c21 aec1 dðtx=c1 Þ
1 I 1 d c c
2 14 2dðbþc 1 dÞðc 1 adÞ 2dðc 1 dbÞðc 1 þadÞ
5 x
L ðr x Þ 4 ¼ ð rx Þ 4 ¼ c2 abebðtx=c1 Þ
2
c21 a2 ead ðtx=c1 Þ
U t
k 212 2 2 2 2 2
c1
ðc1 d b Þðbþad Þ ðc1 a2 d Þðbþad Þ
ð2:138Þ
where U ðt x=c1 Þ is a unit step function, Erf ð yÞ is the error function of the
variable y and Dðt x=c1 Þ is the Dirac delta function. The unit step function has
the values of 0 for t
x=c1 and 1 for t [ x=c1 .
The closed form solution of the stress distribution can be written as:
rx ðx; tÞ ¼ ðrx Þ1 þðrx Þ2 þðrx Þ3 þðrx Þ4 ð2:139Þ
Some additional dimensionless quantities are defined to present the stress dis-
tribution in the dimensionless form, these are:

c1 krx x
c1 ¼ ; rx ¼ and U
¼ t
ð2:140Þ
ad I1 da2 c2 c1
where U is the dimensionless unit step function.
Therefore, for the dimensionless stress distribution, the followings are resulted:

rx 1 ¼ rx 11 þ rx 21 ð2:141Þ
where

rx 11 ¼ rx 111 þ rx 112 þ rx 113 þ rx 114 þ rx 115 þ rx 116 þ rx 117 þ rx 118

ð2:142Þ
and each of these terms are given as:
2 3
pffiffiffiffi
h2 t h x
x

Erfc h t þ pffiffiffiffi þ 7

ðh þ 1Þh2 c2 6h e 2 t
1 6 7
rx 111 ¼ 2 6 7
2ðh 1Þðh c1 Þðh þ b =t Þ 4
2 2 2
2 p ffiffiffi
ffi x 5
h eh t þh x Erfc h t þ pffiffiffiffi
2 t
ð2:143Þ
and
pffiffiffi2 pffiffiffiffiffi 3
b bt x x
6 e Erfc b þ p ffiffiffi
ffi

ðh þ 1Þh c2
1 b

6 2 t 7 7
rx 112 ¼ 2 6 p ffiffiffi 7
2ðb =t þ 1Þðh þ b =t Þðc1 þ b =t Þ 4 b þ b x
2 pffiffiffiffiffi x 5
þe t Erfc b þ pffiffiffiffi
2 t
ð2:144Þ
and
2 pffiffiffiffi 3
2
xffiffiffi

ðh þ 1Þh c3 c2 ec1 t c1 x Erfc c1 t þ 2p þ
1 6 1 t 7
rx 113 ¼ 4 pffiffiffiffi 5
2ðh2 c2 2 2
1 Þð1 c1 Þðc1 þ b =t Þ c2 ec1 t þc1 x Erfc c t þ p
2
xffiffiffi
1 1 2 t
ð2:145Þ
and
2 pffiffiffiffi 3
xffiffiffi

ðh þ 1Þh c2 et x Erfc t þ 2p þ
rx 114 ¼ 1

4 pffiffiffiffi t 5 ð2:146Þ
2 2
2ðh 1Þð1 c1 Þð1 þ b =t Þ e
t þx
Erfc t þ pffiffiffi
x
2 t
and
2 2
3
p2ffiffiffiffiffi exp x
pt 4t
6 7

ðh þ 1Þh2 c2 6 h2 t h x pffiffiffiffi 7
rx 115 ¼ 1 6
6 þh e Erfc h
t þ p xffiffiffi 7
2 2 2 2
2ðh 1Þðh c1 Þðh þ b =t Þ 4 2 t 7
pffiffiffiffi 5
h2 t þh x
h e Erfc h t þ 2pffiffitffi
x
ð2:147Þ
and
2 2
3
6 pt 4t 7

ðh þ 1Þc2 6 qffiffiffiffi pffiffiffi pffiffiffiffiffi 7
1 b 6
b b bt x 7
rx 116 ¼ 6þ t e Erfc b þ 2pffiffitffi 7
x
2t ðb =t þ 1Þðh2 þ b =t Þðc2
1 þ b =t Þ 6 7
4 qffiffiffiffi pffiffiffi
pffiffiffiffiffi 5
b b þ bt x xffiffiffi
t e Erfc b þ 2 tp
ð2:148Þ
and
2 2
3
4t
6 pt 7

ðh þ 1Þc4 6 c2 t c x pffiffiffiffi 7
rx 117 ¼ 1 6
6 þc1 e
1 Erfc c pffiffiffi 7
x
1 t þ 2 t 7
1
2 2 2 2
2ðh c1 Þð1 c1 Þðc1 þ b =t Þ 4
2
pffiffiffiffi 5
xffiffiffi
c1 ec1 t þc1 x Erfc c1 t þ 2p t
ð2:149Þ
and
2
2
3
4t

ðh þ 1Þc2 6 pt pffiffiffiffi 7
1 6 t x þ p xffiffiffi 7
rx 118 ¼ 6 þe Erfc t 7 ð2:150Þ
2ð1 h2 Þð1 c2
1 Þð1 þ b =t Þ 4 pffiffiffiffi 2 t 5

x ffiffiffi
et þx Erfc t þ 2p t

rx 12 in dimensionless form is:

rx 12 ¼ rx 121 þ rx 122 þ rx 123 þ rx 124 ð2:151Þ
where
2 3
h2 t h x
pffiffiffiffi x

Erfc h t þ pffiffiffiffi 7

T0 h c2 6h e
1 6 2 t 7
rx 121 ¼ 6 7 ð2:152Þ
2ðh2 c2 1 Þ 4 2
h t þh x
p ffiffiffi
ffi x 5
þh e Erfc h t þ p ffiffiffi
ffi
2 t
and
2 3

c2 t c1 x
pffiffiffiffi x
6 c1 e 1 Erfc
t þ pffiffiffiffi 7 c1

T0 h2 c1 6 2 t 7
rx 122 ¼ 6 7 ð2:153Þ
2ðh2 c2 1 Þ 4 2
c1 t þc1 x
p ffiffiffi
ffi x 5
þc1 e
Erfc c1 t þ pffiffiffiffi
2 t
and
2
2
2
pffiffiffiffi 3
p2ffiffiffiffiffi exp x xffiffiffi

T0 h c2 pt 4t þ h eh t h x Erfc h t þ 2p t 5
rx 123 ¼ 1 4 pffiffiffiffi
2ðh2 c2 1 Þ h e
h2 t þh x
Erfc h t þ 2p xffiffiffi
t
ð2:154Þ
and
2 2
2
pffiffiffiffi 3
p2ffiffiffiffiffi exp x xffiffiffi

T0 c2 þ c1 ec1 t c1 x Erfc c1 t þ 2p
1 h 4 pt 4t
t 5
rx 124 ¼ pffiffiffiffi
2ðh2 c2 c2 t þc1 x
1 Þ c1 e 1 Erfc c1 t þ 2p xffiffiffi
t
ð2:155Þ

rx 2 in dimensionless form is:

rx 2 ¼ rx 21 þ rx 22 ð2:156Þ
where

rx 21 ¼ rx 211 þ rx 212 þ rx 213 þ rx 214 þ rx 215 þ rx 216 þ rx 217 þ rx 218

ð2:157Þ
and
" sffiffiffiffiffiffiffiffiffiffiffiffiffiffi!#

ðh þ 1Þh c3 1 h2 t x
rx 211 ¼ 2 2 2
1
2 p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ffi þ h e Erf h t U
ðh 1Þðh c1 Þðh þ b =t Þ
pðt x =c1 Þ c1
ð2:158Þ
and
" rffiffiffiffiffiffiffiffiffiffi sffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi!#

ðh þ 1Þc3
1 h

1 b b b x
rx 212 ¼ 2 2 p ffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ffi þ
e Erf b þ U
ðb =t þ 1Þðh þ b =t Þðc1 þ b =t Þ
pðt x =c1 Þ t c1 t
ð2:159Þ
and

ðh þ 1Þc3
1 h

1 c2
x
rx 213 ¼ 2 2 2 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ffi þ c 1 e 1 t Erf c 1 t U
ðh c1 Þð1 c1 Þðc1 þ b =t Þ
pðt x =c1 Þ c1
ð2:160Þ
and

ðh þ 1Þc3
1 h

1 t x
rx 214 ¼ 2 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi

ffi þ e Erf t U
ð1 h Þð1 c1 Þð1 þ b =t Þ
pðt x =c1 Þ c1
ð2:161Þ
and

ðh þ 1Þc3 x 2 h2 x
2 h t c1
rx 215 ¼ 1
2 þ b =t Þ
D t þ h e U
ðh2 1Þðh2 c2
1 Þðh c 1
ð2:162Þ
and

ðh þ 1Þc3 x b b þbcx
rx 216 ¼ 1
D t e 1 U
ðb =t þ 1Þðh2 þ b =t Þðc2
1 þ b
=t Þ c
1 t
ð2:163Þ
and

ðh þ 1Þc3 x 2 c2
rx 217 ¼ 1
D t þ c e 1 t c1 x U
ðh2 c2 Þð1 c 2 Þðc2 þ b =t Þ c 1
1 1 1 1
ð2:164Þ
and

ðh þ 1Þc3 x
t xc
rx 218 ¼ 1
D t þ e 1 U ð2:165Þ
ð1 h2 Þð1 c2
1 Þð1 þ b =t Þ c 1

Also rx 22 is:

rx 22 ¼ rx 221 þ rx 222 þ rx 223 þ rx 224 þ rx 225 þ rx 226 ð2:166Þ
where

T0 c1
ffi U
rx 221 ¼ pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi ð2:167Þ
pðt x =c1 Þ
and

T0 c3 1 h2 t x
rx 222 ¼ 2
1
2
ffi þ h e Erf h t
pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi U
ðh c1 Þ
pðt x =c1 Þ c1
ð2:168Þ
and

T0 c1 h2 1 c2
x
rx 223 ¼ 2 pffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi
ffi þ c 1 e 1 t
Erf c 1 t U
ðh c1 Þ 2
pðt x =c1 Þ c1
ð2:169Þ
and

T0 c1 x
rx 224 ¼ D t U

ð2:170Þ
h c1
and

T0 c3 x 2 h2 x
2 h t c1
rx 225 ¼ 1
D t þ h e U ð2:171Þ
h ðh2 c21 Þ c1
and

T c h x c2
rx 226 ¼ 20 1 2 D t þ c2
1 e 1 t c1 x U
ðh c1 Þ c1
Further

2 b x
c3 c1 t x
1 t e c3 c1 t x
1 t e c2
1 t b e
rx 3 ¼ þ þ
2ðb þ c1 t Þðc1 1Þ 2ðc1 t b Þðc1 þ 1Þ ðb þ t Þðb2 c2 2
1 t Þ
2 t x
c t e
1
ðb þ t Þð1 c2 1 Þ
ð2:172Þ
and
2 b x x
3

c1 t x c1 t þx 2 b þ c1 t t c1
c3
1 t e c3
1 t e c3
1 t b e c3
1 t e
rx 4 ¼ 4 þ 5 U
2ðb þ c1 t Þðc1 1Þ 2ðc1 t b Þðc1 þ 1Þ ðb þ t Þðc2 2
1 t b Þ
2 ðb þ t Þðc2
1 1Þ
ð2:173Þ
Consequently, the dimensionless form of the stress equation is:

r x ¼ r x 1 þ r x 2 þ r x 3 þ r x 4 ð2:174Þ
The dimensionless temperature (Eq. 2.76) and stress distributions (Eq. 2.174)
are computed during the heating pulse.
2.3 Findings and Discussions
The findings of the analytical treatment of laser heating process in relation to

forming and welding are given in line with the previous studies [7, 8]. In the
analytical treatment of the laser heating process, one-dimensional semi-infinite
substrate material is assumed. This assumption can be justified after considering
the size of the absorption depth with the thickness of the workpiece. In this case,
the absorption depth is much smaller than the thickness of the workpiece.
2.3 Findings and Discussions 33
2.3.1 Stationary Source Consideration
The findings of the closed form solution for temperature and stress distributions
due to time exponentially decaying laser pulse and the convective boundary
condition at the surface due to phase change are considered. Steel is employed to
simulate the temperature and stress fields in line with the previous study [7].
Figure 2.2 shows the dimensionless temperature distribution within the sub-
strate material for various dimensionless heating periods. The influence of the heat
transfer coefficient on the temperature distribution becomes significant when the
dimensionless heat transfer coefficient reaches h 2:02 102 ð 108 W/m2 KÞ.
In this case, the temperature and its gradient in the surface region are reduced. The
temperature gradient in the surface region is reduced to its minimum. At the point
of minimum temperature gradient, the internal energy gain by the substrate from
the irradiated area is balanced by the diffusional energy transport from the sub-
strate to the solid bulk. In this case, the depth beyond the point of minimum
temperature gradient diffusional energy transport dominates over the internal
energy gain of the substrate material due to absorption of irradiated field. The point
of minimum temperature gradient changes with the heat transfer coefficient, which
is more pronounced for the heating period of 0.021. Moreover, the sharp decay in
the temperature gradient in the surface region ðx
0:1Þis because of: (i) the
absorption process, i.e. the absorbed energy decreases exponentially with
increasing depth (Lambert’s law), and (ii) the internal energy gain in the surface
region is high and diffusional energy transport due to the temperature gradient
from this region to the solid bulk is low, i.e. the increase in temperature due to
diffusional energy transport in the neighbouring region is low; therefore, the
temperature profile is governed by the internal energy gain in this region.
Figure 2.3 shows the dimensionless stress distribution within the substrate
material for different dimensionless heat transfer coefficients and times. The
thermal stress is zero at the surface as a result of the surface boundary condition
used in the analysis and it increases sharply close to the surface. The thermal stress
is tensile in this region due to expansion of the surface. As the depth increases
ðx [ 0:06Þ, the stress becomes compressive, as a result of the thermal strain
developed in this region, i.e., at this depth and beyond the material contracts
resulting in a compressive thermal stress field. The influence of the heat transfer
coefficient on the stress development is considerable, as illustrated by
h ¼ 0:0202ð109 W/m2 KÞ. In this case, the stress developed is compressive and
with a high magnitude in the vicinity of the surface and decays sharply as the depth
increases. However, the compressive stress wave is developed at some point below
the surface. The magnitude of the stress wave is lower at this point as time
progresses. In addition, the magnitude of the thermal stress levels, corresponding
to a heat transfer coefficient other than h ¼ 0:0202, increases with an increase in
time, provided that this increase is less than 10 %.
1.6E-02
Time = 0.015
TEMPERATURE 1.5E-02
h* = 2.02E-4
h* = 2.02E-3
1.4E-02
h* = 2.02E-2
h* = 2.02E-1
1.3E-02
1.2E-02
0.0 0.1 0.2 0.3 0.4 0.5
DISTANCE
2.0E-02
Time = 0.018
h* = 2.02E-4
1.8E-02
h* = 2.02E-3
TEMPERATURE
h* = 2.02E-2
1.6E-02 h* = 2.02E-1
1.4E-02
1.2E-02
0.0 0.1 0.2 0.3 0.4 0.5
DISTANCE
2.2E-02
2.0E-02
TEMPERATURE
1.7E-02 h* = 2.02E-4
h* = 2.02E-3 Time = 0.021
1.5E-02 h* = 2.02E-2
h* = 2.02E-1
1.2E-02
0.0 0.1 0.2 0.3 0.4 0.5
DISTANCE
Fig. 2.2 Dimensionless temperature distributions within the substrate material

2.0E-01
h = 2.02E-4
1.0E-01 h = 2.02E-3
h = 2.02E-2 Time = 0.015
0.0E+00
STRESS
h = 2.02E-1
-1.0E-01
-2.0E-01
-3.0E-01
0.0 0.1 0.2 0.3 0.4 0.5
DISTANCE
2.0E-01
h* = 2.02E-4
1.0E-01 h* = 2.02E-3
h* = 2.02E-2 Time = 0.018
0.0E+00
STRESS
h* = 2.02E-1
-1.0E-01
-2.0E-01
-3.0E-01
0.0 0.1 0.2 0.3 0.4 0.5
DISTANCE
2.0E-01
h* = 2.02E-4
1.0E-01 h* = 2.02E-3
Time = 0.021
h* = 2.02E-2
0.0E+00
STRESS
h* = 2.02E-1
-1.0E-01
-2.0E-01
-3.0E-01
0.0 0.1 0.2 0.3 0.4 0.5
DISTANCE
Fig. 2.3 Dimensionless stress distribution within the substrate material

2.3.2 Moving Source Consideration
Temperature field due to laser heating of sheet metal is formulated and thermal
stress field is computed using FEM in line with the previous study [9]. Laser beam
is considered as a line source scanning the workpiece surface with a constant speed
(v = constant). The laser beam diameter is taken as 1 mm.
Figure 2.4 shows temperature contours in the workpiece for two laser beam
scanning velocities. Temperature profile in the front region of the laser beam is
higher than that corresponding to the back region. This is because of the internal
energy gain of the substrate material, which is higher in the front region due to the
direction of the motion. Temperature reaches maximum when r/R is low and angular
location (h) is zero. This is the location corresponding to the front of the moving
source. Increasing r/R represents the locations departing away from the heat source,
i.e. r/R is the radial location from centre of the heat source. As the scanning velocity
increases, the magnitude of temperature reduces which is more pronounced in the
region close to the heat source front. In this case, increasing scanning velocity
reduces the internal energy gain of the substrate material, particularly in the region
close to the heat source.
Figure 2.5 shows radial stress contours for two laser scanning speed. Radial
stress component is compressive in the region close to the heat source. This is
because of the temperature distribution in this region. Radial stress component
becomes tensile as the radial distance increases further away from the heat source.
The stress behaviour in the substrate material is because of the thermal strain
(a) (b)
1 mm
Fig. 2.4 Temperature contours (oC) in the workpiece for two laser beam scanning velocities.
a – v = 1 cm/s. b - v = 5 cm/s
(a) (b)
1 mm
Fig. 2.5 Radial stress contours (MPa) for two laser scanning speed. a – v = 1 cm/s. b - v = 5 cm/s
(a) (b) 1 mm
Fig. 2.6 Tangential stress contours (MPa) for two laser scanning speeds. a – v = 1 cm/s. b - v =
5 cm/s
developed due to temperature field. The magnitude of radial stress component

increases slightly for low scanning speed and the compressive stress behaviour is
extended further away from the heat source as scanning speed increases to 5 cm/sec.
The compressive and tensile behaviour of the radial stress is because of the radial
temperature gradient.
Figure 2.6 shows tangential stress contours for two laser scanning speeds.
Tangential stress component is compressive in the region close to the heat source,
which is particularly true for high scanning speed. In this case, increasing scanning
speed results in extended compressive stress component in front of the heat source
while the stress component is tensile behind the heat source. Moreover, stress
distribution varies considerably when scanning speed changes. In this case, the
locations of zero stress line changes. It should be noted that similar behaviour is
observed for the radial stress component (Fig. 2.5), provided that the location of
zero stress line differs. Consequently, radial and tangential temperature distribu-
tion depends highly on the laser scanning speed, which in turn results in variation
in stress distribution.
2.4 Concluding Remarks
The findings from the closed form solution for temperature and stress fields reveal
that the influence of the heat transfer coefficient on temperature profiles is sig-
nificant as the dimensionless heat transfer coefficient at the surface increases. The
temperature gradient is reduced to its minimum at some point below the surface.
At a depth beyond the point of minimum temperature gradient the diffusional
energy transport dominates over the gain in internal energy of the substrate from
the irradiated area. The point of minimum temperature gradient changes for high
heat transfer coefficient. The thermal stress developed in the vicinity of the surface
is tensile and as the depth increases it becomes compressive. This is because the
thermal strain developed in the vicinity of the surface, which is positive, and at
some point below the surface, it becomes negative due to the compressive effect of
the substrate. The thermal stress wave is generated within the substrate material
and the magnitude of stress wave is reduced as the depth increases from the
surface towards the bulk solid.
The analytical solution for the moving source consideration shows that laser
scanning speed influences the temperature distribution, which in turn modifies the
stress field in the substrate material. Temperature distribution in front of the laser
heat source attains high values than that corresponding to behind the source. This
is because of the internal energy gain of the substrate material, which is high in the
region in front of the laser source. Moreover, as the scanning speed increases, the
magnitude of temperature away from the laser-heating source reduces. Radial
stress component is compressive in the region close to the heat source and
as distance increases away from the source, it becomes tensile. This is more
pronounced when scanning speed increases. This occurs because of the radial
distribution of the temperature gradient. Increasing scanning speed results in
extension of compressive stress field in front of the heat source. The similar
situation is observed from the tangential stress component.
References
1. Khan O, Yilbas BS (2004) Laser heating of sheet metal and thermal stress development.
J Mater Process Technol 155–156:2045–2050
2. Cheng P, Fan Y, Zhang J, Yao YL, Mika DP, Zhang W, Graham M, Marte J, Jones M (2006)
Laser forming of varying thickness plate—Part II: process synthesis. ASME J Manuf Sci Eng
128:642–650
3. Kalyon M, Yilbas BS (2002) Closed from Solution for exponentially decaying laser pulse
heating: evaporation at the surface. Japanese J Appl Phys 41:3737–3746
4. Yilbas BS, Sahin A, Davies R (1995) Laser heating mechanism including evaporation process
initiating the laser drilling. Int J Mach Tools Manuf 35(7):1047–1062
5. Yilbas BS, Ageeli N, Kalyon M (2004) Laser induced thermal stresses in solids: exponentially
time decaying pulse case. Laser Eng 14(1):81–101
6. Yilbas BS, Kalyon M (2001) Laser repetitive pulse heating with convective boundary
condition at the surface. J Phys D Appl Phys 34:222–231
7. Yilbas BS, Davies R, Yilbas Z (1991) A study into laser spot welding of sheet metals using
oxygen and argon as assisting gases. J Mater Process Technol 15:139–151
8. Yilbas BS, Aqeeli N (2006) Thermal stresses due to exponentially decaying laser pulse and a
convection boundary at the surface. Laser Eng 16:235–265
9. Khan O, Yilbas BS (2004) Laser heating of sheet metal and thermal stress development.
J Mater Process Technol 155(156):2045–2050
Chapter 3
Numerical Analysis for Laser Forming
and Welding
The numerical treatment of laser forming and welding process gives insight into
the physical processes taking place during the heating. Some of these processes
include solid heating and temperature gradient, phase change, and thermal stress
filed formed in the heated region. Since the mathematical constraints are les in the
numerical analysis, some of the assumptions imposed on the analytical solution of
the heating and thermal stress fields are relaxed in the numerical analysis. Some of
these assumptions include the consideration of semi-infinite solid and stationary
source. The numerical analysis related to the forming and welding process is
presented into two separate sub-headings. In addition, the numerical analysis
related to the forming is also categorized into two groups. The first group includes
the solid heating and stress field formation and the second group is related to the
phase change and the stress field formation. The numerical formulation given
below is presented in line with the previous studies [1–5].
3.1 Numerical Analysis for Laser Forming Process
The stress field can be generated with a moving laser beam source, which can be a
single or multiple beams heating at the surface. The surface temperature of the
substrate material can be kept below the melting temperature of the substrate
material through proper setting of the laser beam output power. Although ther-
mally induced displacement becomes small, in this case, displacement of in the
order of fraction of a millimeter can be achieved. This arrangement may become
sufficient for the optical applications. However, in some cases, the melting at the
substrate surface becomes unavoidable for requirement of the large thermally
induced displacement. In this case, the numerical analysis related to the phase
change should be incorporated and the stress analysis needs to be modified in the
formulations. Consequently, the numerical analysis related to the forming process
is presented under two-subheadings in line with the previous studies [1–3].

40 3 Numerical Analysis for Laser Forming
3.1.1 Solid Heating and Stress Field Formation
The numerical formulation of laser multi-beam heating of steel sheet surface is

presented. In this case, it is considered that the irradiated laser spots are located
along an arc to increase heated surface area during laser scanning at a constant
speed. Temperature and stress fields are predicted for various numbers of spot
along the arc.
Figure 3.1 shows laser heating situation and the coordinate system used in
simulations.
3.1.1.1 Thin Sheet Metal Heating Analysis
Laser heating of a steel sheet is considered such that laser scans the surface at a
constant speed (U) along the x-axis (Fig. 3.1). The enthalpy equation governing
laser heating process can be written as:
r ðqUH Þ ¼ r ðkrT Þ þ S0 ð3:1Þ
where H is the enthalpy, q is the density, U is the laser beam scanning velocity, k is
the thermal conductivity, and So is the volumetric source term, resembling the
absorption of the laser beam, due to the laser spots located around the arc. The
laser beam axis is the z-axis (Fig. 3.1) and the laser beam scans at the surface
along the x-axis. The volumetric source resembling the absorption of the incident
beam is:
0 qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi!2 1
X
n¼nss ðx xn Þ2 þðy yn Þ2
S0 ¼ I0 d 1 rf expðdzÞ exp@ A ð3:2Þ
a
n¼0
Fig. 3.1 A schematic view

of heating situation and the
orientation of laser spots at
the surface
y x
z
= 0
= 0.25
= 0
3.1 Numerical Analysis for Laser Forming Process 41
where Io is laser peak intensity, d is the absorption coefficient, a is the Gaussian

parameter ða ¼ 0:0002 m), rf is the surface reflectivity, x and y are the axes
(Fig. 3.1), nss ¼ ns1
2 (ns is total number of laser spots, ns = 7, 9, 11, 13, 25), n is
the laser spot number, xn and yn are defined
as:
xn ¼ Rarc 1 cos nss and yn ¼ Rarc 1 sin nss
nh nh
. Here, Rarc is the arc radius
(Fig. 3.2), h ¼ Rarc and s ¼ 2 . Laser intensity is kept the same for all spots in the
s Rarc
simulations.
In order to solve Eq. 3.1, two boundary conditions for each principal axis
should be specified. Convection heat loss is the same as the conduction flux at the
surface. The convection heat flux across the irradiated surface is:
oTðx; y; 0; tÞ
k ¼ ht ðTs T1 Þ ð3:3Þ
oz
where ht is the heat transfer coefficient at the free surface. The heat transfer
coefficient predicted earlier [6] is used in the present simulations across the heated
spots ðht ¼ 3 103 W/m2 K) due to the jet impingement and the other regions at
the surface, the natural convection is considered ðht ¼ 10 W/m2 K) to account for
the cooling. It should be noted that high value of heat transfer coefficient at the free
surface of the workpiece resembles the convection effect of assisting gas jet that
impinges onto the surface.
The other boundary conditions, therefore, are:
oT ðx; y; zth ; tÞ ht
z at bottom surface: z ¼ zth : ¼ Ts;b T1 ðspecified)
oz k
ð3:4Þ
where zth is the workpiece thickness, ht is the heat transfer coefficient due to
natural convection ðht ¼ 10 W/m2 K), and Ts;b is the surface temperature at the
workpiece bottom surface.
Fig. 3.2 Laser spots along the arc and top view of the workpiece
x at infinity ð 1Þ: x ¼ 1 : Tð1; y; z; tÞ ¼ To (specified) ð3:5Þ
oTðx; ywidth ; z; tÞ h
y at outer edge: y ¼ ywidth : ¼ Ts;w T1 ð3:6Þ
oy k
where ywidth is the half width of the workpiece and Ts;w is temperature at the outer
edge of the workpiece.
oTðx; 0; z; tÞ
y at the symmetry plane: y ¼ 0: ¼0 ð3:7Þ
oy
3.1.1.2 Thin Sheet Metal Thermal Stress Analysis
For structural response, finite element formulation is incorporated, which is based

on the principle of virtual work. From the principle of virtual work (PVW), a
virtual (very small) change of the internal strain energy (dU) must be offset by an
identical change in external work due to the applied loads (dV). Considering the
strain energy due to thermal stresses resulting from the constrained motion of a
body during a temperature change, PVW yields:
Z Z
T T T
fdug ½B ½D½Bdvfug ¼ fdug ½BT ½D eth dv ð3:8Þ
vol vol
Noting that the fdugT vector is a set of arbitrary virtual displacements common
in all of the above terms, the condition required to satisfy above equation reduces
to:

½K fug ¼ Fth ð3:9Þ
where
Z
½K ¼ ½BT ½D½Bdv ¼ Element stiffness matrix
vol
Z

th
F ¼ ½BT ½D eth dv ¼ Element thermal load vector ð3:10Þ
vol
th
e ¼ fagDT ¼ Thermal strain vector
fag ¼ vector of coefficients of thermal expansion
In the present study, effect of mechanical deformation on heat flow has been
ignored and the thermo-mechanical phenomenon of heating is idealized as a
sequentially-coupled unidirectional problem. Conducting solid element is used in
the thermal analysis. Conducting solid element is also analyzed structurally;
therefore, the element is replaced by an equivalent structural element for the
structural analysis. The element has plasticity, creep, swelling, stress stiffening,
large deflection, and large strain capabilities.
Laser forming process can be extended to include the three dimensional anal-
ogy. In this case, the workpiece thickness is considered to be large enough to affect
the stress levels and the thermal displacement in the substrate material. The
heating analysis and the boundary conditions change as compared the thin sheet
plates. Therefore, the heating and stress analyses are presented below in line with
the previous study [2].
3.1.1.3 Thick Sheet Metal Heating Analysis
The laser heating is formulated using the Fourier heating law for constant speed
workpiece. The enthalpy equation governing the laser heating process can be
written as:
r ðqUH Þ ¼ r ðkrT Þ þ So1 þ So2 þ So3 ð3:11Þ
where H is the enthalpy, q is the density, U is the laser beam scanning velocity, k is
the thermal conductivity, and So1 is the volumetric source term, resembling the
absorption of the laser beam due to the first laser spot, So2 is the source term due to
the second spot, and So3 is the source term due to the third spot, respectively.
The laser beam axis overlays along the z-axis (Fig. 3.3) and the laser beam scans
at the surface along the x-axis. The volumetric source resembling the absorption of
the incident beam is:
ðx x1 Þ2 þðy y1 Þ2
So1 ¼ Io1 d 1 rf expðdzÞ exp@ A ð3:12Þ
a
and
a
and
a
where Io1 is the peak intensity at the first and Io2 is peak intensities at the second
and third spots, d is the absorption coefficient, a is the Gaussian parameter, rf is the
surface reflectivity, x and y are the axes (Fig. 3.3), x1 is the distance between the
origin of the coordinate system and the first beam center, and x2 is the distance
between the origin of the coordinate system and the second beam center along the
Laser Beams
U
y x
z
Symmetry Plane
Laser Spots Symmetry Line
Fig. 3.3 Schematic view of three-beam laser heating and the coordinate system
x-axis (Fig. 3.3). Moreover, y1 is the distance between the origin of the coordinate
system and the second beam center along the y-axis, and y2 is the distance between
the origin of the coordinate system and the third beam center along the y-axis. It
should be noted that jy1 j ¼ jy2 j. The laser beam axis is the z-axis (Fig. 3.3). The
laser intensity is kept the same for the second and the third spots in the simulations.
In order to solve Eq. 3.11, two boundary conditions for each principal axis
should be specified. Convection heat loss is the same as the conduction flux at the
surface. The convection heat flux across the irradiated surface is:
oTðx; y; 0; tÞ
k ¼ ht ðTs T1 Þ ð3:15Þ
oz
where ht is the heat transfer coefficient at the free surface. The heat transfer
coefficient predicted earlier [6] is used in the present simulations across the heated
spots ðht ¼ 3 103 W/m2 K) due to the jet impingement and the other regions at
the surface, the natural convection is considered ðht ¼ 10 W/m2 K) to account for
the cooling. It should be noted that high value of heat transfer coefficient at the free
surface of the workpiece is due to the assisting gas jet impingement at the surface.
The other boundary conditions, therefore, are:
oT ðx; y; zth ; tÞ ht
z at bottom surface: z ¼ zth : ¼ Ts;b T1 ðspecified)
oz k
ð3:16Þ
where zth is the workpiece thickness, ht is the heat transfer coefficient due to
natural convection ðht ¼ 10 W/m2 K), and Ts;b is the surface temperature at the
workpiece bottom surface.
x at infinity ð 1Þ: x ¼ 1: Tð1; y; z; tÞ ¼ To ðspecified) ð3:17Þ
oTðx; ywidth ; z; tÞ h
y at outer edge: y ¼ ywidth : ¼ Ts;w T1 ð3:18Þ
oy k
where ywidth is the half width of the workpiece and Ts;w is temperature at the outer
edge of the workpiece.
oTðx; 0; z; tÞ
y at the symmetry plane: y ¼ 0: ¼0 ð3:19Þ
oy
3.1.1.4 Thick Sheet Metal Stress Analysis

The elastic strain increment vector, Deel is related to the stress increment vector,
fDrg by Hooke’s law:

fDrg ¼ ½D Deel ð3:20Þ
where ½D contains the elastic constants related to temperature-dependent elastic
modulus, E and Poisson’s ratio, m. The total strain vector, fDeg, may be expressed
as follows:

fDeg ¼ Deel þ Deth Depl ð3:21Þ
el th
where De is the strain increment vector, De
elastic is the thermal strain
increment vector, Depl is the plastic strain increment vector. The incremental

thermal strain vector, Deth arises from the volume changes that accompany the
temperature increment, DT, which is calculated by the thermal analysis. It is
normally accounted for in stress analyses through a temperature-dependent dif-
ferential thermal expansion coefficient, a(T). The differential and total thermal
expansion (aT ) coefficients are related to each other through:
ZT
1
aT ðTÞ ¼ aðTÞdT ð3:22Þ
T To
To
where T o is a reference temperature designating the point at which the material

exhibits no dilatational strain.
The coupled thermal stress analysis requires identifying the displacement–strain
relations, which are expressed in dimensionless form as follows [7]:
o
u ov o
w
exx ¼ ; eyy ¼
; ezz ¼ ð3:23Þ
oxoy oz

1 ou ov 1 ov ow 1 ow ou
exy ¼ þ ; eyz ¼ þ ; ezx ¼ þ ð3:24Þ
2 oy ox 2 oz oy 2 ox oz
An exact implementation of Newton’s method involves a nonsymmetrical
Jacobian matrix which is stress–strain relation in dimensionless form as is illus-
trated in the following matrix representation of the coupled equations [7]:
8 9 2 38 9 8 9
> r
xx > 1m m m 0 0 0 > e 1
>
> >
> > xx > > > > > >
>r
> >
yy > 6 m 7>> eyy >> > > 1>
>
>
> >
> 6 1m m 0 0 0 7>> >
> >
> >
>
>
<r > 6 7>> >
> >
> >
>
zz = E 6 m m 1m 0 0 0 7< ezz = < 1 = T
aE
¼ 6 7
> r
> ð 1 þ m Þ ð 1 2m Þ 6 0 0 0 1 2m 0 0 7 > e > > 0 > 1 2m
> yz >
> > 6 7> yz > > >
>
> >
> 6 7>> > > > > > >
> r

> zx >
>
>
> 4 0 0 0 0 1 2m 0 5> >
>
> e >
>
zx >
>
> 0
> >
>
>
: ; : > ; > : > ;
r
xy 0 0 0 0 0 1 2m exy 0
ð3:25Þ
Solving this system of equations requires the use of the unsymmetrical matrix
storage and solution scheme. von Mises (equivalent stress) stress can be written as:
1 h 2 2
i1=2
reqv ¼ pffiffiffi rx ry þ ry rz þðrz rx Þ2 þ6 s2xy þ s2yz þ s2zx ð3:26Þ
2
Furthermore, the mechanical and thermal equations are solved simultaneously.
3.1.1.5 Numerical Solution
Finite element model is used to solve governing equations heat transfer and
thermal stress. COMSOL Finite Element Code [8] is used in the simulations. In the
case of thin plate, 3-D heating model, and non-uniform rectangular grid is used
with 100 150 cells. The grids are dense near the heat sources in order to
accurately resolve for temperature and stress distributions. The solver was run
until the converged results were obtained. In this case, the residual error for the
energy equation was less than the limit set in the simulations, which is 10-8.
It should be noted that the error related to the predictions is minimized through
setting the residual error in the energy equation in COMSOL [8]. In the case of
three-dimensional heating situation, a non-uniform rectangular grid is used with

100 150 100 cells. The grids are dense near the two heat sources in order to
accurately resolve for temperature distribution. The solver was run until the
converged results were obtained. In this case, the residual error for the energy
equation was less than the limit set in the simulations. It should be noted that the
error related to the predictions is minimized through setting the residual error in
the energy equation in COMSOL [8]. In the simulations a laser delivering three
beams with the varying intensity is used.
3.1.2 Phase Change and Stress Field Formation
The phase change at the irradiated region of the workpiece requires modification
of the heat equation. In this case, specific heat capacity of the workpiece is
modified to include the latent heat of enthalpy across the solidus and liquidus
temperatures. The below analysis is given in the light of the previous study [3].
3.1.2.1 Heating Analysis
In the analysis, the solid body heat conduction with temperature-dependent con-
ductivity, internal energy (including latent heat effects due to phase change), and
convection and radiation boundary conditions are considered. The Fourier heat
transfer equation for the laser heating process can be written as:
DE
q ¼ ðrðkrTÞÞ + So ð3:27Þ
Dt
where E is the energy gain of the substrate material, k is the thermal conductivity,
and So is the heat source term resembling the laser beam, i.e.:
x2 þy2
So ¼ Io 1 rf e a2 ð3:28Þ
Io is laser power peak density, a is the Gaussian parameter, rf is the surface
reflectivity, q is the density, and x and y are the axes while the laser beam scans the
surface along the x-axis. The laser beam axis is the z-axis (Fig. 3.4). It should
be noted that the laser beam intensity distribution is assumed to be Gaussian at the
irradiated surface.
In the case of a moving heat source along the x-axis with a constant velocity U,
energy gain of the substrate material yields:
DE oE oE
q ¼q þ qU ð3:29Þ
Dt ot ox
Fig. 3.4 Schematic view of laser heating situation and coordinate system
or
DE oðCpTÞ oðCpTÞ
q ¼q þ qU ð3:30Þ
Dt ot ox
Combining Eqs. 3.27 and 3.30 yields:
oðCpTÞ oðCpTÞ
q ¼ ðrðkrTÞÞ qU þSo ð3:31Þ
ot ox
Equation 3.31 is solved numerically with the appropriate boundary conditions
to predict the temperature field in the substrate material. However, to analyze the
phase change problem, the enthalpy method is used [9]. The specific heat capacity
is associated with the internal energy gain of the substrate material, i.e.
CpðTÞ ¼ oEoT . However, the internal energy gain during the phase change is asso-
ciated with the latent heat of fusion, which is given separately in terms of solidus
and liquidus temperatures (the lower and upper temperature bounds of the phase
change range) and the total internal energy associated with the phase change,
called the latent heat [9].
3.1.2.2 Stress Analysis

The elastic strain increment vector, Deel is related to the stress increment vector,
fDrg by Hooke’s law:

fDrg ¼ ½D Deel ð3:32Þ
where ½D contains the elastic constants related to temperature-dependent elastic

modulus, E and Poisson’s ratio, m. The total strain vector, fDeg, may be expressed
as follows:

fDeg ¼ Deel þ Deth Depl ð3:33Þ
el th
where De is the strain increment vector, De
elastic is the thermal strain
increment vector, Depl is the plastic strain increment vector. The incremental

thermal strain vector, Deth arises from the volume changes that accompany the
temperature increment, DT, which is calculated by the thermal analysis. It is
normally accounted for in stress analyses through a temperature-dependent
differential thermal expansion coefficient, a(T). In ABAQUS/Standard analysis, a
spatially varying thermal expansion can be defined for homogeneous solid con-
tinuum elements by using a distribution, which includes the tabulated values for
the thermal expansion [9]. ABAQUS uses an implicit backward-difference scheme
for time integration of both temperature and displacements at every material
integration point [9].
3.2 Numerical Analysis for Laser Welding Process
In laser welding process, the deep penetration of laser beam is required and the
phase change takes place not only at the surface, but it is extended to include
inside the substrate material. In this case, laser beam moves with a constant speed
while the workpiece is stationary. The process involves the phase change and
the heat conduction from the melt pool to the surroundings. In order to assess the
physical insight of the process, the phase chase change and cavity formation at the
surface is formulated. In order to achieve the deep penetration of the laser beam
into the substrate partial evaporation of the substrate material at the surface is
unavoidable. In addition, during the melting and evaporation process, mushy zone
is generated, which needs to be incorporated in the analysis. Consequently, the
formulation of phase change using the energy method is introduced first and
enthalpy porosity method pertinent to the phase change process are presented
below in the light of the previous studies [4, 5].
3.2.1 Phase Change Including Melting and Evaporation
Energy equation for each phase is modelled independently as well as being cou-
pled across the interfaces of the two-phases, where both phases exist mutually
(mushy zones). In the formulations, the laser output power intensity distribution at
the workpiece surface is considered being Gaussian and its centre is located at the
centre of the co-ordinate system. The temporal variation of laser power intensity
resembling the actual laser pulse is accommodated in the simulations. This
arrangement results in an axisymmetric heating of the workpiece material. The
heat conduction equation for a solid phase heating due to a laser irradiation pulse
with a Gaussian intensity profile can be written as:

oT ks o oT o2 T
qs c p s ¼ r þ k s 2 þ So ð3:34Þ
ot r or or oz
where So is the volumetric source term and it is:
2
r
So ¼ Io d 1 rf expðdzÞ exp 2 ð3:35Þ
a
Io ; d; rf and a are the laser peak power intensity, absorption coefficient,
reflectivity and the Gaussian parameter, respectively.
The initial and boundary conditions for Eq. 3.34 are given below. Initially, the
substrate material is assumed at a uniform temperature, To , i.e.:
At time zero ) t ¼ 0 : Tðr; z; 0Þ ¼ To ðspecified) ð3:36Þ
At a distance considerably away from the surface (at infinity) in the radial
direction a constant temperature To is assumed. Since the heating has no effect on
the temperature rise at a depth of infinity below the surface, temperature is
assumed to be constant and equals to the initial temperature of the substrate
material in this region. The respective boundary conditions are:
r at infinity ) r ¼ 1 : Tð1; z; tÞ ¼ To ðspecified) ð3:37Þ
and
z at infinity ) z ¼ 1 : Tðr; 1; tÞ ¼ To ðspecified) ð3:38Þ
At the symmetry axis, maximum temperature is assumed and the convection
boundary condition is considered at the workpiece surface, i.e.:
oTð0; z; tÞ
At symmetry axis ) r ¼ 0 : ¼0 ð3:39Þ
or
and
oTðr; 0; tÞ
At the surface ) z ¼ 0 : k ¼ hðTðr; 0; tÞ To Þ ð3:40Þ
oz
where h is taken as 10 W/m2K due to natural convection from the surface.
Since the evaporation temperature depends on the pressure and this relation is
not known for steel vapor, it is assumed that the substrate material has single
melting and boiling temperatures. Moreover, once the phase change initiates, a
mushy zone (mutually existence of two-phases) is introduced across the interface
3.2 Numerical Analysis for Laser Welding Process 51
of two phases. During the phase change process, including the mushy zone,
temperature of the substrate material remains the same, but its enthalpy changes in
this region. This situation can be formulated via energy balance in the mushy zone.
It should be noted that nominal laser pulse length is 24 ns, therefore, the flow in
the vapor and liquid layer during the heating process is neglected, i.e. 1 lm of fluid
motion, in radial or axial direction, in the liquid layer requires the liquid velocity
in the order of 1000 m/s, which may not be the case that occurs in the present
simulations; consequently, during the short heating period, the fluid motion in the
liquid and vapor layers is assumed not to influence the heat transfer mechanism in
these zones. Consider a differential element in the mushy zone (Fig. 3.5) and let xm
is the mass fraction of liquid present in the element. Energy content ðDU Þ of the
differential element with volume D8 at melting temperature Tm can be written as:

DU ¼ qm D8 xm Lm þ cpm Tm Tref þ cps ð1 xm Þ Tm Tref ð3:41Þ
where,
mm
xm ¼ ð3:42Þ
mm þ ms
Tref ; xm ; mm ; ms are reference temperature for enthalpy, quality (mass fraction)
of liquid, mass of liquid and mass of solid in the element, respectively.
After assuming
specific heat of melt is the same as the solid at the melting
temperature cps ¼ cpm at T ¼ Tm , and differentiation of Eq. 3.41 with time
yields:
ou oxm
¼ qm Lm ð3:43Þ
ot ot

since cpm Tm Tref ¼ const: and u ¼ U8 . Substituting Eq. 3.43 into Eq. 3.34 and
re-arrangement results the energy equation for the differential element in the
mushy zone:

oxm km o oT o2 T
qm Lm ¼ r þ k m 2 þ So ð3:44Þ
ot r or or oz
Equation 3.44 is valid in the mushy zone where 0 xm 1, i.e. temperature of
the cells with 0 xm 1 is set to melting temperature ðT ¼ Tm Þ. For the situation
xm ¼ 1, liquid phase occurs and Eq. 3.34 is used to determine the temperature rise
in the liquid heating with using the liquid thermal properties in the equation.

of a mushy zone Mushy Cell Mushy Cell Mushy Cell
Mushy Cell
Moreover, the liquid heating continues until the boiling point is reached in the
substrate material; in which case, a new mushy zone is formed. In this case,
Eq. 3.44 is modified for a differential element in the mushy zone subjected to
evaporation, i.e.:

oxb kb o oT o2 T
qb Lb ¼ r þ kb 2 þ S o ð3:45Þ
ot r or or oz
Equation 3.45 is applicable for temperature T ¼ Tb and 0 xb 1 in the
mushy zone (partially liquid partially vapor zone); in which case, temperature of
the cells with 0 xb 1 is set to the boiling temperature of the substrate material
ðT ¼ Tb Þ. It should be noted that xm is replaced with xb in Eq. 3.44, which rep-
resents the fraction of vapor phase in the differential element. The calculation of xb
is the same as xm , provided that latent heat of fusion is replaced with latent heat of
evaporation of the substrate material in Eq. 3.44 in the later.
The boundary condition at the evaporating surface is introduced in relation to
Eq. 3.45. In this case, the temperature along the evaporated surface is kept at
boiling temperature of the substrate material, i.e., the cells in the evaporated region
are kept at boiling temperature, i.e.:
In the mushy zone, at z ¼ zb ) Tðr; zb ; tÞ ¼ Tb ð3:46Þ
where zb represents the axial location at the evaporated surface.
Equations 3.44 and 3.45 provide the relative position of solid–liquid and
liquid–vapor interface in the substrate material. Liquid–vapor interface determines
the shape and size of the cavity generated during evaporation process.
3.2.2 Transiently Developing Vapor Jet
The vapor front emerging from the laser induced cavity is modeled numerically
using a control volume approach and the fluid dynamic/mass transfer model is
accommodated in the analysis. The laser produced cavity shape and its temporal
progression are employed in the simulations. In this case, the time-varying cavity
shape, mass flux of the vapor and the temperature distribution at the cavity surface
are the inputs for the simulations. In the flow analysis, the time averaged con-
servation equations are accommodated for an unsteady, incompressible, axisym-
metric turbulent flow situation resembling the vapor jet expansion. The Standard
K-e turbulence model is used to account for the turbulence. Moreover, the species
transport model is also used to account for the mass transfer of the vapor jet from
the cavity into the stagnant water ambient. It should be noted that all the unknown
quantities are time-averaged since the RANS equations are used.
Continuity Equation:
1 oðrVr Þ oVz
þ ¼0 ð3:47Þ
r or or
Radial momentum:

oðqVr Þ 1 o qrVr2 oðqVr Vz Þ op 2 o oVr o oVr
þ þ ¼ þ l r þ l
ot r or oz or r or eff or oz eff oz
o oVz Vr
þ l 2leff 2
oz eff oz r
ð3:48Þ
Axial momentum:

oðqVz Þ 1 oðqrVr Vz Þ o qVz2 op 1 o oVz o oVz
þ þ ¼ þ l r þ2 l
ot r or oz oz r or eff or oz eff oz
1o oVr
þ leff r
r or oz
ð3:49Þ
where
qCl K 2
leff ¼ l þ lt ; lt ¼ ; Cl ¼ 0:09 ð3:50Þ
e
Energy Equation:

oðqEÞ 1 oðrVr qEÞ oðVz qEÞ 1 o oT o oT
þ þ ¼ rkeff þ keff
ot r or oz r or or oz oz

1o l oYvapor o l oYvapor
þ rhvapor qD þ t þ hvapor qD þ t
r or Sct or oz Sct oz

1o l oYwater o l oYwater
þ rhwater qD þ t þ hwater qD þ t
ð3:51Þ
where E ¼ Ywater hwater þ Yvapor hvapor , after neglecting the contribution of kinetic
energy. Enthalpy of vapor and water are:
ZT

hvapor ¼ Cpvapor dT ¼ Cpvapor T Tref ð3:52Þ
Tref
and
ZT

hwater ¼ Cpwater dT ¼ Cpwater T Tref ð3:53Þ
Tref
In Eqs. 3.51–3.53, the properties and parameters are:

lt
keff ¼ k þ kt ; kt ¼ Cp ; and D ¼ 2:88 105 m2 s ;
Prt
Sct ¼ 0:7; Pr ¼ 0:85; Tref ¼ 298:15 K ð3:54Þ
t
Turbulence Kinetic Energy Equation, K:

oðqK Þ 1 oðqrVr K Þ oðqVz K Þ 1 o leff oK o leff oK
þ þ ¼ r þ qe þ PK
ot r or oz r or rK or oz rK oz
ð3:55Þ
Rate of Dissipation Equation, e:

oðqeÞ 1 oðqrVr eÞ oðqVz eÞ 1 o leff oe o leff oe e
þ þ ¼ r þ C1 PK
ot r or oz r or re or oz re oz K
e2
C2 q
K
ð3:56Þ
where,
" ( ) #
oVz 2 oVr 2 Vr 2 oVz oVr 2
PK ¼ leff 2 þ þ þ þ ð3:57Þ
or or r or oz
and
rK ¼ 1; re ¼ 1:3; C1 ¼ 1:44; and C2 ¼ 1:92 ð3:58Þ
Species Transport Equation:

o qYvapor 1 o rVr qYvapor o Vz qYvapor
þ þ
ot r or oz
1o lt oYvapor o l oYvapor
¼ r qD þ þ qD þ t ð3:59Þ
where Ywater ¼ 1 Yvapor .
Initial and Boundary Conditions:
Figure 3.6 shows the solution domain.
Symmetry Axis: ðr ¼ 0Þ
Fig. 3.6 Solution domain of

an axisymmetric jet
expansion emanating from
cavity
At the symmetry axis all the unknown quantities are considered to be maximum
accept the r-direction velocity, which is zero.

oVz oT oK oe
¼ 0; Vr ðz; 0Þ ¼ 0; ¼ 0; ¼ 0; ¼ 0;
or z¼z or z¼z or z¼z or z¼z
r¼0 r¼0 r¼0 r¼0 ð3:60Þ
oYvapor
¼ 0
or z¼z
r¼0
Outflow: ðz ¼ 0Þ
At the outflow boundary perpendicular to the z-axis the normal derivatives of
all the unknown quantities are considered to be zero accept the r-direction velocity,
whose value is zero as required from the continuity equation.

oVz oT oK oe
¼ 0; Vr ð0; r Þ ¼ 0; ¼ 0; ¼ 0; ¼ 0;
oz z¼0 oz z¼0 or z¼0 or z¼0
r¼r r¼r r¼r r¼r ð3:61Þ
oYvapor
¼ 0
or z¼0
r¼r
Outflow: ðr ¼ rmax Þ
At the outflow boundary perpendicular to the r-axis the normal derivatives of all
the unknown quantities are considered to be zero accept the z-direction velocity,
whose value is zero as required from the continuity equation.

oVr oT oK
Vz ðz; rmax Þ ¼ 0; ¼ 0; ¼ 0; ¼ 0;
or z¼z or z¼z or z¼z
r¼rmax r¼r max r¼rmax
ð3:62Þ
oe oYvapor
¼ 0; ¼ 0
or z¼z or z¼0
r¼rmax r¼rmax
Solid Wall: z ¼ f ðr; tÞ

The surface of the substrate material including the cavity surface acts like a
solid wall in the solution domain and hence a no-slip and no-temperature jump
boundary conditions are considered. At the cavity surface, the vapor mass fraction
is considered to be one whereas the water mass fraction is considered to be zero. In
this case, the function f ðr; tÞ defining the cavity wall shape, as obtained from the
heat transfer analysis, can be presented in algebraic form. It should be noted that
the temporal development of the cavity in axial and radial directions can be
computed using above equations. Once the cavity profile in axial and radial
directions is obtained for each time step, an algebraic equation is introduced, using
the polynomial fitting technique, to resemble the temporal behavior of the cavity
shape through using the data computed for time steps. Moreover, the algebraic
equation resembling the cavity shape is in the functional form such that time
(t) and radial location (r) are set as independent variables for fixed axial ðDzÞ and
radial space ðDr Þ increments. Consequently, for each heating time step and defined
radial locations, cavity shape can be obtained from the functional relation. This
arrangement is necessary due to in the solution of flow equations; solid wall at the
cavity surface should be defined. In addition, this functional arrangement of the
temporal behavior of cavity surface enables to solve flow equations. The vapor
mass generated due to recession of the cavity surface varies in time as well as in
space coordinates. The vapor mass flow generated during cavity surface recession
is the source of the mass of the vapor jet emanating from the cavity surface.
Equation resembling the temporally recessing cavity surface is:
h

i
f ðr; tÞ ¼ c ðr=Dr Þ2 rmax
2
þ d ðr=Dr Þ3 rmax
3
ðql =qv ÞDz ð3:63Þ
where the coefficients c and d can be written as:
c ¼ cc3 þ cc2 ðt=DtÞ þ cc1 ðt=DtÞ2 ð3:64Þ

where,
cc1 ¼ 4:4499 106 : cc2 ¼ 0:0026726 : cc3 ¼ 0:15077 ð3:65Þ

and
d ¼ cd3 þ cd2 ðt=DtÞ þ cd1 ðt=DtÞ2 ð3:66Þ

where,
cd1 ¼ 1:7974 107 : cd2 ¼ 0:00011187 : cd3 ¼ 0:0062477 ð3:67Þ

and
rmax ¼ cr5 þ cr4 ðt=DtÞ þ cr3 ðt=DtÞ2 þcr2 ðt=DtÞ3 þcr1 ðt=DtÞ4 ð3:68Þ
where
cr1 ¼ 2:5982 109 : cr2 ¼ 2:0455 106 : cr3 ¼ 0:00059955 :

cr4 ¼ 0:086871 : cr5 ¼ 7:7763
ð3:69Þ
The numerical values for the space increments are:
Dz ¼ 3:2415 108 m; Dr ¼ 8:3333 108 m ð3:70Þ

The recession velocity of the cavity wall as obtained from the above analysis
can be represented in the algebraic form. In this case, the cavity size is limited with
0 r rmax . Therefore, the recession velocity of the cavity along the vertical
direction is:
ql of of =or
Vz ðf ðr; tÞ; r Þ ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi for 0 r rmax ð3:71Þ
qv ot
1 þ ðof =or Þ2
Moreover, outside of the cavity a stationary solid wall is considered (Fig. 3.6).
The recession velocity along the vertical axis is, therefore:
Vz ðf ðr; tÞ; r Þ ¼ 0 for r [ rmax ð3:72Þ
The recession velocity of the cavity along the radial direction is:
ql of 1
Vr ðf ðr; tÞ; r Þ ¼ qffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffiffi for 0 r rmax ð3:73Þ
qv ot
1 þ ðof =or Þ2
Outside of the cavity a stationary solid wall is considered (Fig. 3.6). The
recession velocity along the radial direction is, therefore:
Vr ðf ðr; tÞ; r Þ ¼ 0 for r [ rmax ð3:74Þ
Temperature at the cavity wall is determined from the heat transfer analysis and
can be presented in algebraic form. In this case, temperature at the cavity wall is
the same as the boiling temperature of the substrate material ðTb Þ, i.e.:
T ðf ðr; tÞ; r Þ ¼ Tb
0 r rmax
2 ð3:75Þ
T ðf ðr; tÞ; r Þ ¼ 2834 exp btempððr=Dr Þ rmax Þ þ 300 r [ rmax
where
btemp ¼ ctempðt=DtÞ2 þdtempðt=DtÞ þ etemp; ctemp ¼ 1:0370442955011 106

dtemp ¼ 3:39682806506743 105 ; etemp ¼ 0:0241815719639816
ð3:76Þ
Turbulence kinetic energy and dissipation of the jet in the cavity wall region is
assumed to be constant and taken as:
K ðf ðr; tÞ; r Þ ¼ 1 (m/s)2 and eðf ðr; tÞ; r Þ ¼ 1 (m/s)2 ð3:77Þ

and
Yvapour ðf ðr; tÞ; r Þ ¼ 1 0 r rmax
ð3:78Þ
Yvapour ðf ðr; tÞ; r Þ ¼ 0 r [ rmax
Initial Conditions:
Initially the ambient water is assumed as stagnant; therefore, the z and
r-directions velocity components are zero. Initially, temperature is considered to be
uniform and equal to 300 K in water ambient and the vapor mass fraction is zero
whereas the water mass fraction is one throughout the domain.
Vz ðz; r Þ ¼ 0; Vr ðz; r Þ ¼ 0; T ðz; r Þ ¼ 300 K; K ðz; r Þ ¼ 1; eðz; r Þ ¼ 1;

Yvapour ðz; r Þ ¼ 0
ð3:79Þ
3.2.3 Numerical Solution
3.2.3.1 Phase Change Process
Equation 3.34 is applicable to solid and liquid heating, Eq. 3.44 is applicable to
mushy zone at solid–liquid interface and Eq. 3.45 is applicable to mushy zone at
liquid–vapor interface. To discretize the governing equations, a finite difference
scheme is introduced. The details of the numerical scheme are given in [10].
To compute the equations discretized for temperature field and relative positions
of solid–liquid and liquid–vapor interface, an implicit scheme is used, i.e. using
the initial conditions, the temperature in the whole domain is calculated for fol-
lowing time steps with the respective conditions.
The calculation domain is divided into grids and grid independence test is
performed for different grid size and orientation and the grid size resulting grid
independent solution is used, which is 100 9 120 mesh points in the r and z-axes.
A computer program based on implicit scheme is developed to compute the
temperature field.
3.2.3.2 Transiently Developing Vapor Jet (Front)
A control volume approach is employed when discretizing the governing equations

[11]. A staggered grid arrangement is used in which the velocities are stored at a
location midway between the grid points, i.e. on the control volume faces. All other
variables including pressure are calculated at the grid points. This arrangement
gives a convenient way of handling the pressure linkages through the continuity
equation and is known as Semi-Implicit Method for Pressure-Linked Equations
(SIMPLE) algorithm. The details of this algorithm are given in Ref. [11].
The computer program used for the present simulation can handle non-uniform
grid spacing. Along the radial direction fine uniform grid spacing is allocated at the
inlet (in cavity symmetry axis region) while gradually increasing spacing is con-
sidered away from the inlet (in the cavity edge region). Along the axial direction,
again fine uniform grid spacing is used inside and near the cavity while the grid
spacing gradually increases away from the cavity. The number of grid points in the
radial direction is 300 while 215 grid points are used in the axial direction. Since
the problem is involved with the moving boundary, the moving meshes are
accommodated in the cavity to account for the cavity recession with time. The grid
independence test is conducted and grid size (215 300) resulting in grid inde-
pendent solution is used in the simulations.
Eight variables are computed at all grid points. These are the two velocity
components, local pressure, two turbulence quantities, temperature and two mass
fractions.
3.2.4 Analysis for Laser Welding
3.2.4.1 Heating Analysis
Figure 3.7 shows the schematic view of the laser welding process and the coordinate
system. In the analysis, the solid body heat conduction with temperature-dependent
conductivity, internal energy (including latent heat effects), and convection and
radiation boundary conditions are considered [5].
The details of the mathematical arguments are given in [12]. The Fourier heat
transfer equation pertinent to the laser heating process can be written as:
oHðTÞ
q ¼ ðr ðkðTÞrTÞÞþSo ð3:80Þ
ot
where H is the temperature dependent enthalpy including the latent heat of
solidification, k is the temperature dependent thermal conductivity, and So is the
heat source term resembling the laser beam, i.e.:
dz
x2 þy2
So ¼ Io de 1 r f e a2 ð3:81Þ
Io is laser power peak density, d is the absorption coefficient, a is the Gaussian
parameter, rf is the surface reflectivity, q is the density, and x and y are the axes.
The absorption coefficient of the laser beam ðdÞ after the key-hole formation is
considered, i.e.:
Fig. 3.7 A schematic view of laser welding and coordinate system

1 Io
d ¼ ln ð3:82Þ
L IL
where L is the thickness of the workpiece, Io is the peak power intensity at the
workpiece surface, IL is the laser power intensity at the workpiece thickness. The
laser beam axis is parallel to the z-axis (Fig. 3.7). It should be noted that the laser
beam intensity distribution is assumed to be Gaussian at the irradiated surface.
The convective and radiation boundary conditions are considered at the free
surface of the workpiece. Therefore, the corresponding boundary condition is:
oT hf er 4 4

At the irradiated surface ðtop surfaceÞ: ¼ ðTs Tamb Þ þ T Tamb
oz k k s
ð3:83Þ
where hf ¼ 3000 W/m2 K [6] is the forced convection heat transfer coefficient due
to the assisting gas.
At the bottom surface:
oT h er 4 4
oT h er 4 4

¼ ðTs Tamb Þ þ Ts Tamb ; ¼ ðTs Tamb Þ þ Ts Tamb
oy k k oz k k
ð3:84Þ
where h ¼ 20 W/m2 K is the heat transfer coefficient due to natural convection,
and Ts and Tamb are the surface and ambient temperatures, respectively, e is the
emissivity (e = 0.9 is considered), r is the Stefan-Boltzmann constant
(r = 5.67 9 10-8 W/m2K). At far away boundary (at edges of the solution
domain) constant temperature boundary is assumed (T = 293 K), i.e.
x ¼ 1; y ¼ 1; z ¼ th ! T ¼ 293 K ð3:85Þ
where th is the thickness of the workpiece. Initially (prior to laser welding), the
substrate material is assumed to be at constant ambient temperature, i.e. T = Tamb,
which is considered as constant (Tamb = 293 K).
Equation 3.80 can be solved numerically with the appropriate boundary con-
ditions to predict the temperature field in the substrate material. However, to
analyze the phase change problem, the enthalpy method is used [9]. The specific
heat capacity is associated with the internal energy gain of the substrate material,
i.e. CpðT Þ ¼ oU
oT . However, the internal energy gain during the phase change is
associated with the latent heat of fusion, which is given separately in terms of
solidus and liquidus temperatures (the lower and upper temperature bounds of the
phase change range) and the total internal energy associated with the phase
change, called the latent heat [9].
The molten flow takes place due to surface tension and density variation in the
melt pool. Therefore, the conservation equations need to be incorporated to
account for the flow field. The conservation equations are:
oq
Continuity þ r ðqvÞ ¼ 0
ot

ov
Momentum q þ v rv ¼ rp þ lr2v þ gbE T þ Sz ð3:86Þ
ot
oH
Energy q þ r ðqvH Þ ¼ r ðkrTÞ þ So Sh
ot
where v is the velocity vector in the melt pool, p is the pressure, bE is
the volumetric thermal expansion coefficient, which is formulated using the
Boussinesq approximation, Sz is the momentum sink, and Sh is a phase related
source term convection–diffusion phase change (Sh ¼ oðqDH ot
sÞ
þ r ðqvDHs Þ,
where DHs ¼ H qCp T) [13]. The term gbE T is the buoyancy term, which is used
to introduce natural convection in the melt pool.
3.2.4.2 Stress Analysis
Solidification involves small strain, so the assumption of small strain is adopted in

this work. The thermal strains which dominate thermo-mechanical behavior during
solidification are on the order of only a few percent. Several previous solidification
models [14] confirm that the solidified metal part indeed undergoes only small
deformation during solidification. The displacement spatial gradient is small ru ¼
ou=ox so ru : ru 1 and the linearized strain tensor becomes [15]:
1h i
e¼ ru þ ðruÞT ð3:87Þ
2
The small strain formulation can be used, where Cauchy stress tensor is
identified with the nominal stress tensor r, and b is the body force density with
respect to initial configuration:
r: rðxÞ þ b ¼ 0 ð3:88Þ
The rate representation of total strain in this elastic-viscoplastic model is given
by [9]:
e_ ¼ e_ el þ e_ ie þ e_ th ð3:89Þ
where e_ el ; e_ ie ; e_ th are the elastic, inelastic (plastic ? creep), and thermal strain rate
tensors respectively. Stress rate r_ depends on elastic strain rate, and in this case of
linear isotropic material and negligible large rotations, it is given by Eq. 3.90 in
which ‘‘:’’ represents inner tensor product:
r_ ¼ D : ðe_ e_ ie e_ th Þ ð3:90Þ
D is the fourth order isotropic elasticity tensor given by Eq. 3.91.

2
D ¼ 2lI þ KB l I I ð3:91Þ
3
Here l; KB are the shear modulus and bulk modulus respectively and are in
general functions of temperature, while I; and I are fourth and second order
identity tensors and ‘‘ ’’ is the notation for outer tensor product.
Inelastic strain includes both strain-rate independent plasticity and time
dependent creep. Creep is significant at the high temperatures of the solidification
processes and is indistinguishable from plastic strain [16]. The inelastic strain-rate
is defined here with a unified formulation using a single internal variable [17, 18],
equivalent inelastic strain e_ el characterize the microstructure. The equivalent
inelastic strain-rate e_ el is a function of equivalent stress r
, temperature T, equiv-
alent inelastic strain eel .
e_ el ¼ f ðr
; T; eel Þ ð3:92Þ
rffiffiffiffiffiffiffiffiffiffiffiffiffi
2 0 0
¼
r r r ð3:93Þ
3 ij ij
0
r is a deviatoric stress tensor defined in Eq. 3.94.
0 1
rij ¼ rij rkk dij ð3:94Þ
3
The workpiece is assumed to harden isotropically, so the von Mises loading

surface, associated plasticity and normality hypothesis in the Prandtl-Reuss flow
law is applied [19]:
0
2 rij
ðe_ ie Þij ¼ e_ el ð3:95Þ
3 r

Thermal strains arise due to volume changes caused by both temperature dif-
ferences and phase transformations, including solidification and solid-state phase
changes, i.e.:
ZT
ðeth Þij ¼ aðT Þ dT dij ð3:96Þ
To
where a is temperature dependent coefficient of thermal expansion, and To is the

reference temperature and dij is Kronecker’s delta. It is normally accounted for in
stress analyses through a temperature-dependent differential thermal expansion
coefficient, aðT Þ. ABAQUS/Standard analysis a spatially varying thermal expan-
sion can be defined for homogeneous solid continuum elements by using a dis-
tribution, which includes the tabulated values for the thermal expansion [9].
ABAQUS uses an implicit backward-difference scheme for time integration of
both temperature and displacements at every material integration point [9].
3.2.4.3 Numerical Solution
Finite element discretization was carried out using the ABAQUS software [9].
The simulation is performed in ABAQUS/Standard and consists of sequential
thermal-stress analysis. In the sequential thermal-stress analysis, 142468 elements
are incorporated while 132140 hexahedral elements are used for the thermal-stress
analysis. In addition, for the heat transfer analysis, mesh used elements of type
DC3D8 (8-node linear heat transfer brick) and stress analysis used C3D8 (8-node
linear stress brick). The temperature data are transferred to the elements used for
the stress analysis through the connectivity matrix. This provided less computa-
tional time for the converged results. The fixed boundary conditions are applied
on the both ends of the workpiece resembling the experimental laser welding
situation. In the stress analysis, displacements are stored by ABAQUS at the
nodal positions as a solution variable, and loads are defined as prescribed dis-
placements and forces. Employing the interpolation functions, it is possible to
calculate the strain and stress increments at any point within the element using the
compatibility and constitutive equations. ABAQUS transforms the mechanical
equilibrium equations into a set of simultaneous equations, such that the nodal
displacements and forces are related to each other through the elemental stiffness
matrix.
Laser heat flux with Gauss distribution and prescribed velocity of 10 cm/s along
the x-axis through user subroutine DFLUX is applied to the thermal model. The
Gauss parameter ‘‘a’’ is a = 0.000333 m, in accordance with the experimental
power intensity distribution. The thermal model consisted of two steps. The first
step, which lasts 0.05 s, simulates the response of plate under moving laser heat
flux. The second step, which lasts for 1000 s, simulated the continued cooling in
the model. Cooling was allowed to continue until all of the plate reaches initial
temperature (room temperature). The temperature–time history resulted from the
thermal analysis is used as input to the thermal stress analysis. The workpiece is
considered as an elastic body, which is modeled as von Mises elastic–plastic
material with isotropic hardening and with a yield stress that changes with tem-
perature. Moreover, temperature dependent properties of steel are used in the
simulations rather than the mechanical properties of each constituting elements.
This is necessary because of the difficulties associated with incorporating the field
of species transport in stress equations, which is complicated to implement. In
addition, the data for the coupling field of species transport and the momentum
equations are not readily available in the open literature. Nevertheless, the use of
temperature dependent bulk properties of the alloy is the common practice in the
thermal stress analysis.
3.3 Efficiency Analysis of Welding
The welding efficiency is presented incorporating to the first and second laws of
thermodynamics. The mass and energy parameters are introduced to quantify the
first and second law efficiencies. The mathematical arrangement of the efficiency
analysis is given in line with the previous study [13].
The conservation of energy in a laser welding process can be written as
dE X _ X
¼ Ein E_ out ð3:97Þ
dt
Considering the infinitesimal control volume, as shown in Fig. 3.8, for the
steady state laser welding operation:
X X
E_ in ¼ E_ out ð3:98Þ
or in an explicit form
Pin þ Rm_ in hin þ m_ gas; in hgas; in ¼ Prefl þ Q_ cond þ Q_ conv þ Q_ rad

ð3:99Þ
þ Rm_ out hout þ m_ gas; out hgas; out
Let
Q_ loss ¼ Prefl þ Q_ cond þ Q_ conv þ Q_ rad þ m_ gas; out hgas; out m_ gas; in hgas; in ð3:100Þ
3.3 Efficiency Analysis of Welding 65
Fig. 3.8 A schematic view LASER

of Laser welding process Nozzle
Pin
(Dia. 1.5 mm)
mgas,out
mgas,in Qrad
Qconv
V
Prefl
x
min
Qcond D
Workpiece
mout
then
Pin þ Rm_ in hin ¼ Q_ loss þ Rm_ out hout ð3:101Þ

or
Pin ¼ Q_ loss þ Rm_ out hout Rm_ in hin ð3:102Þ

For steady welding process the conservation of mass is
Rm_ in ¼ Rm_ out ¼ m_ bond þ m_ loss ð3:103Þ
where the rate of the bonding material m_ bond is
m_ bond ¼ m_ flatbond þ Dm_ b ð3:104Þ
in which Dm_ b is the difference between the actual bond material rate and the
required flat bond material rate. Thus, the conservation of energy can be written as
Pin ¼ Q_ loss þ m_ flatbond ðhb hin Þ þ Dm_ b ðhb hin Þ þ m_ loss ðhloss hin Þ ð3:105Þ
The bonding material enters the control volume in a solid form at room tem-
perature. It is melted to for the bond and is partially lost as a result of excess
heating and evaporation.
Equation (3.105) can be integrated with time for the duration of the welding
operation
ZDt ZDt ZDt ZDt

Pin ¼ Q_ loss þ m_ flatbond ðhb hin Þ þ Dm_ b ðhb hin Þ
0 0 0 0
ZDt
þ m_ loss ðhloss hin Þ ð3:106Þ
0
where Dt is the time taken for the welding process. Assuming no variation in the
properties during the welding process
Qin ¼ Qloss þ mflatbond ðhb hin Þ þ Dmb ðhb hin Þ þ mloss ðhloss hin Þ ð3:107Þ
RDt RDt
where Qin ¼ Pin dt and Qloss ¼ Q_ loss dt.
0 0
3.3.1 First Law (Energy) Efficiency
Based on Eq. (3.107), the energy efficiency can be defined as:

Energy used in producing the bont mbond ðhb hin Þ
gI ¼ ¼ ð3:108Þ
Total Energy Input Qin
or
mflatbond ðhb hin Þ þ Dmb ðhb hin Þ
gI ¼ ð3:109Þ
Qloss þ mflatbond ðhb hin Þ þ Dmb ðhb hin Þ þ mloss ðhloss hin Þ
where
hb hin ¼ cps ðTm To Þ þ Lm ð3:110Þ
and
hloss hin ¼ cps ðTm To Þ þ Lm þ cpl ðTevap Tm Þ þ Levap ð3:111Þ
Defining a weld bond excess mass ratio a
Dmbond
a¼ ð3:112Þ
mflatbond
weld mass loss ratio b
Dmloss
b¼ ð3:113Þ
mflatbond
and heat loss ratio c
Qloss Qloss
c¼ ¼ ð3:114Þ
Qideal mflatbond ðhb hin Þ
the energy efficiency can be re-written as
1þa
gI ¼
ð3:115Þ
hin
1 þ a þ b hhloss
b h in
þc
3.3 Efficiency Analysis of Welding 67
3.3.2 Second Law (Exergy) Efficiency
Similarly, the exergy efficiency can be defined (based on the exergy balance) as:
Exergy used in producing the bond mbond ðxb xin Þ
gII ¼ ¼ ð3:116Þ
Total Exergy Input Xin
RDt
where Xin ¼ Pin dt.
0
The specific exergy change of bonding material can be written as:
xb xin ¼ ðhb hin Þ T0 ðsb sin Þ ð3:117Þ
or

Tm Lm
xb xin ¼ cps ðTm To Þ þ Lm T0 cps ln þ ð3:118Þ
T0 Tm
Thus the exergy efficiency in laser welding process can be written as
n h
io
mbond cps ðTm To Þ þ Lm T0 cps ln TTm0 þ LTmm
gII ¼ ð3:119Þ
Xin
Using Eqs. (3.101) and (3.113), the exergy efficiency becomes
8 h
i9
< Tm Lm =
1þa T 0 cps ln T0 þ Tm
gII ¼
1 ð3:120Þ
1þaþb hloss hin
þc : c ps ðT m T o Þ þ L m ;
hb hin
or
8 h
i9
< T0 cps ln TTm0 þ LTmm =
gII ¼ gI 1 ð3:121Þ
: cps ðTm To Þ þ Lm ;
Equations 3.115 and 3.121 are used to compute the first and second law effi-
ciencies of the welding process.
Numerical treatment of the laser forming and welding processes are presented
under different sub-headings in line with the previous studies [1–4]. Since the
process is involved with the phase change, the findings of the phase change are
presented initially and later thermal stress analysis is presented in the following
manner.
3.4.1 Phase Change Process Including Surface Evaporation
High power laser heating of surfaces results in melting and evaporation at the
surface. This becomes necessary to achieve deep penetration process such as
welding and forming. Since the phase change process involves with the mushy
zone generation at the interfaces, the findings of the model study becomes
important for correct predictions of the stress filed. The findings of the phase
change process are given below in line with the previous finding [4].
Figure 3.9 shows temperature contours in the region of irradiated spot for the
heating period of 12 ns. The cavity wall lies on the contour of evaporation tem-
perature. Evaporation of the surface initiates after 1.35 ns of the initiation of the
laser pulse. Once the evaporation starts, the depth of evaporated region extends
towards the axial direction and radial extension is mainly associated with the laser
intensity distribution in the radial direction, which is Gaussian. Extension of
contours of vapor temperature along the symmetry axis is more than its counterpart
along the axial direction at any other radial locations. This is because of the
absorption of laser beam. In this case, at the symmetry axis, laser beam intensity is
high and energy distribution in the absorption depth along the symmetry axis
enables more energy being absorbed by the substrate material than any other radial
location. This in turn, enhances the evaporation rate and increases the depth of
evaporated region along the symmetry axis. This situation is more pronounced
with increasing heating period.
Figure 3.10 shows the cavity cross-section. It is observed that near conical
cavity shape is resulted after the initiation of evaporation. In addition, the mushy
zone formation at interfaces of solid–liquid and liquid–vapor is evident. The size
of mushy zone is small particularly at interface of solid–liquid phases. This is
because of the latent heat of fusion, which is considerably smaller than that of
vapor. Moreover, the size of the mushy zone is small in the region close to the
cavity edges. This is because of the laser incident energy, which is less in this
region due to Gaussian distribution of laser beam intensity. In addition, temper-
ature gradient in the axial direction in this region is smaller than that of symmetry
Fig. 3.9 Temperature

contours at two different
heating periods
Fig. 3.10 Cavity cross

section after 6.5 ns heating
duration
axis, i.e. heat flow in the axial direction at the cavity centre is larger than that
corresponding to in the region of cavity edges.
Figure 3.11 shows velocity contours in and around the cavity for two vapor
densities. The size of the domain presented in Fig. (3.11) is larger than the cavity
size. This is due to that the vapor jet emanating from the cavity expands into cavity
as well as into the cavity surrounding, which is water ambient, during the time
domains considered. In this case, the size of the jet expansion is much larger than
the cavity size. In the case of vapor density 50 kg/m3, temporal extension of cavity
in the radial direction results in complex flow structure in the region of the cavity
edge. Moreover, jet expansion along the symmetry axis is high in the early heating
period. As the heating period increases, jet expansion in the radial direction also
increases. This is because of the relatively low density jet expansion into high
density stagnant water. In the early heating period, jet emanating from cavity
purges the stagnant water in the axial direction more than in the radial direction.
Due to the momentum change during this process, pressure builds up in the frontal
region of the jet. In this case, radial expansion of the jet becomes unavoidable due
to blockage affect of the pressure build up in the frontal area of the jet in the axial
direction. The radial expansion of the jet is evident in the early heating period. It is
observed that next to the jet boundary particularly in the region of cavity edge,
circulation cell is generated. Moreover, the detachment of the jet after emanating
from the cavity is also observed. This is more pronounced during the late heating
period ðt
7:74 nsÞ. This indicates that the initial acceleration of the jet is high and
once the pressure increases in the frontal area of the jet, radial expansion of the jet
becomes more than its axial component. However, continuous vapor mass addition
to the jet in the axial direction in the cavity, due to cavity recession, pushes the jet
front penetrating into the surrounding fluid. This forms like a jet neck immediately
above the cavity exit. Due to the rate of fluid strain, which is high in this region,
forms a circulation cell in the vicinity of the jet neck. This appears as a detachment
of jet front from the mainstream jet flow.
Fig. 3.11 Velocity contours for vapor front density of 50 kg/m3

3.4.2 Laser Multi Spot Heating of Thin Metal Sheet
Laser heating of a sheet metal and thermal stress predictions are presented in line
with the previous study [1]. Multi spot beam forming an arc at the workpiece
surface is considered.
Figure 3.12 shows temperature distribution along the arc for different number
of laser irradiated spots. The difference in the maximum and the minimum
temperature is considerably high for seven irradiated spots at the surface. Large
difference indicates that overlapping of heated regions between the irradiated spots
is small. Therefore, temperature uniformity along the arc, where the irradiated
spots are situated, becomes less likely. Consequently, the temperature gradients in
the edges of the irradiated spots become high. As the number of laser spots along
the arc increases, temperature difference between the maximum and the minimum
temperatures becomes small. This is more pronounced for 15 laser spots at the
surface. Although complete overlapping of temperature field is not evident, tem-
perature difference along the arc is small, which can be considered as homoge-
neous temperature distribution along the arc. Therefore, the temperature gradient is
expected to be rather smaller at the edges of the irradiated spot along the arc.
Figure 3.13 shows von Mises stress variation along the arc for various numbers
of irradiated spots. von Mises stress attains high values in the region of the large
temperature gradients. The difference between the maximum and the minimum
von Mises stress is large for seven irradiated spots at the surface. This difference is
on the order of 400 MPa which is considered to be high across the small distances
along the arc. The difference in von Mises stress increases slightly towards the end
of the arc. This is associated with high rate of heat conduction towards this region
while resulting in high temperature gradients in the arc end region. von Mises

distribution along the arc due
to various numbers of spots.
ns represents number of spots
Fig. 3.13 Von Mises

distribution along the arc due
to various number of spots. ns
represents number of spots
stress increases as the number of irradiated spots increases along the arc length.
The difference between the maximum and the minimum von Mises stress becomes
small. This, in turn, results in stress continuity along the arc length. Therefore,
increasing number of irradiated spots minimizes the local heating and thermal
stress field along the arc. When comparing difference in von Mises stress due to
laser spots 13 and 15, it is evident that both number of spots result in almost the
same stress field continuity along the arc, provided that 13 irradiated spots gives
rise to less von Mises stress level than that of 15 irradiated spots. Although stress
level is high, it is less than the yielding limit of the substrate material.
3.4.3 Three-Dimensional Heating and Stress Fields
Laser multi-beam heating of moving steel sheet at constant velocity is considered.

Influence of location of laser spots and beam intensity on temperature and stress
fields are examined. Since the heating and thermal stress generation are symmetric
along the symmetry plane, the results obtained for temperature and stress varia-
tions are presented for the half domain of the irradiated workpiece in line with the
previous study [2].
Figure 3.14 shows temperature and von Mises stress variation along the x-axis
for different laser intensities at the irradiated spots and y-axis location is y = 0.
It should be noted that the workpiece material moves at a constant speed along the
x-axis, z-axis is the depth below the surface y-axis is normal to the laser scanning
direction and Io is the peak intensity at the front and back spots, respectively. Two
temperature peaks are observed at y = 0 location, which corresponds to the center
of the first laser spot. The presence of the first peak is due to the first spot and the
Fig. 3.14 Temperature and

von Mises stress variation
along the x-axis for different
laser intensities at the
irradiated spots
second peak corresponds to other two spots, which are situated at y = 0.25 mm
behind the first spot. Temperature increases sharply to reach its peak value in the
region of the first peak and temperature decay is also sharp towards the second
temperature peak provided that temperature between the peaks is high along the
x-axis. Temperature decay is gradual behind the second spot. This is attributed to
laser scanning at a constant speed along the x-axis; in which case, region behind
the second spot is initially heated by the scanning laser beam. Convection, con-
duction, and radiation losses do not reduce temperature substantially in this region.
As the power intensity changes at the irradiated spots, the location of the peak
temperature charges along the x-axis, provided that behavior of temperature rise in
the vicinity of the first spot and temperature decay after the second spot do not
alter notably. In the case of Fig. (3.14)b, the peak power intensities at the irra-
diated spots and y = 0 and z = 0.25 mm below the surface. von Mises stress
attains high values in the region where temperature gradient is high; however, von
Mises stress becomes low in the region where temperature is high. This is asso-
ciated with temperature dependent elastic modules, which reduces with increasing
temperature. Consequently, the location of the maximum temperature does not
coincide with the location of the maximum stress. von Mises behavior is different
than that of temperature behavior. In this case, a secondary stress peak is formed
further away from the second spot along the x-axis. The occurrence of the second
stress peak is due to temperature gradient in this region. In addition, the maximum
stress at the second peak is almost half of that of the first peak. Therefore, low
value of von Mises stress is resulted due to relatively lower temperature gradient at
the second peak location. The influence of laser intensities at two irradiated spots
on von Mises stress is significant. In this case, reducing power intensity at the first
spot while increasing at the second spot moves the location of von Mises stress
peaks further away from the irradiated spots center. This is associated with the
location of the maximum temperature gradients along the x-axis, which changes
with varying laser intensities at two spots. In addition, the peak value of von Mises
stress reduces for low beam intensity at the first spot. Consequently, beam intensity
at the first spot has significant effect on the peak value of von Mises stress as
compared to that at the second spot.
Figure 3.15a shows temperature distribution along the z-axis at location x = 0
and y = 0 for different power intensities at the first and second spots. Temperature
decay is sharp in the surface region and temperature decay becomes gradual as the
depth below the surface increases towards the solid bulk. The sharp decay of
Fig. 3.15 Temperature and

stress distributions along the
z-axis at location x = 0 and
y = 0 for different power
intensities at the first and
second spots
temperature in the surface region is attributed to convection cooling at the surface.

Moreover, high temperature gradient at the surface enhances the conduction heat
transfer from the surface region towards the solid bulk, which contributes to sharp
decay of temperature in the surface region. As the laser intensity at the first spot
reduces, value of the maximum temperature reduces giving rise to relatively small
temperature gradients than that of at high intensities. In addition, low temperature
gradient suppresses conduction heat transfer from surface region towards the solid
bulk. Consequently, temperature decay becomes more gradual along the x-axis
than that of high intensities. Figure 3.15b shows von Mises stress distribution
inside the substrate along the z-axis at x and y-axes locations at x = 0 and y = 0
for different laser intensities. von Mises stress reduces sharply from its maximum
value as the depth below the surface increases. In this case, von Mises stress
distribution follows almost the same as temperature distribution in the surface
region. However, as the depth below the surface increases, decay of von Mises
stress becomes gradual. This is attributed to the thermal strain developed in this
region. Consequently, in this region, von Mises stress does not follow temperature
distribution. The influence of laser intensity on von Mises stress is significant; in
which case, increasing intensity enhances the temperature gradient while
increasing von Mises stress levels, particularly, in the surface region.
3.4.4 Temperature and Stress Fields for Welding

and Forming Process
The numerical treatment of laser welding process is presented for Haynes 188
sheet metal in line with the previous study [3].
Figure 3.16 shows temperature distribution along the x-axis for different
cooling times. The cooling period initiates at t = 0.05 s from the welding starts.

variation along the x-axis for
different cooling periods. The
cooling cycle is initiated at
t = 0.05 s
The laser transverses the workpiece along the x-axis at a constant scanning speed of
10 cm/s during welding. Consequently, the laser power ceases at x-axis location
x = 0.005 m. At the point of cooling initiation, temperature in the in the irradiated
surface remains high, which is well above the melting temperature of the substrate
material. In this case, superheating of the liquid phase takes place at the irradiated
spot center. The decay of temperature is sharp in the region x [ 0.005 m i.e. in front
of the irradiated spot. The sharp decay of temperature is associated with the tem-
perature difference between the irradiated spot edge, which is well above the melting
temperature, and the base material temperature, which is at 300 K. However, tem-
perature decays gradually in the region 0 B x B 0.005 m, which corresponds to the
initially welded region. This is attributed to the initial heating of this region during
the welding process. Consequently, conduction, convection, and radiation heat
transfer reduces temperature gradually in this region. The presence of almost flat
temperature region behind the irradiated spot (0.0038 m B x B 0.0045 m) attri-
butes to the phase change in this region. Since the solidus and liquidus temperatures
difference is small, the phase change takes place not at constant temperature, but
across the solidus and liquidus temperature. As the cooling period progresses,
temperature decays rapidly along the x-axis; in which case, the maximum temper-
ature reduces to 850 K at 0.07 s in the cooling period.
Figure 3.17 shows von Mises stress distribution along the x-axis for different
cooling periods as similar to those shown in Fig. 3.16. von Mises stress reduces in
the region of the laser irradiated spot at the initiation of the cooling period. This
occurs because of the attainment of the low elastic modulus at high temperatures.
However, in the region next to the laser irradiated spot edge (x [ 0.005 m), von
Mises stress attains high values in the early cooling period (t = 0.05 s). This is
attributed to the attainment of the high temperature gradients in this region.
Consequently, high temperature gradient results in high thermal strain in the
welding region. Moreover, in the region where temperature decay is gradual
(0 B x B 0.05 m), von Mises stress attains low values for the cooling periods
Fig. 3.17 Von Mises stress

variation along the x-axis for
t = 0.05 s
t B 0.07 s. This occurs because of: (i) temperature is reasonably high lowering the
elastic modulus in this region, and (ii) thermal strain developed along the x-axis
due to the temperature gradient is not significantly high. However, as the cooling
period progresses further, temperature reduces to almost initial temperature of the
workpiece. von Mises stress becomes the residual stress in the welding section. It
is evident that the residual stress attains high values and remains almost the same
along the x-axis, except towards x B 0.005 m where the residual stress increases
slightly.
Figure 3.18 shows temperature distribution along the z-axis for different
cooling periods. It should be noted that the z-axis is the laser beam axis. Since the
workpiece thickness is 1 mm, temperature profiles terminate at z = 0.001 m
below the surface and it reduces as the depth below the surface increases towards
the bottom surface. Temperature remains high along the thickness of the work-
piece; in which case superheating of liquid phase takes place. This is more pro-
nounced in the region of the top surface of the workpiece. Moreover, temperature
remains above the liquidus temperature of the substrate material at the bottom
surface of the workpiece. Temperature decay in the mid-section of the workpiece
(0.00025 m B x B 0.00075 m) is slightly sharper than that corresponding to top
surface of the workpiece. This is attributed to the internal energy gain from the
irradiated field, which is high in the surface region because of the exponential
decay of the absorbed laser energy with increasing depth below the surface.
Consequently, temperature difference between the mid-section and the bottom
surface of the workpiece results in sharp decay of temperature in the mid-section
of the workpiece.
Figure 3.19 shows von Mises stress distribution along the z-axis for different
cooling periods as similar to those shown in Fig. 3.18. von Mises stress attains
high values in the top and bottom surfaces of the melted section during the long
cooling periods (t B 0.17 s); however, it remains low during the early cooling
period because of the attainment of low elastic modulus at high temperatures. The
residual stress reduces in the mid-section of the workpiece, which is associated
with the low cooling rate. The heat conduction along the z-axis is less resulting in

distribution along the z-axis
for different cooling periods.
The cooling cycle is initiated
at t = 0.05 s
Fig. 3.19 Von Mises stress

variation along the z-axis for
t = 0.05 s
slow cooling rate in the mid-section of the workpiece. The low heat conduction is
because of the small temperature difference between the laser beam axis and its
neighborhood inside the workpiece. The slow cooling rate results in stress relax-
ation in the mid-section of the workpiece and the bottom surfaces of the
workpiece.
3.4.5 Welding Efficiency
The findings for the welding efficiency are presented in line with the previous
study [20].
Figure 3.20 shows the tensile test results for welds obtained for aluminum,
nickel and steel workpieces at different thicknesses. The breaking load decreases at
the upper end of the thickness. This effect is significant in the case of nickel and
steel. It should be noted that nickel and steel have higher thermal diffusivity than
aluminum; consequently, small surface plasma may develop, which in turn,
improves the bonding in the weld zone. In this case, the surface plasma acts as a
heat source enhancing the mass removal rate and allowing more laser energy
penetrating into the material. On the other hand, when the size of the surface
plasma increases, it absorbs the incident laser beam and reduces the laser energy
reaching the workpiece. This may be the case occurring for thick samples, since it
was reported that the recoil pressure at the interaction zone increases for thicker
materials, which in turn increases the mass removal rate from the workpiece
material. When comparing the breaking load corresponding to the different size of
the spot diameters, large spot diameter (0.8 mm) results in low breaking load. This
may be due to that the energy intensity available at the workpiece surface becomes
less, therefore, the full penetration of the laser beam may not be achieved, i.e. the
sound weld may not be obtained.
Fig. 3.20 Breaking load for

three different materials and
two laser spot diameters. a
Aluminum workpieces. b
Nickel workpieces. c Steel
workpieces
First law efficiencies are shown in Fig. 3.21 for aluminum, nickel and steel
workpieces. In general, the first law efficiency increases as the diameter of the spot
weld increases. This is due to that the energy requirement for a large diameter
weld is high. This is also true for the thick samples, i.e. as the weld thickness
increases, the first law efficiency increases. On the other hand, as the thickness
increases, the weld strength in Fig. 3.20a reduces and after 0.8 mm thickness the
welding ceases. In this case, laser energy output, which is 20 J, is not sufficient to
melt the material for the completion of the welding process. It is also true that
when the weld diameter increases, the energy intensity available for full pene-
tration of both sheet metals become less, therefore, the weld strength drops. It
should also be noted that use of oxygen in the welding region triggers the high
temperature exothermic reaction, which in turn provide excess energy for
improved penetration. In this case, the material removed from the top and bottom
Fig. 3.21 First law

efficiency of laser welding. a
Aluminum workpieces. b
Nickel workpieces. c Steel
workpieces
weld pools increases. When considering the material properties, it is evident that
materials having high specific heat and latent heat of melting results in high first
law efficiency. Consequently, the first law efficiency is higher in the case of steel,
and then follows nickel and aluminum.
Figure 3.22 shows the second law efficiencies corresponding to aluminum,
nickel and steel. In general, second law efficiency increases as the thickness of
the welded parts and the diameter of the spot weld increase. The behavior of the
second law efficiency with workpiece thickness and the weld spot diameter is
similar to that obtained for the first law efficiency. It is evident that second law
efficiency corresponding to steel is higher, then follows nickel and aluminum. It
should be noted that the laser output energy is set to increase with increasing
Fig. 3.22 Second law

efficiencies and laser output
energy. a Aluminum
workpieces. b Nickel
workpieces. c Steel
workpieces
thickness of the welded parts. This is due to the requirements of the energy for
achieving a sound weld. When comparing the first and second law efficiencies, it is
evident that the second law efficiencies are lower than the first law efficiencies.
This may be due to that the laser beam source is considered as a clean energy
source.
The findings from the phase change process and from the numerical treatment of
the heating and thermal stress formation are presented in relation to laser forming
and welding processes.
The numerical treatment of the phase change due to high intensity laser pulse
indicates that the cavity extension along the symmetry axis is larger than that
corresponding to the cavity edge. This is due to absorption of laser power intensity,
which is high along the symmetry axis. The size of the mushy zones in the region
close to the cavity edge is smaller than that of symmetry axis. The size of solid–
liquid mushy zone is smaller than the size of liquid–vapor mushy zone. The jet
density has significant influence on the flow structure within and around the
expanding jet. Since the mass flow rate emanating from the cavity is kept the same
for both densities, jet with low density secures high velocity at cavity exit. This, in
turn, results in formation of circulation cell next to the jet boundary immediately
after the cavity exit. The recoil pressure developed in the cavity is considerably
high. As the time progresses, the recoil pressure reduces in the vicinity of the
cavity due to cavity recession.
The findings of the multi-spot heating in the form of arc at the surface reveal that
the number of irradiated spots has significant effect on temperature and stress fields.
In this case, local heating is resulted along the arc for less number of irradiated
spots. This, in turn, causes high temperature difference between the maximum and
the minimum temperatures along the arc length while resulting in local high stress
regions along the arc, i.e. the difference between the maximum and the minimum
von Mises stress is on the order of 400 MPa. As the number of irradiated spots
along the arc length increases, difference between the maximum and the minimum
temperatures becomes small and high temperature region along the arc length
extends almost uniformly. This gives rises to almost uniform high stress region
along the arc, i.e., the difference between the maximum and the minimum von
Mises stress becomes considerably small. Consequently, introducing multi-spots
along the arc length increases heated region at the workpiece surface. The conti-
nuity in high temperature and high stress fields along the arc is possible through
proper selection of the number of irradiated spots along the arc length.
The numerical findings of three-dimensional multi beam heating and stress for-
mulation for the thick substrates indicate that two temperature peaks are formed
along the x-axis. The first temperature peak has higher value than the second peak.
von Mises stress attains high values along the x-axis at y = 0 and z = 0. The
location of von Mises stress peak differs from the location of temperature peak; in
which case, von Mises stress attains high values where the temperature gradient is
high. The number of temperature peaks reduces to 1 at y-axis location is
y = 0.25 mm. In this case, the maximum value of temperature increases with laser
intensity at the second spot. Temperature decay is sharp in the surface region along
the z-axis and it becomes gradual as the depth below the surface increases. This
causes similar behavior of von Mises stress below the surface. von Mises stress
increases notably as the intensity at both spots increases. The presence of second and
third spot modifies significantly temperature and stress fields in the heated region.
The findings of finite element analysis for thermal and stress fields due to
welding application show that temperature in the welding section, particularly in
the irradiated spot, temperature well exceeds the melting temperature of the
substrate material and superheating in the liquid phase takes place in the melt pool.
3.5 Concluding Remarks 83
The high temperature gradients are formed in the vicinity of melted zone. Tem-
perature dependent elastic modulus reduces von Mises stress in the welding region
during the heating cycle. However, once the cooling cycle is initiated, von Mises
stress attains high values, particularly in the region of the high temperature gra-
dients. This results in high residual stress levels in the vicinity of the irradiated
spot edge. In addition, slow cooling rate in the welding core causes stress relax-
ation and the residual stress reduces in this region. Due to high temperature heating
in the melt pool, the surface tension gradient causes curvature of the top surface of
the welding section. This appears as a cavity upon the solidification.
The thermal efficiency analysis of the spot welding process indicates that the
first law analysis reveals that the first law efficiency increases as the workpiece
thickness increases. In addition, an increase in the diameter of the laser spot weld
increases the first law efficiency. These arguments are also true for the second law
efficiencies. However, it is evident from the tensile tests that the weld strength
drops at certain workpiece thickness and above, even though the first and second
law efficiencies improve. Therefore, a relation may exist between the weld strength
and thermal efficiencies of the welding process, in this case, for a sound weld, the
thermal efficiency may not be the maximum.
References
1. Shuja SZ, Yilbas BS (2012) Multi-beam laser heating of steel: temperature and thermal stress
analysis. Trans Can Soc Mech Eng 36(3):373–382
2. Shuja SZ, Yilbas BS (2013) Laser multi-beam heating of moving steel sheet: thermal stress
analysis. Opt Lasers Eng 51(4):446–452
3. Yilbas BS, Akhtar SS, Karatas C (2011) Laser welding of Hayness 188 alloy sheet: thermal
stress analysis. J Adv Manuf Technol 56(1–4):115–124
4. Yilbas BS, Mansour SB (2007) Laser heating: jet emanating from laser induced cavity. Int J
Therm Sci 46(4):385–398
5. Yilbas BS, Akhtar SS (2013) Laser welding of AISI 316 steel: microstructural and stress
analysis. ASME J Manuf Sci Eng 135(3), 034502 (May 27, 2013) (4 pages). doi:10.1115/
1.4024289
6. Shuja SZ, Yilbas BS (2000) The influence of gas jet velocity in laser heating—a moving
workpiece case. In: Proceedings of the institution mechanical engineers, part C, J Mech Eng
Sci 214:1059–1078
7. Ootao Y, Tanigawa Y (2005) Three-dimensional solution for transient thermal stresses of
functionally graded rectangular plate due to non-uniform heat supply. Int J Mech Sci
47:1769–1788
8. COMSOL multiphysics finite element analysis software (2012). http://www.comsol.com
9. ABAQUS Theory Manual (2009) version 6.2. ABAQUS Inc., Pawtucket
10. Naqavi IZ (2001) Conduction and non-conduction limited laser heating process—
mathematical simulation. M.Sc. thesis, Mechanical Engineering Department. KFUPM,
Dhahran, Saudi Arabia
11. ANSYS fluent (2012). http://www.ansys.com/Products/Simulation+Technology/
Fluid+Dynamics/Fluid+Dynamics+Products/ANSYS+Fluent
12. Yilbas BS, Arif AFM, Karatas C, AbdulAleem BJ, Tabet N (2010) Laser gas-assisted
nitriding of steel: residual stress. Ind Lubr Tribol 62(4):214–223
13. Sahin AZ, Ayar T, Yilbas BS (2010) Laser welding: the first and second law analysis. Int J
Exergy 7(5):535–546
14. Risso JM, Huespe AE, Cardona A (2006) Thermal stress evaluation in the steel continuous
casting process. Int J Numer Methods Eng 65(9):1355–1377
15. Mase GE, Mase GT (1999) Continuum mechanics for engineers, 2nd edn. CRC Press, Boca
Raton
16. Li C, Thomas BG (2004) Thermo-mechanical finite-element model of shell behavior in
continuous casting of steel. Metall Mater Trans B 35B(6):1151–1172
17. Anand L (1982) Constitutive equations for the rate dependant deformation of metals at
elevated temperatures. ASME J Eng Mater Technol 104:12–17
18. Lush AM, Weber G, Anand L (1989) An implicit time–integration procedure for a set of
integral variable constitutive equations for isotropic elasto viscoplasticity. Int J Plast
5:521–549
19. Mendelson A (1983) Plasticity: theory and applications. Krieger Pub Co, Melbourne
20. Yilbas BS, Kar A (1997) Laser spot welding and efficiency consideration. J Mater Eng
Perform 6:766–770
Chapter 4
Experimental Analysis for Laser Forming
and Welding
4.1 Introduction
Laser forming and welding are important areas of laser processing of metallic
substrates. Although analytical and numerical analyses provide physical insight
into the processes, experimental study of these processes is necessary to validate the
predictions and to achieve the required quality of the end product. In addition, the
theoretical predictions lower the experimental cost, since they provide information
between the affecting parameters and the dependent process characteristics.
Therefore, the selection of the experimental parameters and their levels becomes
easier for the cost effective experimental investigations. Nevertheless, the experi-
mental study compliments the theoretical predictions. The main focuses for the
experimental study are to examine: the microstructure of the laser treated sections,
mechanical properties of the laser welded sections, such as elastic modulus,
residual stress development, and microhardness, and the quality assessment of the
end product. In the this chapter, the analytical tools used for the microstructural
analysis, assessment of the mechanical properties of the resulting welds, and quality
analysis for the end product are given in line with the previous studies [1–10].
4.2 Lasers Used for Welding and Forming
High power lasers are used for laser forming and welding processes. However, in
sheet metal forming laser output power is lower than that of used for the welding
processes. Although many kinds of high power lasers are available for forming and
welding applications, CO2 and Nd:YAG lasers are widely used in industry because
of their low process cost and ready availability. CO2 lasers are usually in CW
mode or delivering high frequency pulses with adjustable duty cycle. Nd:YAG
laser operates in a pulse mode and it has many advantages for the application of
laser spot welding of small size parts including electrical wires. In this section, the
brief description of CO2 and Nd:YAG lasers are given.

86 4 Experimental Analysis
Table 4.1 Typical laser welding conditions used in the experiment

Welding Power Frequency Nozzle gap Nozzle Focus N2 Gas
speed (cm/s) (W) (Hz) (mm) diameter setting pressure (kPa)
(mm) (mm)
10 2000 1500 1.5 1.5 127 600
The data presented in this book are based on the previous studies [1–10], in
which CO2 laser was used. In this case, CO2 laser (LC-ALPHAIII) delivering
nominal output power of 2 kW was used to irradiate the workpiece surface. The
nominal focal length of the focusing lens was 127 mm. The laser output power is
in Gaussian mode (TEM00 mode) and 1/e points are located at ±0.2 mm.
Nitrogen assisting gas emerging from the conical nozzle and co-axially with the
laser beam was used. The use of nitrogen as an assisting gas is due to its cost and
higher specific heat capacity as compared to that of argon. The mass flow rate of
assisting gas was on the order of 1.9 9 10-4 kg/s. Typical laser welding condi-
tions are given in Table 4.1.
Nd: YAG laser used in the previous studies are presented in this section. In this
case, a Nd YAG laser of output energy in the range of 10–21 J with a nominal pulse
length of 1.48 ms giving output power intensity of the order of 100 GW/m2 was
used to irradiate the workpieces. An energy power meter was used to measure the
instantaneous laser output power and energy. A nominal focal length of 51 mm
focusing lens was used to focus the laser beam. The laser output power at the
workpiece surface is Gaussian (TEM00 mode) and 1/e points are located
at ±0.75 mm. A nozzle of 0.6 mm inner diameter was placed 2 mm away and
1.5 mm above the center of the irradiated spot. The nozzle axis was set at an angle
of 45o to the axis of the laser beam. Argon and oxygen were used as assisting gas.
Steel (EN58), titanium, and nickel were used as workpieces. In the experiments, the
focal position and laser output power were set to give a spot of 1 mm diameter on
the workpiece surface and a laser peak-power intensity of about 1010 W/m2.
Welding process was carried out using different assisting gas pressures. However,
after the welding process, tensile testing of resulting welds was carried out to assess
the weld strength. A schematic view of laser spot welding is shown in Fig. 4.1.
The validation of the simulation results is important for the assessment of the
accuracy of the numerical method. One of the methods is to validate temperature
predictions. In order to validate temperature predictions, a thermocouple can be
used to monitor the temporal variation of surface temperature. In this case, the
location of thermocouple is important. In line with the previous studies [1–10],
thermocouple was located at 0.5 mm away from the weld center line at the surface
(Fig. 4.2). This was necessary to avoid the melting of the surface of the thermo-
couple during the laser welding process. The thermocouple output was calibrated
according to the previous study [11]. The experimental error was determined using
the experimental repeatability; therefore, the experiments were repeated three
times and the error was estimated in the order of 5 %.
4.3 Analytical Tools for Microstructural Analysis 87
Fig. 4.1 A schematic view of laser system for spot welding
Fig. 4.2 A schematic view of laser welding situation and the thermocouple location
4.3 Analytical Tools for Microstructural Analysis
The microstructural and morphological analysis of the laser treated layer is very
important. To achieve this, scanning electron and optical microscopes needs to be
used. Since the previous studies [1–10] are presented in this section, JEOL JDX-
3530 scanning electron microscope (SEM) was used to obtain photomicrographs
of the cross-section and surface of the workpieces after the tests. Fully computer
integrated optical microscope was used to obtain optical micrographs of laser

welded sections.
The compounds formed in the bended and welded sections are important. One
of the methods to characterize the compounds is to use X-ray diffraction (XRD).
Since the data presented in the previous studies [1–10] were presented in this book,
XRD analysis used in the previous studies were presented herein. In this case,
XRD analysis was carried out incorporating Bruker D8 Advanced XRD unit with
having Cu-Ka radiation. A typical setting of XRD was 40 kV and 30 mA. The
parabolically-shaped Göbel Mirror was used in Bruker D8 Advanced, which
provided the highly-parallel X-ray beams. The parallel beams from it removed the
traditional limitations of the Bragg–Brentano geometry for powder diffraction and
thin film investigations.
The surface of the workpiece was ground to obtain the average surface
roughness of 1.5 lm prior to laser treatment process. This was due to that the
surface roughness of 1.5 lm did not result the considerable surface reflection
during the laser irradiation. However, lowering the surface roughness through fine
surface polishing resulted in substantial reflection of the laser beam while dam-
aging the focusing lens. Therefore, the surface roughness of the workpieces was
kept almost the same during experiments.
4.4 Mechanical Characterization of Laser Welded

and Bended Sections
Mechanical characterisations of laser bended and welded sections are important

for the end product quality point of view. The main mechanical characterizations
of the resulting welds and bends include microhardness, residual stress measure-
ments, and tensile properties of the weld sections.
The microhardness tests were presented in line with the previous studies [1–10].
In this case, microphotonics digital microhardness tester (MP-100TC) was used to
obtain microhardness at the surface of the nitride layer. The standard test method
for Vickers indentation hardness of advanced ceramics (ASTM C1327-99) was
adopted. Microhardness was measured at the workpiece surface after the laser
treatment process. The measurements were repeated three times at each location
for the consistency of the results.
The residual stress measurement relies on the stresses in fine grained poly-
crystalline structure. The position of the diffraction peak exhibits a shift as the
specimen is rotated by an angle W. The magnitude of the shift is related to
the magnitude of the residual stress. The relationship between the peak shift and
the residual stress (r) is given [12]:
E ðdn do Þ
r¼ ð4:1Þ
ð1 þ tÞSin2 w do
4.4 Mechanical Characterization of Laser Welded and Bended Sections 89
where E is Young’s modulus (220 GPa), m is Poisson’s ratio (0.28), W is the tilt
angle, dn are the d spacing measured at each tilt angle, and do is the stress-free
lattice spacing. If there are no shear strains present in the specimen, the d spacing
changes linearly with sin2W. For example, the bcc ferrite steel peak at [211] planes
takes place at 2h = 88.3o with the inter-planer spacing of 0.11709 nm. In addition,
knowing the elastic modulus of steel, one can determine the residual stress in the
weld section from the X-ray data. The XRD experiment for the residual stress
measurement should be repeated three times at different locations in the weld or
bend section to secure the repeatability of the results. It is expected that the error
related to the measurements is in the order of 3 %.
In general, laser welding resulted in slightly unsymmetrical weld geometry
along the central line at the cross-section. This might effect the formulation of the
elastic modulus through 3-point bending tests. However, the weld size variation at
the weld cross-section due to unsymmetrical geometry could be assumed to be
small. Therefore, 3-point bending tests can be carried out to measure the weld
elastic modulus. Figure 4.3 shows a schematic view 3-point bending tests. The 3-
point bending tests can be realized with a constant speed, such as of 1 mm/min,
and a span of 46.4 mm. Before the 3-point bending, the samples needs to be
machined to flatten the workpeices resulting in samples dimension of approxi-
mately 20 mm in width and 2 mm in thickness. The fluxural modulis of elasticity
can be measured using the equation:
mL3
E ¼ ð4:2Þ
4bd3

of three-point bending test Applied Load
unit
Indent
Workpiece
Weld Section
Support
L
where L is the support span (mm), m is the slope of the tangent to the initial
straight-line portion of the load deflection curve (N/mm), b is the width of the test
sample (mm), and d is the thickness of the sample (mm).
Experimental findings for forming and welding processes are classified under the
relevant sub-headings. In this case, bend angle measurement is presented first,
metallurgical and morphological changes, residual stress measurements, microh-
ardness, and elastic modules estimations are presented later in line with the pre-
vious findings [8–10, 13].
4.5.1 Bending Angle Measurements, Morphology

and Microstructure for Laser Forming
The discussions on the bend angle data is presented in line with the previous study [10].
Figure 4.4 shows bending angle with laser output power intensity. The laser
scanning speed is kept at 1.7 cm/s and the number of line scans is 20 during the
experiments. Increasing laser output power increases the bending angle and this
variation appears to be almost linear. The increase in the bending angle with
increasing laser power is attributed to depth of the melt and the heat affected zone;
in which case, depth of the melt zone enhances with increasing laser power
intensity. Although the depth of heat affected zone increases, the rate of this
increase is not greater than the rate of increase in the size of the melt zone. This is
because of the absorption of laser beam, which takes place at the surface and
energy absorbed is converted into internal energy grain of the substrate material in
Fig. 4.4 Bending angles

predicted and measured with
laser power intensity
the surface vicinity. This, in turn, enhances the melt depth with increasing laser
power intensity in the irradiated region. However, the conduction heat transfer
from the melt zone mainly determines the size of the heat affected zone. Since the
conduction heat transfer depends on the thermal conductivity and the temperature
gradient at the melt pool vicinity, the heat transfer rates from the melt vicinity to
the solid bulk remains almost the same for all laser power intensities. This is
because of the fact that temperature at the melt pool vicinity remains almost the
same since the melting temperature of the substrate material and the bulk tem-
perature does not change significantly. Therefore, temperature gradient becomes
almost constant in this region, which in turn results in almost constant heat transfer
rate from the melt pool vicinity to the solid bulk. In addition, thermal strain
developed in the heated region depends on the temperature gradient. Hence,
almost constant temperature gradient causes the attainment of the constant thermal
strain in this region. Consequently, bending angle, which is associated with the
thermal strain in the irradiated region changes almost linearly with increasing laser
output power intensity. The small difference is attributed to the experimental error,
which is estimated as of the order of 5 %, and the assumptions made in the
analytical formulation, which assumes the constant material properties. Never-
theless, the difference in both results is acceptably small.
Figure 4.5 shows the residual stress predicted along the y-axis for different x-
axis locations. The residual stress is compressive in the surface region, which is
particularly true at locations corresponding to x = 0.6 mm and x = 1.2 mm.
However, it becomes tensile towards the bottom surface of the workpiece at
location x = 0.6 mm. This is attributed to thermal strain developed in this region;
in which case, differences in the cooling rates during laser scanning alters the
thermal strain in the workpiece. The residual stress remains low in the surface
region at location corresponding to x = 0, y = 0 and z = 0. This occurs because
of the self-annealing effect of the laser heating process. In this case, when laser
moves away from this region, temperature remains high in this region due to
Fig. 4.5 Residual stress

predicted along the y-axis for
different x-axis locations and
z-axis location is z = 0 mm
conduction heat transfer from the laser heated spot. This, in turn, lowers the
cooling rate while reducing the residual stress levels in this region. The maximum
residual stress is on the order of -110 MPa below the surface (y = 0.18 mm).
However, the maximum value of the residual stress at the surface is -80 MPa at
locations corresponding to x = 0.6, y = 0, z = 0 and x = 1.2 mm, y = 0 mm,
z = 0.
Figure 4.6 shows optical photograph of laser bended workpieces while Fig. 4.7
shows SEM micrographs of laser scanned surface. It is evident that the overlapping
ratio of the laser irradiated spot at the surface is about 70 %. This, in turn, results
continuous melting at the surface along the laser scanning tracks. The melt flow
between the laser scanning track is not observed indicating the control melting at
the surface, i.e., excessive melting at the surface is avoided. In addition, cavity
formation at the surface is not observed which reveals that the evaporation at the
surface does no take place. It should be noted that increasing laser power intensity
at the surface initiates surface evaporation and a cavity formation. Consequently,
Fig. 4.6 Optical

photographs of laser bend Bended Workpiece
section
Workpiece Holder
20 mm
Laser Scan Tracks
5.5 mm
Fig. 4.7 SEM micrograph of

laser scanned surface,
overlapping ratio and Laser Scan Tracks
microcrack at the surface
Overlapping of Laser Spots
Microcrack
evaporation is avoided through control melting at the surface. Some microcracks

are observed at the surface due to high cooling rates; however, crack extension
forming a crack network is not visible at the surface. This indicates that once the
crack is formed, the strain energy releases locally suppressing the crack propa-
gation in the surface region. The roughness of the laser treated surface is on the
order of 10 lm. The attainment of high surface roughness is associated with deep
penetration melting and subsequent solidification; therefore, the surface tension of
the melted region forms textures with curvature appearance around the laser
scanned regions.
Figure 4.8 shows SEM micrographs of the cross-section of the laser treated
region. It is evident that the melt layer extends almost thickness of the work
piece, which is in good agreement with the predictions. The surface scanning
resulted in bending of the work piece. In addition, increase in surface roughness in
the laser treated region is evident because of the surface tension effect of the
melted material during the scanning. It should be noted that the bending angle at
the laser treated region contributes to the roughness at the surface. The closed
(a)
(b) (c)
Cellular Structure Fine Structure s
Dendiritic Structure
(d) (e)
Dendiritic Structure Demarcation Zone
Fig. 4.8 SEM cross-section of laser bended section: a general view of bend cross-section,
b dense layer at the surface vicinity, c fine structure next to the surface vicinity, d dentritic
structures in the melted region, e melted-heat affected region interface
examination of SEM micrograph reveals that a dense layer consisting of fine grains
is evident at surface vicinity. This is attributed to high cooling rates at surface due
to convection heat transfer by the presence of the assisting gas during the laser
treatment process. Since nitrogen assisting gas at high pressure is used, nitride
species, such as Fe4N is formed at the surface vicinity. This results in volume
shrinkage due to low density at the surface vicinity while contributing to the
formation of the dense layer in the surface region. The melted region close to the
Fig. 4.9 XRD diffractogram

of laser scanned region
surface contains dendrites and some cellular structures because of high cooling
rates. In addition, the ferritic-pearlitic microstructure is transformed to martensite
at the surface region of the laser treated layer due to the high cooling rates.
However, the solidification cracking in the surface vicinity does not take place
indicating the low solidification strains, which are below the ductility of the
solidifying molten metal. The demarcation zone between the laser treated region
and the base material is clearly observed.
Figure 4.9 shows XRD diffractograms of the laser treated surface. The peaks of
d-Fe, and c-Fe phases, and a-Fe phases and Fe4N compounds are visible from the
X-ray diffractograms. The incomplete transformation of d-c-phases take place at
the surface vicinity and metastable d-Fe also remains at the surface, which is
unavoidable due to convection cooling at the surface. The residual stress deter-
mined from XRD technique using equations is on the order of -90 MPa. The
formation of compressive stress at the surface vicinity is associated with defor-
mation during the solidification of the melted region under the high cooling rates.
The residual stress predicted (-80 MPa) agrees well with that obtained for the
XRD technique (-90 MPa).
4.5.2 Morphology, Microstructure and Mechanical

Properties for Laser Welding Process
Experimental results associated with the laser spot welding morphology and
mechanical properties are presented in line with the previous findings [13].
The experimental data obtained for the laser spot welding of copper wires with
different diameters on steel, nickel, and titanium is presented. The laser power
intensity and focusing conditions are varied in the welding process. The tensile
testing of resulting welds are carried out to determine the mechanical strength of
the resulting welds. Optical microscopy is conducted to examine the metallurgical
changes across the weld zone.
40
35
Breaking Load (N) 30
25
20
15
10 Nickel Base
Stainless Steel Base
5
Titanium Base
0
30 35 40 45 50 55 60
Laser Output Power Intensity (W/cm 2) x 104
Fig. 4.10 Tensile strength for multi core wire diameter 2.5 mm
60
50
Breaking Load (N)
40
30
20
Nickel Base
10
Titanium Base
0
30 35 40 45 50 55 60
Three workpiece materials (Nickel, Titanium and Stainless Steel) and four
different wire diameters are considered. In Fig. 4.10 the tensile test results are
plotted for laser welded multi core wire having 2.5 mm outer diameter. It is
evident that the tensile strength of the welds is not very high when they are
compared to those given in Fig. 4.11. This may be explained by either or both of
the following circumstances: (i) with multi core wire, there are always air gaps
between the wires. This increases convective heat losses during welding and (ii)
the focal position relative to the wire surface is not constant across the surface,
since the free surface of the wire is not flat and consequence of this is that the
Fig. 4.12 Multicore wire

welds after cutting by laser
beam
power intensity varies across the surface, resulting in non-uniform welding sec-
tions. In Fig. 4.11 the tensile test results are plotted for 1.55 mm thick laser welded
multi core wire. It can be seen that the tensile strength of the welds are relatively
higher than those obtained for 2.5 mm thick multi core wire. This may be due to
the effect of the thickness of the wire. In this case relatively less energy is required
for melting of the wire and this leaves more available energy to vaporise the base
material. Therefore, the vapour pressure of the base material increases and a
relatively larger crater diameter results. This produces a large area of base material
to interface with the core material, However, the effect of base material on tensile
strength can be seen in both Figs. 4.10 and 4.11. The tensile strength is relatively
higher for materials having relatively higher thermal diffusivities and lower
absorption depths (such as Ni). In addition, the effect of power intensity is obvious,
i.e. welding improves for any given material at particular power intensity. This
case is obtained at 0.443*106 W/cm2 for 1.5 mm thick multicore wire material.
Photographs of multicore wire welds are given in Fig. 4.12. It can be seen that
some wires in the surface region are cut by the laser beam. This is especially true
for 2.5 mm thick wires. However, this reduces the tensile strength of the welds.
The tensile test results are given in Figs. 4.13 and 4.14 for single core wires at
different outer diameters, noting that the wire material is copper. The tensile
strength is relatively higher in the case of 0.3 mm thick wire. This may be due to
large liquid interface between the wire and base material. The tensile strength of
the welds having stainless steel and nickel base materials are higher than those
having titanium base material. This is again because of the thermal properties of
titanium. The tensile strength of 0.2 mm thick wire welds are considerably low as
seen in Fig. 4.13). This may be due to following reasons: (i) the wire diameter is
small and during the welding some amount of wire material evaporates resulting in
less wire material. This reduces the amount of material in the weld pool, i.e. the
joints are shallow, and (ii) the wire diameter is small; as a consequence of this, the
joint area is expected to be narrow. In addition, the tensile strength of the weld is
again relatively lower in the case of titanium than nickel and stainless steel.
However, as the wire diameter increases above 0.3 mm, the tensile strengths of the
welds decrease. This may be well explained after considering the heat losses
during the welding. Since the wire material being copper (having high thermal
diffusivity), heat losses due to diffusion are high during welding. This results in
less energy to reach the base material. Consequently, only the surface of the base
material may melt and evaporate resulting in shallow joints.
7
Breaking Load (N) 6
3
Titanium Base
2 Stainless Steel Base
Nickel Base
1
0
30 35 40 45 50 55 60
30
25
Breaking Load (N)
20
Titanium Base
15
Nickel Base
10
0
30 35 40 45 50 55 60
4.5.3 Laser Welding of Hayness 188 Alloy
The findings are presented in line with the previous study [7].
Figure 4.15 shows optical micrographs of the top and the bottom surfaces of the
weld section and Fig. 4.16 shows SEM micrographs of the weld cross-section. The
presence of small cavity at the top surface of the melt section is evident. In
addition, some small molten material accumulates at the weld bottom surfaces.
The small cavity formation at the top surface and the presence of some out flow of
molten metal at the bottom surface of the weld section are because of the
attainment of high temperature in the welding section. In this case, superheating in
Top surface Bottom Surface
Top surface Bottom Surface
Fig. 4.15 Optical photographs of top and bottom surfaces of the laser welding sections
the liquid phase alters the surface tension gradient (dr/dT, r is the surface tension)
across the heated section while modifying the surface tension force on the melt
surface. This causes the convex melt profile at the surface, which appears as a
shallow cavity upon solidification on the surface of the weld. Moreover, increasing
viscosity at elevated temperature reduces the shear stress in the molten flow, which
results in the out flow of the molten metal from the bottom surface of the weld.
The close view of the SEM micrographs reveals that dense structures are formed in
the welding core region. The solidification cracking in this region is not observed.
This is associated with the relatively slow cooling rates along the laser beam axis.
It should be noted that when solidification strains exceed the ductility of the
solidifying weld material due to the formation of low melting-point liquid film
along the solidification grain boundaries, solidification cracking occurs. This is
known as the hot-cracking. In addition, the presence of silicon in the alloy con-
tributes positively to the formation of the hot-cracking in the weld core region.
SEM micrograph also reveals that the heat affected zone is narrow and dendrites
are formed in the region next to the weld core, which is associated with the high
cooling rates in this region. Moreover, no secondary dendrites are observed from
the micrographs.
Fig. 4.16 SEM micrographs of weld cross-section
Figure 4.17 shows microhardness distribution across the weld zone along the y-
axis. The microhardness is measured at mid-thickness of the workpiece. It is evi-
dent that microhardness attains the minimum at the laser beam axis and increases
towards the edge of the irradiated spot. The attainment of the minimum hardness at
the welding zone center is associated with the slow cooling rates in this region. The
maximum hardness is in the order of 1.4 times the base material hardness. However,
increasing the distance from the irradiated spot edge along the y-axis, microhard-
ness reduces. This is because of the heat affected zone formed in the region next to
the welding zone, i.e. microhardness reduces across the heat affected zone.
4.5.4 Welding of AISI 316 Steel
The results associated with the metallurgical changes in the weld section and
mechanical properties of the resulting welds are presented in line with the previous
studies [8, 9].
1.3
HARDNESS RATIO
1.2
1.1
1.0
0.0000 0.0001 0.0002 0.0003 0.0004 0.0005 0.0006
DISTANCE ALONG Y-AXIS (m)
Fig. 4.17 Microhardness variation along the core of the weld zone
Laser welded region primarily consists of the fusion and heat affected zones.
This can be seen from Fig. 4.18, in which cross-section of the weld is shown. The
fusion zone contains austenite phase consisting of dendrite and cellular forms due
to high cooling rates. It should be noted that the ratio of temperature gradient to
cooling rates determines the solidification mode of the microstructure in the fusion
zone; in which case, this ratio is the maximum in the fusion boundary and the
minimum in the weld center. Consequently, the locally scattered cellular structure
occurs in the fusion boundary, which changes to dendritic structures due to
reduction of this ratio at the weld center. The ferritic-pearlitic microstructure is
transformed to martensite at the top and bottom surfaces of the welding section
because of significantly high cooling rates from the weld surfaces. The solidifi-
cation cracking is not observed in the fusion zone indicating that the solidification
strains do not exceed the ductility of the solidifying weld metal. The close view of
the weld structure features an increased amount of austenite; which was formed
through coarsening of pre-existing austenite grain boundary and intergranular
sites. In this case, the partial decomposition of ferrite takes place by the growth of
austenite phase according to a Widmanstatten-type mechanism from original
coarse grain boundaries and needle-like aspects inside the ferrite grains. The EDS
analysis reveals that the austenitic former elements are concentrating in the den-
drites and the ferritic former elements are moving to the interdentritic structures.
Moreover, the fluctuation of Cr, Ni and Mo in fusion zone is larger than that in the
heat affected zone. This is attributed to the non-uniform crystallization process and
the phase transition and diffusion of alloying elements in the welding joints under
the high cooling rates. At the fusion zone interface, the fine grains consist of ferrite
and fine pearlite structures occur (Fig. 4.18). The pearlitic modulus grows from the
nuclei along the austenite boundaries and the scattered colonies of alternating
ferrite and cementite lamellae are formed within the pearlitic structure.
Fusion Zone
Fusion Zone
Boundary
SEM micrograph of laser welded section
Dendrites Optical Micrograph
Fusion Zone
Cellular Structures Heat Affected
Zone
Fusion/heataffected zone Fusion/heataffected zone
Cellular Structures
Dendrites
Fusion zone Close viewoffusion boundary
Fig. 4.18 SEM and optical micrographs of laser weld cross-section
Figure 4.19 shows XRD diffractograms of the laser welded surface. d-Fe, c-Fe
and a-Fe phases are present in the diffractogram. The high cooling rates at the
surface during the solidification results in incomplete transformation of d- c-
phases and metastable d-Fe could be remained, which is unavoidable. Figure 4.20
Fig. 4.19 XRD

diffractogram of laser welded
region
Fig. 4.20 Linear dependence

of d (211) with sin2w
shows the linear dependence of d (211) with sin2w. The slope of the curve is
9.55 9 10-3 nm and intercept is 0.11712 nm. The residual stress calculated at the
surface vicinity is tensile and it is on the order of 140 MPa. The tensile residual
stress is associated with deformations during the expansion of the surface region in
the cooling cycle; i.e. high cooling rates at the surface region is responsible for the
formation of tensile residual stress at the surface vicinity. When comparing the
residual stress predicted from the simulations and obtained from the XRD tech-
nique, it is evident that both results are in good agreement.
The microhardness measurement results show that the microhardness decays
towards the center of the fusion zone due to self-annealing effect in this region
during the cooling period. However, the microhardness increases towards the
fusion zone edges and decreases with increasing distance along the heat affected
zone. The increased hardness in the region of fusion zone edges is associated with
the martensite formation in this region due to the high cooling rates. Moreover,
microhardness becomes less than the base material hardness, which is associated
with the softening of the material in this region. The formation of soft zone is
attributed to tempering of pre-existing martensitic phase and bainite in the ferritic
matrix. The peak microhardness is 1.4 times of the hardness of the base material.
The three-point bending tests reveals that the elastic modulus of the weld varies for
different joining configurations of the welds; consequently, elastic modulus varies
within 50–65 GPa. This variation is attributed the size of the weld zone and the
cooling rates. The slow cooling rates able the weld section to become ductile while
reducing the elastic modulus. Nevertheless, the elastic modulus of the weld section
is well below that of the workpiece material indicating the softening of the weld
section enhancing ductility of the substrate material in the weld region.
The concluding remarks for the experimental findings include the laser forming
and welding processes and they are given herein.
The findings from the laser forming process reveal that the linear increase in the
bending angle with the laser output power intensity is associated with almost
constant temperature gradient in the vicinity of the melt pool wall. Laser scanned
tracks consists of overlapped laser spots due to high frequency laser pulsing. No
micro-cracks and crack network is evident at the treated surface. This is attributed
to the self-annealing effect of the initially formed laser tracks, which suppresses
the increase of the cooling rates. The laser treated layer cross-section consists of
two regions. In the first region, the melt pool is formed and ferritic-pearlitic
structures are formed into martensite in this region because of the high cooling
rates. The feathery like structure is formed in the surface vicinity. The presence of
nitride compound is also evident from the X-ray diffractorgam at the treated
surface because of the high pressure nitrogen assisting gas.
The results obtained from the welding of Hayness 188 workpiece material show
that the core section of the welding zone consists of the fine structures and the heat
affected zone next to the welding core is narrow. The dendritic structures are
observed in the vicinity of the welding core indicating the high cooling rates in this
region. The microhardness profile across the welding zone follows almost the
residual stress distribution in this region. The findings from the welding of AISI
316 steel indicate that the residual stress predicted agrees well with that obtained
from the X-ray diffraction data. von Mises stress reduces at the mid-thickness of
the weld section, which is associated with the self-annealing effect during the
cooling period in this region. Fine grains are formed at the front and back surfaces
of the weld due to high cooling rates. In the fusion zone, dendrites and cellular
structures are observed; in which case, dendritic structures are dominantly present
at the fusion boundary and scattered cellular structures occur at the weld central
region. The solidification cracking is not present in the fusion zone. The fine grains
4.6 Concluding Remarks 105
consisting of ferrite and fine pearlite structures take place at the fusion zone
boundary. The microhardness of the weld zone increases and the maximum
hardness is almost 1.4 times the base material hardness. The annealing effect in the
central region of the weld section lowers the microhardness in this region. The
residual stress predicted from the simulations agrees well with that obtained from
the X-ray diffraction technique. The three-point bending tests show that the elastic
modulus of the weld varies for different joining configurations of the welds;
consequently, elastic modulus varies within 50–65 GPa. This variation is attrib-
uted the size of the weld zone and the cooling rates. The slow cooling rates able
the weld section to become ductile while reducing the elastic modulus. Never-
theless, the elastic modulus of the weld section is well below that of the workpiece
material indicating the softening of the weld section enhancing ductility of the
substrate material in the weld region. In the case of laser spot welding of electrical
wires, the findings indicate that at some power intensities the mechanical prop-
erties of the welds improve. The multi core wires can be welded to any base
material giving a sufficient tensile strength. However, in the case of 2.5 mm thick
multi core wires, the tensile strength of the welds decrease to 20–40 N. A base
material having low thermal diffusivity and high absorption depth gives low tensile
strength and a large heat affected zone. This may be due to the presence of the
conduction effects during the welding process.
References
1. Yilbas BS, Davies R, Yilbas Z (1991) A study into laser spot welding of sheet metals using
oxygen and argon as assisting gases. J Mater Process Technol 15:139–151
2. Yilbas BS, Kar A (1997) Laser spot welding and efficiency consideration. J Mater Eng
Perform 6:766–770
3. Sunar M, Yilbas BS, Boran K (2006) Thermal and stress analysis of a sheet metal in welding.
J Mater Process Technol 172:123–129
4. Bayraktar E, Kaplan D, Yilbas BS (2008) Comparative study: mechanical and metallurgical
aspects of tailored welded blanks (TWBs). J Mater Process Technol 204:440–450
5. Yilbas BS, Arif AFM, Abdul Aleem BJ (2010) Laser welding of low carbon steel and thermal
stress analysis. Opt Laser Technol 4(5):760–768
6. Abu-Dheir N, Yilbas BS (2010) Quality assessment and metallurgical examination of laser
welded sheets. Adv Mater Res 83–86:611–615
7. Yilbas BS, Akhtar SS, Karatas C (2011) Laser welding of Hayness 188 alloy sheet: thermal
stress analysis. J Adv Manuf Technol 56(1–4):115–124
8. Abu-Dheir N., Yilbas BS (2011) Laser welding of steel laminates: measurement of elastic
modulus with 3-point bending, AMPT 2011, 14th International conference on advances in
materials & processing technologies, Istanbul, Turkey, 13–16 July, 2011
9. Yilbas BS, Akhtar SS (2013) Laser welding of AISI 316 steel: microstructural and stress
analysis, ASME. J Manuf Sci Eng 135(3), 034502 (May 27, 2013) (4 pages). doi:10.1115/
1.4024289
10. Yilbas BS, Arif AFM, Abdul Aleem BJ (2012) Laser bending of AISI 304 steel sheets:
thermal stress analysis. Opt Laser Technol 44:303–309
11. Yilbas BS, Davies R, Gorur A, Yilbas Z, Begh F, Kalkat M, Akcakoyun N (1990) Study into
the measurement and prediction of penetration time during CO2 laser cutting process. In:
Proceedings of the institution mechanical engineers, Part B, vol 204, pp 105–113
12. Khana ZA, Hadfield M, Tobe S, Wang Y (2005) Ceramic rolling elements with ring crack
defects—a residual stress approach. Mater Sci Eng A 404:221–226
13. Yilbas BS, Hyder SJ (2000) Laser spot welding of electrical wires. In: 33rd international
MATADOR conference, Manchester, pp 13–14

Laser Forming and Welding Processes

Uploaded by

Document Information

Copyright

Available Formats

Share this document

Share or Embed Document

Sharing Options

Did you find this document useful?

Is this content inappropriate?

Copyright:

Available Formats

Laser Forming and Welding Processes

Uploaded by

Copyright:

Available Formats

Materials Forming, Machining and Tribology

Bekir Sami Yilbas

Laser Forming and

For further volumes:

Shahzada Zaman Shuja

Laser Forming and

ISSN 2195-0911 ISSN 2195-092X (electronic)

Library of Congress Control Number: 2013940956

Ó Springer International Publishing Switzerland 2013

Printed on acid-free paper

Springer is part of Springer Science+Business Media (www.springer.com)

We would like to acknowledge to role of King Fahd University of Petroleum and

1 Introduction to Laser Forming and Welding Processes . . . . . . . . . 1

2 Analytical Treatment of Laser Forming

3 Numerical Analysis for Laser Forming and Welding . . . . . . . . . . . 39

3.4.4 Temperature and Stress Fields for Welding

4 Experimental Analysis for Laser Forming and Welding . . . . . . . . 85

B. S. Yilbas et al., Laser Forming and Welding Processes, 1

radiation on boundaries on the temperature distribution can be insignificant and a

1. Dearden G, Edwardson SP (2003) Some recent developments in two-and three-dimensional

2.1 Analytical Formulation of Heating and Stress Field

B. S. Yilbas et al., Laser Forming and Welding Processes, 5

Fig. 2.1 Schematic view of

T ¼ eðkvnÞ Fðn; y; zÞ ð2:5Þ

T ¼ AeðkvnÞ Ko ½kvr  ð2:12Þ

The transverse shrinkage force per unit length is:

The solution of Eq. 2.32 yields:

2.2 Analytical Formulation of Heating and Stress Field

where V is the vapor front velocity and can be written as [4]:

Therefore, the complete solution to the transformed equation (Eq. 2.42)

L1 ½uðp þ aÞ ¼ eat L1 ½uðpÞ ð2:49Þ

Finally the term:

After substituting the values of A1 and A2 in Eq. 2.71, it yields:

sÞ ¼ I1 dðh þ dkÞegx hT0 egx I1 dedx

The dimensionless temperature distribution for a full pulse is then:

I1 dðh þ dkÞc2 s2 hT0 c2 s

Substituting Eq. 2.87 in Eq. 2.85 yields:

The second part of the particular solution is:

Substituting Eq. 2.89 in Eq. 2.85 yields:

So, the general solution for the stress field becomes:

From the boundary condition ðx ¼ 1 ) rx ¼ 0Þ, this yields A3 ¼ 0.

Therefore, the Laplace inversion of Terms ðr

The Laplace transform of ðr

ðb þ ad2 Þðh2 a þ k2 bÞðc21 þ abÞ be Erf bðt x=c1 Þ

rx 11 ¼ r x 111 þ r x 112 þ r x 113 þ r x 114 þ r x 115 þ r x 116 þ r x 117 þ r x 118

rx 12 in dimensionless form is:

rx 12 ¼ r x 121 þ r x 122 þ r x 123 þ r x 124 ð2:151Þ

rx 2 in dimensionless form is:

rx 21 ¼ r x 211 þ r x 212 þ r x 213 þ r x 214 þ r x 215 þ r x 216 þ r x 217 þ r x 218

rx 22 ¼ r x 221 þ r x 222 þ r x 223 þ r x 224 þ r x 225 þ r x 226 ð2:166Þ

2.3 Findings and Discussions

The findings of the analytical treatment of laser heating process in relation to

2.3.1 Stationary Source Consideration

Fig. 2.2 Dimensionless temperature distributions within the substrate material

Fig. 2.3 Dimensionless stress distribution within the substrate material

2.3.2 Moving Source Consideration

developed due to temperature field. The magnitude of radial stress component

2.4 Concluding Remarks

3.1 Numerical Analysis for Laser Forming Process

B. S. Yilbas et al., Laser Forming and Welding Processes, 39

3.1.1 Solid Heating and Stress Field Formation

The numerical formulation of laser multi-beam heating of steel sheet surface is

3.1.1.1 Thin Sheet Metal Heating Analysis

Fig. 3.1 A schematic view

T ¼ AeðkvnÞ Ko ½kvr ð2:12Þ

L1 ½uðp þ aÞ ¼ eat L1 ½uðpÞ ð2:49Þ

rx 11 ¼ rx 111 þ rx 112 þ rx 113 þ rx 114 þ rx 115 þ rx 116 þ rx 117 þ rx 118

rx 12 ¼ rx 121 þ rx 122 þ rx 123 þ rx 124 ð2:151Þ

rx 21 ¼ rx 211 þ rx 212 þ rx 213 þ rx 214 þ rx 215 þ rx 216 þ rx 217 þ rx 218

rx 22 ¼ rx 221 þ rx 222 þ rx 223 þ rx 224 þ rx 225 þ rx 226 ð2:166Þ

x at infinity ð 1Þ: x ¼ 1 : Tð1; y; z; tÞ ¼ To (specified) ð3:5Þ

where ½D contains the elastic constants related to temperature-dependent elastic