Laeser Pulse Heating of Surfaces and Thermal Stress Analysis - BYilbas

Materials Forming, Machining and Tribology
Bekir S. Yilbas
Ahmad Y. Al-Dweik
Nasser Al-Aqeeli
Hussain M. Al-Qahtani
Laser Pulse
Heating of Surfaces
and Thermal Stress
Analysis
Materials Forming, Machining and Tribology
Series Editor
J. Paulo Davim
For further volumes:

http://www.springer.com/series/11181
Bekir S. Yilbas Ahmad Y. Al-Dweik

Nasser Al-Aqeeli Hussain M. Al-Qahtani

Laser Pulse Heating

of Surfaces and Thermal
Stress Analysis
123
Bekir S. Yilbas Ahmad Y. Al-Dweik
Nasser Al-Aqeeli Department of Mathematics and Statistics
Hussain M. Al-Qahtani King Fahd University of Petroleum
Mechanical Engineering Department and Minerals
King Fahd University of Petroleum Dhahran
and Minerals Saudi Arabia
Dhahran
Saudi Arabia
ISSN 2195-0911 ISSN 2195-092X (electronic)

ISBN 978-3-319-00085-5 ISBN 978-3-319-00086-2 (eBook)
DOI 10.1007/978-3-319-00086-2
Springer Cham Heidelberg New York Dordrecht London
Library of Congress Control Number: 2013940306
Springer International Publishing Switzerland 2014

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 Publishers 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
High power lasers can be used as the tools to treat and machine the materials in
industry with precision of operation, high processing speed, and low cost.
Although laser processing of metallic materials has several advantages over the
conventional methods, the development of high temperature gradients in the
treated region gives rise to the formation of high stress levels in the region treated.
In some instances, this limits the practical applications of the material processed
by a high power laser beam. Although considerable research studies were carried
out to examine the thermal stress developed during laser processing, the models
developed have short comings to reflect the actual physical problems. This is
mainly because of the complicated nature of the problem, which involves with
multi-physics phenomena. Online experimentation of the thermally induced stress
field is extremely difficult and costly because of the limitations in sensing systems,
which operate at tremendously elevated temperatures during laser processing.
Therefore, the model studies still satisfactorily provide physical insight into the
thermally induced processes enabling to understand the relations between the
process parameters and thermal response of the treated material.
Although the physical processes related to the laser heating and thermal stress
development are complicated, analytical solution to the problem is possible
through incorporating some useful assumptions in the analysis. Since the tem-
perature gradient along the absorption depth is much higher than that of the
direction normal to the absorption depth, heating situation may reduce to one-
dimensional problem. In addition, the assumption of heating of a semi-infinite
body can be justified after comparing the depth of absorption with the thickness of
the substrate material, which is significantly larger than the absorption depth.
Analytical solution to the laser pulse heating problem yields the functional relation
between the dependent and independent parameters, despite the unavoidable
number of assumptions, which may be reduced in the numerical solutions. Nev-
ertheless, validation through the experimentation of the findings is necessary to
fulfill the required accuracy of the solutions.
In this book, thermal stress development during laser pulse heating of metallic
surfaces is formulated for various laser pulse parameters and heating conditions
incorporating the temporal variation of laser pulse intensity, the convection
cooling of the surface resembling the assisting gas, and the consideration of
v
vi Preface
surface and volumetric heat sources. In addition, equilibrium and non-equilibrium

heating situations are classified presenting the closed form solutions, accordingly.
Analytical solution is also presented for two-dimensional heating situation for non-
equilibrium energy transfer in the irradiated region. However, some cases pertinent
to laser heating and thermal stress development are not presented in this book due
to space limitations and, therefore, these cases are left for the future treatments.
Acknowledgments
We would like to acknowledge the role of King Fahd University of Petroleum and
Minerals in extending a strong support from the beginning to end facilitating every
means during the preparation of this book. The author wishes to thank the col-
leagues who contributed to the work presented in the book through previous
cooperation of the author. In particular, thanks to Dr. Muammer Kalyon, Dr. S. Z.
Shuja, Dr. Saad Bin Mansoor, and all our graduate students.
vii
Contents
1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 1
2 Equilibrium Laser Pulse Heating and Thermal Stress Analysis . . . 5

2.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 5
2.2 Step Input Laser Pulse Heating . . . . . . . . . . . . . . . . . . . . . . . . 6
2.2.1 Stress Free Boundary at the Surface . . . . . . . . . . . . . . . 6
2.2.2 Stress Continuity Boundary at the Surface . . . . . . . . . . . 19
2.3 Time Exponentially Varying Laser Pulse Heating . . . . . . . . . . . 26
2.3.1 Stress Free Boundary at the Surface . . . . . . . . . . . . . . . 26
2.3.2 Stress Free Boundary and Convection at the Surface. . . . 36
2.3.3 Stress Boundary at the Surface . . . . . . . . . . . . . . . . . . . 54
2.4 Entropy Analysis Due to Thermal Stress Field . . . . . . . . . . . . . 64
2.5 Findings and Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 67
2.5.1 Step Input Pulse Heating . . . . . . . . . . . . . . . . . . . . . . . 67
2.5.2 Time Exponentially Varying Laser Pulse Heating . . . . . . 71
2.5.3 Entropy Analysis Due to Thermal Stress Field . . . . . . . . 79
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 84
3 Analytical Solution of Cattaneo and Thermal Stress Equations . . . 85

3.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . ........ . . . 85
3.2 Surface Heat Source Consideration . . . . . . . . . . ........ . . . 86
3.2.1 Step Input Pulse Heating . . . . . . . . . . . . ........ . . . 87
3.2.2 Exponential Pulse Heating . . . . . . . . . . . ........ . . . 92
3.3 Volumetric Source Consideration . . . . . . . . . . . ........ . . . 95
3.3.1 Step Input Pulse Heating . . . . . . . . . . . . ........ . . . 96
3.3.2 Exponential Pulse Heating . . . . . . . . . . . ........ . . . 103
3.4 Entropy Analysis . . . . . . . . . . . . . . . . . . . . . . . ........ . . . 108
3.5 Findings and Discussion . . . . . . . . . . . . . . . . . . ........ . . . 110
3.5.1 Surface Heat Source Consideration . . . . . ........ . . . 110
3.5.2 Volumetric Heat Source Consideration . . ........ . . . 114
3.5.3 Entropy Generation Rate . . . . . . . . . . . . ........ . . . 117
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ........ . . . 119
ix
x Contents
4 Analytical Treatment of Hyperbolic Equations

for Stress Analysis . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 121
4.1 Introduction . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . ...... 121
4.2 Formulation of Energy Transport in Metallic Substrates
at Microscopic Level . . . . . . . . . . . . . . . . . . . . . . . . . ...... 122
4.3 Thermal Stress Field: Consideration of Surface
and Volumetric Sources . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 134
4.3.1 Surface Heat Source Consideration . . . . . . . . . . . . . . . . 134
4.3.2 Volumetric Heat Source Consideration . . . . . . . . . . . . . 140
4.4 Thermal Stress Field: Two-Dimensional Consideration . . . . . . . 147
4.5 Findings and Discussions . . . . . . . . . . . . . . . . . . . . . . . . . . . . 153
4.5.3 Two-Dimensional Analysis. . . . . . . . . . . . . . . . . . . . . . 160
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 164
5 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

5.1 Equilibrium Heating. . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167
5.1.1 Step Input Laser Pulse Intensity . . . . . . . . . . . . . . . . . . 168
5.1.2 Exponential Laser Pulse Intensity . . . . . . . . . . . . . . . . . 168
5.2 Cattaneo Heating Model and Thermal Stresses . . . . . . . . . . . . . 170
5.2.1 Exponential Laser Pulse Intensity . . . . . . . . . . . . . . . . . 171
5.2.2 Step Input Laser Pulse Intensity . . . . . . . . . . . . . . . . . . 171
5.3 Non-Equilibrium Heating . . . . . . . . . . . . . . . . . . . . . . . . . . . . 173
5.3.3 Two-Dimensional Heating . . . . . . . . . . . . . . . . . . . . . . 174
References . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 174
Index . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 177
Chapter 1
Introduction
Abstract Laser heating offers considerable advantages over the conventional

methods such as precision of operation, local treatment, and low cost. Laser at high
intensity when interacts with the solid surface, the absorption takes place. This in
turn causes internal energy gain of the substrate material and heat release from the
irradiated region. Since the process, in general, is fast, temperature gradients remain
high in the irradiated region. This results in high thermal strain and thermally
induced stresses in this region. Moreover, in laser treatment process, the end product
is important from the application point view. The high stress levels formed in the
irradiated region may cause failure of the surface through stress induced cracking.
Consequently, a care needs to be taken during the laser treatment process. This
chapter provides the information about the importance and limitations of the laser
treatment process in terms of the thermal stresses formed in the irradiated region.
Lasers are considered to be one of the effective tools for laser treatment of metallic
surfaces. High power laser, when focused at the surface, generates excessive heat,
which enables the surface to reach the melting temperature of the substrate
material. In some cases, laser heating is associated with the phase change, such as
melting and evaporation at the surface. In this case the heating process becomes
complicated because of the size of the heated, which is in the order of fraction of
millimeter. Therefore, controlling the phase process becomes necessary to avoid
surface asperities, such as cavities, which are formed during surface evaporation.
One of the methods to secure the control melting at the surface is to model the
physical processes involved during the laser interaction with surface. Although
model studies involve with assumptions simplifying the mathematical arrange-
ments, experimenting the process is difficult, timely, and expensive because of the
high temperature involvement and small size of heated region during the short
time period. The time limitation in measurements is attributed to the laser pulse
length, which has to be shorter to be effective for the treatment of the surfaces.
Laser heating process can be classified into two categories. The first category
involves with the solid state heating without the phase change at the irradiated
surface. The heating is governed by the conduction process, which can be
B. S. Yilbas et al., Laser Pulse Heating of Surfaces 1

and Thermal Stress Analysis, Materials Forming, Machining and Tribology,
DOI: 10.1007/978-3-319-00086-2_1, Springer International Publishing Switzerland 2014
2 1 Introduction
categorized as conduction limited heating. The conduction heating is, in general,

governed by the Fourier heating law. This consideration has shortcomings when
the heating duration becomes comparable with the thermalization time of the
substrate material, which is on the order of fraction of nanoseconds for most of the
metals. The Fourier heating low fails to predict the correct temperature rise in this
region due to the consideration of infinite heat wave speed resulting in equilibrium
heating in the irradiated region. Since the absorption depth of the incident radiation
is small for metallic substrates, the heat wave propagation at a finite speed occurs
in the irradiated region. This is particularly true for laser short-pulse heating of the
metallic surfaces. In this case, the short heating duration of solid surfaces, by a
laser beam, initiates non-equilibrium energy transport in the irradiated region. In
addition, electrons gain excess energy from the irradiated field and increase their
temperature. This, in turn, results in thermal separation of electron and lattice sub-
systems in the irradiated region. Since electron excess energy transfer to lattice site
takes place through the collisional process, lattice site temperature increases
gradually as the heating period progresses. The rates of electron and lattice site
temperature increase change for different absorption depths and laser pulse shapes.
The energy transfer in the irradiated region governs by non-equilibrium transport,
which can be modeled through electron kinetic theory approach. The resulting
lattice temperature equation becomes hyperbolic due to wave nature of the heating
process during the short interaction time. Since the depth of absorption of the
irradiated field is considerably small and heating duration is extremely short, the
closed form solution of the resulting hyperbolic equation provides useful insight
into the heating process. On the other hand, the phase change takes place at high
laser power intensities. The irradiated material undergoes first melting and sub-
sequent evaporation during the heating period. Although phase change involves
with the energy transport by conduction and convection, the convection effect may
be neglected due to the small depth of the melt layer. Therefore, introducing the
convective boundary condition at the surface resembling the surface melting and
evaporation satisfies the physical phenomena taking place during the high intensity
laser irradiation at the surface. Moreover, this assumption simplifies the heating
problem such that the analytical solution to the resulting energy equation becomes
feasible.
In laser heating, thermal strain is developed because of the presence of the
temperature gradient in the irradiated solid substrate. Depending on the thermal
expansion coefficient of the substrate material and the temperature gradient,
thermal strain results in excessive thermal stress levels in the substrate material. In
laser non-conduction heating situation, melting and evaporation of the substrate
material occur. Since the materials expands freely in the vapor and molten phases,
the stress level reduces to zero in these phases. Moreover, below the melt zone
conduction heating takes place in the solid substrate during the laser surface
treatment process. The thermal stresses can only developed in the solid phase of
the substrate material and the energy transfer taking place in the solid substrate is
limited with the conduction heating process. Consequently, the solid phase of the
substrate material is subjected to the stress levels during the surface treatment
1 Introduction 3
operations. However, the evaporation of the surface results in recoil pressure

development across the vapor front and the melt surface. The magnitude of recoil
pressure depends on the laser power intensity across the irradiated spot and
material properties. Since the laser pulse, in general, decays exponentially with
time, the pressure is expected to decay with a similar trend of the power intensity.
The recoil pressure can be considered as a stress source, which acts externally at
the surface of the substrate material. In some cases, such as low rate of evapo-
ration, the recoil pressure induced forces can be ignored at the surface. This allows
introducing stress free condition at the surface. However, the presence of thin films
at the surface causes stress continuity at the interface between the substrate
material and the coating. In this case, the stress gradient may be considered ay zero
rather than the stress free surface.
Laser short-pulse heating results in non-equilibrium energy transfer in the
irradiated region and the speed of the thermal cannot be ignored as indicated
earlier. However, lattice-site temperature rise becomes gradual during the laser
short-pulse heating because of the collisional energy transport between the excited
electrons and the lattice phonons. The thermal stress is developed in the lattice
sub-system only because of the work done against the thermal expansion.
Application of the classical treatment of thermal stress analysis is useful to
understand the mechanical response of the laser irradiated region to the short-pulse
thermal loading. In addition, the energy deposited in a laser irradiation pulse is
stored in the substrate material in terms of the internal energy gain and the
mechanical work done during the thermal expansion. If the heating process takes
place in a short time period, thermal energy transport becomes non-equilibrium
while influencing the mechanical response of the material in the heated region.
Although the thermomechanical coupling between the temperature and stress
fields is small, it cannot be neglected for the non-equilibrium heating situation.
Consequently, uncoupled solutions of thermal and mechanical fields lead to the
incorrect results associated with the actual physical process. On the other hand,
during the long laser pulse heating (pulses longer than the thermalization time),
thermomechanical coupling may not be important in the energy equation so that
the uncoupled solution of the problem becomes visible to explore both the tem-
perature and the stress fields. The analytical solution for the coupled non-equi-
librium energy transport or for uncoupled equilibrium energy transport provides
the functional relations between the laser parameters and the workpiece properties
in time and space for thermal and stress fields.
Chapter 2
Equilibrium Laser Pulse Heating
and Thermal Stress Analysis
Abstract When the heating duration becomes greater than the thermalization time
of the substrate material, equilibrium heating takes place in the laser irradiated
region. In this case, the classical Fourier heating law governs the energy transport.
Although the heating process is complicated, some useful assumptions enable to
obtain the closed form solution for temperature and stress fields. Since the ana-
lytical solution provides the functional relation between the dependent variable
and the independent parameters, it provides better physical insight into the heating
problem than that of the numerical analysis. In this chapter, equilibrium heating of
solid surfaces heated by a laser beam is considered. The closed form solution for
the resulting temperature and stress fields are presented for various heating situ-
ations. The study also covers the phase change taking place at the irradiated region
during the laser treatment process.
2.1 Introduction
Laser pulse heating of metallic surfaces can be modeled using the Fourier heating
law for the pulse lengths greater than the thermalization time of the substrate
material. Laser pulses can be considered as a step input pulse type or a time
exponentially decaying pulses. In addition, the boundary conditions are important
and the solution changes for different boundary conditions resembling the actual
physical configurations. Therefore, the analytical solution for the laser heating
problem depends on the pulse type and the boundary conditions incorporated in
relation to the physical situations. During the laser heating process, thermal strain is
developed because of the temperature gradient generated in the solid substrate.
Depending on the thermal expansion coefficient of the substrate material and the
temperature gradient, thermal strain results in excessive thermal stress levels in the
substrate material. Therefore, thermal stress developed because of step input laser
and time exponentially varying pulses for various boundary conditions are con-
sidered, and the closed form solutions for temperature and stress fields are presented
in line with the previous studies [18].

6 2 Equilibrium Laser Pulse Heating and Thermal Stress Analysis
The laser spot diameter at the workpiece surface is small and heat transfer in the
radial direction is considerably smaller than that corresponding to the axial direc-
tion. Therefore, one-dimensional modeling of the laser heating process predicts
reasonably accurate temperature distribution at the surface and inside the substrate
material. In addition, the absorption depth of laser irradiated radiation is consider-
ably small in the surface as compared to the thickness of the substrate material;
consequently, the substrate material can be assumed as a semi-infinite body.
2.2 Step Input Laser Pulse Heating
The boundary conditions influence the thermal stress distribution in the irradiated
region. In, general two-boundary conditions can be considered associated with the
practical laser heating applications. The first boundary condition involves with
stress free surface where the surface free to expand during the heating cycle. The
second boundary condition is stress continuity at the surface because of the
presence of the film at the surface. The closed form solutions for laser heating and
thermal stress development are given below the appropriate sub headings in line
with the previous studies [13].
2.2.1 Stress Free Boundary at the Surface
The laser heating pulse can be considered as a step input pulse. In the actual laser
pulse, there exists the pulse rise and fall times and the actual laser pulse deviates
slightly from the step input pulse due to small pulse rise and fall times. However, it
is convenient to use a step input pulse in the analysis for the mathematical sim-
plicity. In order to construct a single step intensity pulse, it is necessary to subtract
two unit step functions with a time shift between them. The first step function will
start at t 0 and the other will start at t Dt. The step intensity pulse is therefore:
SPt 1t 1t Dt 2:1
where:
( )
1; jt [ 0
1t 2:2
0; t\0
and
( )
1; t [ Dt
1t Dt 2:3
0; t\Dt
The mathematical arrangement of temperature rise in the solid substrate due to

a single step intensity pulse is given here; however, the analyses related to two
2.2 Step Input Laser Pulse Heating 7
successive pulses are similar to the single step intensity pulse and, therefore, it is
avoided herein. The Fourier heat transfer equation for a laser heating pulse can be
written as:
o2 T I 1 d 1 oT
2
C1 SPtedx 2:4
ox k a ot
where:
I1 1 rf Io 2:5
Initially, the substrate material is assumed to at a constant uniform temperature;
therefore, the initial condition is:
At t = 0 ) Tx; 0 T0 2:6
Since the heating duration is short and no assist gas cooling is assumed;
therefore, the radiation and convective losses from the surface are negligible.
Consequently, the corresponding boundary condition is:

oT
At x 0 ) 0 2:7
ox x0
Since the depth of the irradiated region is limited with the depth of absorption,
the size of the absorption depth is considerably smaller than the depth of the
workpiece; therefore a semi-infinite workpiece is considered. This assumption
leads to a boundary condition of constant temperature at an infinitely depth below
the surface, i.e. the influence of the laser heating pulse is negligible at a depth
infinitely long from the surface. Therefore,
At x 1 ) T1; t 0 2:8
The Laplace transform of Eq. 2.4 with respect to t, results:
o2 T I1 d 1
2
C1 SPsedx sTx; s Tx; 0 2:9
ox k a
where:
1 eDts
SPs 2:10
s s
Introducing the initial condition and rearranging Eq. 2.9 yields:
o2 T

I1 d C1 SPsedx T0
k2 T 2:11
ox2 k a
where k2 = s/a and s is the transform variable. Equation 2.11 has the solution:
I1 d C1 SPs dx T0
Tx; s Aekx Bekx e 2:12
k d2 s=a s
where A and B are constants. Introducing the boundary conditions determines the
constants A and B, i.e.:
I1 d2 C1 SPs
B 0; and A :
kkd2 s=a
Substitution of A and B in Eq. 2.12 yields:
2 p kx
s I1 d apC 1 SPse I1 dC1 SPsedx T0
Tx; 2 2:13
k s d k2 k d2 k2 s
which gives the thermal solution in the Laplace transform domain.

The inverse Laplace transform of Eq. 2.13 gives the temperature distribution
inside the substrate material in non-dimensional form. This is possible by defining
dimensionless quantities, which are:
Tkd
x xd : t ad2 t : T 2:14
I1
The dimensionless temperature distribution becomes:
8 2
39
> x
p x >
>
> e erfc t p
7>>
> 6
>
>
>
x t
e 6 t 2 t
7>>
>
>
>
> e e U1 6 7 >
>
< 2 6
7 =
4 x p x 5
T T0 C1 e erfc
t p
>
> 2 t >
>
>
> >
>
>
> p
>
>
>
> 2 t x2 x >
>
>
: p e 4t x erfc p
>
;
p 2 t
8 x t Dt 9
>
> e e U1 Dt >
> 2:15
>
> 2
3>>
>
> p >
>
>
> x

x >
>
>
> 6
e erfc t Dt p 7 >
>
>
> e

t Dt 6 2 t Dt 7 >
>
>
< 6 7>=
2 4 6
7
C1 p x 5
>
> e erfcx
t Dt p >
>
>
> 2 t Dt >
>
>
> >
>
>
> " p
# >
>
>
> Dt >
>
>
> 2 t x2
4t Dt x >
>
>
: p e x erfc p >
;
p
2 t Dt
where erfc is the complementary error function, which is:

Zz
2 2
erfcz 1 erf z and erfcz p et dt
p
0
Equation 2.15 is the closed form solution for temperature distribution.

To solve for the stress distribution inside the substrate material, equation
governing the momentum in a one-dimensional solid for a linear elastic case can
be considered [9], i.e.:
o2 rx 1 o2 rx o2 T
2
2 2 c2 2 2:16
ox c1 ot ot
where c1 is the wave speed in the solid
s
E 1t
c1 and c2 qaT
q 1t
where t is Poissons ratio, q is the density of the solid and aT is the thermal
expansion coefficient of the solid.
Initial and boundary conditions for the temperature field in Eq. 2.16 are similar to
those given for Eq. 2.4. Initially substrate material is considered as free from the
stress. In addition, as time extends to infinity, the stress field vanishes in the substrate
material. Therefore, the initial and final conditions for the stress field become:
At t 0 ) rx 0 2:17
and
At t 1 ) rx 0 2:18
The consideration of no mechanical force at the surface prior to laser irradiation
pulse leads to stress free boundary conditions at the surface. In addition, the
effective depth of laser irradiation is considerably smaller than the workpiece
thickness. Therefore, the assumption of semi-infinite body holds in the stress
analysis. In this case, the thermal strain disappears at a depth infinitely long from
the surface. Consequently, the corresponding boundary conditions yield:
At x 0 ) rx 0 2:19
and
At x 1 ) rx 0 2:20
Taking the Laplace transform of Eq. 2.16 with respect to time yields:
o2 r
x 1 :
2
2 s2 rx x; s srx x; 0 rx x; 0
ox c 2:21
h1 : i
s sTx; 0 Tx; 0
c2 s2 Tx;
where r s are the Laplace transforms of thermal stress and tem-

x x; s and Tx;
perature, respectively in the x and s domains.
By substituting the initial conditions, Eq. 2.21 reduces to:
o2 r
x s2 s c2 sT0
x x; s c2 s2 Tx;
r 2:22
ox2 c1 2
Considering the temperature distribution for a pulse with exponential temporal

variation, Eq. 2.13, and substituting into Eq. 2.22 and solving for the stress field,
yields:
" #
2 p
o2 r
x s2 2 I1 d aC1 SPsekx I1 dC1 SPsedx T0
x c2 s
r p c2 sT0
ox2 c1 2 k s d2 k2 k d2 k2 s
2:23
The complementary and the particular solutions for Eq. 2.23 are:
sx sx
r
x c D1 ec1 D2 e c1 2:24
While the particular solution has two parts, the first part is:
ps
x p1 D3 e ax
r 2:25
The second part of the particular solution is:
x p2 D4 edx
r 2:26
Solving for the particular solutions yields:

p
c2 s2 I1 d2 C1 SPs a
D3 p 2 2:27
k s d k2 k2 s2 =c21
and
c2 s2 I1 dC1 SPs
D4 2 2:28
k d k2 s2 =c21 d2
So, the general solution for the stress field becomes:
sx sx
x g D1 ec1 D2 e c1 D3 egx D4 edx
r 2:29
Form the boundary condition (x = ? ) rx = 0), it yields D1 = 0:

Then,
sx
ps
x g D2 e c1 D3 e ax D4 edx
r 2:30
x 0, the
Consider the boundary condition at the surface, where at x 0 ) r
constant in Eq. 2.30 becomes:
D2 D3 D4 2:31
Then:
ps sx sx
x x; s D3 e
r ax D3 e c1 D4 edx D4 e c1 2:32
Finding the solution for rx in the x and t domain, we should take the inverse
Laplace transform for each term in Eq. 2.32. To accomplish this, the following
definitions are introduced:
ps sx
Term1 D3 e ax Term2 D3 esx
c1
2:33
Term3 D4 edx Term4 D4 e c1
Consequently, the solution for the stress distribution becomes the summation of
the inverse Laplace transforms of the above terms.
Therefore, the inverse of Laplace transform of Terms (Term1, Term2, Term3,
Term4) are:
p " ps ps #
I 1 d2 C 1 c 2 a e ax e ax eDts
Term1 = p 2 p 2
k s d s=a 1=a s=c21 s d s=a 1=a s=c21
2:34
After partial fraction of the above equation:
Term1
2 ps p 4 psx p ps 3
ae ax sa e a c21 a se ax
p 6 psd2 a2 d2 c2 sa c2 d2 a2 d2 c2 ad2 s 7
I 1 d2 C 1 c 2 a 6 6 1 1 1
7
7
6
k 6 p
as x Dts p 4 psx Dts 2
p psx Dts 77
4 ae e sa e a e c1 a se a e 5
p 2 2 2 2

2 2 2 2
2

sd 2
a d c1 sa c1 d a d c1 ad s
2:35
p
let G1 I1 d Ck1 c2 a, then:
2
Term1 G1 Term11 G1 Term21 G1 Term31 2:36

where
ps p 4 psx
ae s
1 eDts
a sa e a 1 eDts
Term11 p 2 ; Term21 2 and
sd a2 d c1 2 sa c1 2
p ps 2:37
c1 2 a se ax 1 eDts
Term31 2
s2 a2 d c1 2 ad2 s
The inverse of Laplace transform of Term1 is -1[Term1], i.e.:
1 Term1 1 Term11 1 Term21 1 Term31 2:38

The inverse of Laplace transform of terms becomes:

"

!#
1 a 1 x2 1 x2
G1 Term11 G1 2 p exp p exp
d pt 4at pt Dt 4at Dt
2:39
1 G1 Term21
2 2
r
3
2 x c1 c21 tc1 x t x
6 p
exp p e a a erfc c1 p 7
6 pt 4 a a 2 at 7
6 r !
7
6 c1 c21 tc1 x t x 2 x2 7
6 p e a a erfc c1 p p exp 7
6 a 2 at 4at Dt 7
6 a pt Dt 7
a3 6 ! 7
G1 2 6 r 7
2 a2 d c21 66 c1 c21 tDtc1 x t Dt x 7
7
6 p e a a erfc c1 p 7
6 a a 2 at Dt 7
6 r ! 7
6 2 7
4 c1 c1 tDtc1 x t Dt x 5
p e a a erfc c1 p
a a 2 at Dt
2:40
1 G1 Term31
2 2

3
2 x pad2 tdx p x
6 p exp a erfc d at p 7
6 pt 4at 2 at
6

7
7
6 p 2 p x 2 x 2
7
6 adead tdx erfc d at p p exp 7
6 2 at pt Dt 4at Dt 7
c21 a 6 ! 7
G1 2 2 6 7
2
2d c1 a d 62 6 p ad2 tDtdx p x 7
ade erfc d at Dt p
7
6 7
6 2 at Dt 7
6 ! 7
6 p 7
4 p ad2 tDtdx x 5
ade erfc d at Dt p
2 at Dt
2:41
However, the Term 2 is:
p " sx sx
#
c2 d2 I1 C1 a e c1 e c1 eDts
Term2 p 2 p 2
k s d s=a 1=a s=c21 s d s=a 1=a s=c21
2:42
Let:
p
c2 d2 I1 C1 a
G2 2:43
k
After the partial fraction, Term2 can be written as:
Term2 G2 Term12 Term22 Term32 2:44
where
sx p s x
a e c1 c21 sae c1
Term12 2 p : Term22 2
d s d c1 2 a2 d2 ad2 s
and
p 4 cs x
sa e 1
: Term32 2 2 2:45
d a d c1 2 as c1 2
2
The inverse of Laplace transform of Term 2 comes out to be:
1 Term2 1 G2 Term12 1 G2 Term22 1 G2 Term32 2:46

or
"
1 a 1 x
G2 Term12 G2 2 p 1 t
d pt x=c1 c1

# 2:47
1 x
p 1 t Dt
pt Dt x=c1 c1
and
1 G2 Term22
2 r
!
3
1 p ad2 t p x x
6 p
a de erf a d t 1 t 7
c2 a 6 pt x=c1 c1 c1 7
6 7
G2 2 12 2 6 r
!
7
2
d a d c1 46 p ad2 tDt p 7
1 x x 5
p ade erf ad t Dt 1 t Dt
pt Dt x=c1 c1 c1
2:48
and
1 G2 Term32
2 r
!
3
1 c1 c21 t c1 x x
6 p p e a erf p t 1 t 7
a3 6 pt x=c1 a a c1 c1 7
6 7
G2 2 6 r
!
7
a2 d c1 6
2
4 1 c1 c21 tDt c1 x x 5
7
p p e a erf p t Dt 1 t Dt
pt Dt x=c1 a a c1 c1
2:49
where erf (y) is the error function of the variable y.
However, Term3 is:
" #
c2 I1 dC1 sedx sedx eDts
Term3 2:50
k s2 =c21 d2 d2 s=a s2 =c21 d2 d2 s=a
Let:
c2 I1 dC1
G3 2:51
k
The Term3 can be written as:
After the partial fraction, the Term 3 becomes:
c21 aedx c21 aedx

Term13 : Term23
2ds c1 dc1 ad 2ds c1 dc1 ad
and
c1 2 a2 edx
Term13 2:53
c1 2 a2 d2 s ad2
The inverse of Laplace transform of Term3 comes out to be:

where
c21 aedx h c1 dtDt i

1 G3 Term13 G3 e ec1 dt 2:55
2dc1 ad
and
c21 aedx h c1 dt i
1 G3 Term23 G3 e ec1 dtDt 2:56
2dc1 ad
and
c21 a2 edx h ad2 t ad2 tDt

i
1 G3 Term33 G3 e e 2:57
c21 a2 d2
However, Term4 is:
" #
sx sx
c2 I1 dC1 se c1 se c1 eDts
Term4
k s2 =c21 d2 d2 s=a s2 =c21 d2 d2 s=a
2:58
Let:
c2 I1 dC1
G4 2:59
k
Term4 can be written as:

After partial fraction of Term4, it yields:
sx sx
c1 2 ae c1 c1 2 ae c1
Term14 : Term24
2ds c1 dc1 ad 2ds c1 dc1 ad
and
sx
c1 2 a2 e c1
Term34 2:61
c1 2 a2 d2 s ad2
The inversion of Term 4 is:

where
1 G4 Term14
"

#
c21 a c1 d tDtcx x c1 d tcx x
G4 e 1
1 t Dt e 1
1 t
2dc1 ad c1 c1
2:63
and
1 G4 Term24
"

#
c21 a c1 d tcx x c1 d tDtcx x
G4 e 1
1 t e 1
1 t Dt
2dc1 ad c1 c1
2:64
and
1 G4 Term34
"

#
c21 a2 ad2 tcx x ad2 tDtcx x
G4 2 e 1
1 t e 1
1 t Dt
c 1 a2 d2 c1 c1
2:65

where 1 t cx1 is a unit step function.
The closed form solution of stress distribution can be written as:
rx x; t 1 Term1 1 Term2 1 Term3 1 Term4 2:66

where -1 represents the inverse sign of Laplace transform.
Presenting the stress distribution in dimensionless form, the additional

dimensionless quantities are defined, i.e.:
c1 krx
c1
and rx 2:67
ad I1 da2 c2 C1

U[1] is the unit step function, which is U1 t cx , and U1 Dt is the

1
unit step function for the time shift, i.e. U1 Dt 1 t Dt cx

1
Therefore, for the dimensionless stress distribution, the following results are
obtained:

rx 1 rx 11 rx 21 rx 31 2:68
where
2

1 x 1 x2
rx 11 p exp p exp 2:69
pt 4t pt Dt 4t Dt
and
2 2

3
2 x c2 t c1 x
p x

6 p exp c1 e erfc c1 t p
1
pt 4t 2 t 7
6
7
6 p
7
6 2 x 7
6 c1 ec1 t c1 x erfc c1 t p 7
6 2 t 7
6
7
6 2 x 2 7
6 7
1 6 p exp 7
rx 21 2 6 pt Dt 4t Dt 7
2 1 c 1 6 !7
6 p 7
6 x 7
6 c1 ec2
1 t Dt c1 x erfc c
t Dt p 7
6 1
2 t Dt 7

6 7
6 ! 7
6 p
7
4 c2 t Dt c1 x
x 5
c1 e 1 erfc c1 t Dt p

2 t Dt
2:70
and

rx 31
2 2

3
2 x t x
p x p x
p exp e erfc t p et x erfc t p
6 pt 4t 2 t 2 t 7
6 7
6
!7
6 2 x 2 p x 7
c 2 6 7
1 2 6 p exp
et Dt x erfc t Dt p 7
2 1c1 6 6 p t Dt 4t Dt 2 t Dt 7
7
6 ! 7
6 p x 7
4
t Dt x
5
e erfc t Dt p
2 t Dt
2:71

However, rx 2 is:

rx 2 rx 12 rx 22 rx 32 2:72
where
1 1
rx 12 q
U1 q
U1 Dt 2:73

p t x =c1
p t Dt x =c1
and
20 s!
1 3
1 x
B
6 @q et erf
C 7
6 t A U1 7
6 p t x =c c1 7
2
c1 6 1 7
rx 22 6 0 1 7
1c2 6 s! 7
1 6 7
6 B 1
Dt
x C 7
4 @q
e t
erf t Dt A U 1Dt 5

p t Dt x =c1 c 1
2:74
and
2 0 1 3
s!
1 x
6 B c2 t C 7
6 @q
c1 e 1 erf c1 t A U1 7
6
p t x =c1 c1 7
1 6 7
rx 32 6 0 1 7
1 c2 6 s
! 7
1 6 7
6 B 1 c2 x C 7
4 @q
c e 1 t Dt erf c
t Dt
A U 1 Dt 5
1 1
c1
p t Dt x =c1
2:75

The term rx 3 is:

rx 3 rx 13 rx 23 rx 33 2:76
where
x h i
c2
1 e
rx 13
ec1 t Dt ec1 t 2:77
2 c1 1
and
c2 ex h c t i

rx 23 1 e 1 ec1 t Dt 2:78
2 c1 1
and
ex h t i

c2
1
rx 33 2 e et Dt 2:79
c1 1

The term rx 4 is:

rx 4 rx 14 rx 24 rx 34 2:80
where
" #
c2 c
t
Dt x
c
t x

rx 14 1 e U1 Dt e
1 c 1 c
1 1 U1 2:81
2 c1 1
and
" #
c2 c
t x
c
t
Dt x

rx 24 1 e
1 c 1 c
1 U1 e 1 U1 Dt 2:82
2 c1 1
and
" #
c2
1

t xc

t Dt xc
rx 34 2 e 1 U1 e 1 U1 Dt 2:83
c1 1
Consequently, the dimensionless form of stress equation becomes:

rx rx 1 rx 2 rx 3 rx 4 2:84
The solution of Eqs. 2.15 and 2.84 provides data for temperature and stress
fields. The mathematical arguments for the stress equation due to two successive
pulse heating situation are not given inhere due to the length arguments. However,
the closed form solution for the stress field (Eq. 2.84) can be used for heating
situation due to two successive pulses, provided that the following conditions are
satisfied:

rx rx first pulse rx second pulse 2:85

rx first pulse can be obtained after setting Dt t1 dt1 h1 in Eq. 2.84, where dt*1
is the pulse length of the first pulse and h1 is the cooling period after the first pulse.
It should be noted that the duration of the cooling period of the first pulse is the
same of the first pulses length (h1 = dt*1). This requires that 0 t t1 dt1 h1
(covering the heating and the cooling periods;
the second pulse initiates after the
cooling period of the first pulse. rx f secondpulse can be obtained when setting
Dt t2 dt2 h2 in Eq. 2.84, where dt*2 is the pulse length of the second pulse
and h2 is the cooling period after the second pulse. This requires that
t1 dt1 h1 t t2 dt2 h2 . In the present simulations the pulse lengths and
cooling periods of the first and second pulses of the two successive pulses are set
the same (dt*1 = dt*2 and h1 = h2).
2.2.2 Stress Continuity Boundary at the Surface
The zero stress gradient at the surface represents the case occurring during laser
heating of the coated surfaces. In this case, the surface of the substrate material is
coated by a high absorbent material, such as paint, to improve the absorptivity at
the surface. However, the absorbent material has different mechanical properties,
but the stress continuity across the substrate material and thin layer of the coat
occurs. This results in zero stress gradient at the surface of the substrate material.
The heat transfer equation for a laser step input heating pulse can be written
similar to Eq. 2.5, which is:
o2 T I1 d dx 1 oT
e 2:86
ox2 k a ot

where, I1 1 rf Io (similar to Eq. 2.5), x is the distance, t is the time, k is the
thermal conductivity, a is the thermal diffusivity, d is the absorption coefficient rf
is the reflection coefficient, and Io is the peak power intensity. Since the solution of
Eq. 2.86 is similar to Eq. 2.4, the boundary conditions and the solution of Eq. 2.86
is given lightly.
The initial and boundary conditions are:
At time t 0 ! Tx; 0 0 2:87
and

oT
At the surface x 0 ! 0
ox x0 2:88
and at x 1 ! Tt; 1 0
The solution of Eq. 2.86 can be obtained possibly through Laplace transfor-
mation method, i.e., with respect to t, the Laplace transformation of Eq. 2.86
yields:
o2 T I1 d dx 1
2
e sT Tx; 0 2:89
ox ks a
where T Tx; s. Using the initial condition, Tx; 0 0; Eq. 2.87 yields:
o2 T I1 d dx
q2 T e 2:90
ox2 ks
where q2 as . Equation 2.89 has a solution:
I1 ad
T Aeqx Beqx edx 2:91
ks s ad2
where A and B are the constants and they are calculated through the boundary
conditions. Substituting boundary condition, ooxT 0 at the surface (x 0), it gives:
I1 ad2
AB 2:92
qks s ad2
The boundary condition, T 0 at x 1, results A 0 in Eq. 2.91. Therefore,

B yields:
I1 d2
B 2:93
qks q2 d2
Hence, Eq. 2.91 becomes:
I1 ad I 1 d2
T 2
edx 2
eqx 2:94
ks s ad 2
qks q d
Rearrangement of Eq. 2.94 yields:

I1 ad dx I1 d eqx eqx
T e 2:95
ks s ad2 2kqs q d q d
The Inverse Laplace Transformation of Eq. 2.95 results:

I1 p x 1
Tx; t 2 atierfc p edx
k 2 at 2d

1 ad2 tdx p x
e erfc d at p 2:96
2d 2 at

1 ad2 tdx p x
e erfc d at p
2d 2 at
where erf is the error function, erfc is the complementary error function, and ierfc
is the integral of complementary error function, which are:
Z x
2 2
erf v p ev dv
p 0
erfcv 1 erf v 2:97
1 v2
ierfcv p e verfcv
p
Introducing dimensionless quantities as:
kd
s ad2 t : x0 xd : T 0 T 2:98
I1
Substituting the dimensionless quantities in Eq. 2.96, it yields:
0
p x 1 0
T 0 x0 ; s 2 sierfc p ex
2 s 2

1 sx0 p x0
e erfc s p 2:99
2 2 s

1 sx0 p x0
e erfc s p
2 2 s
Equation 2.99 is used to compute the dimensionless temperature profiles inside
the substrate material.
The equation governing the momentum in one-dimensional solid for a linear
elastic case can be used to formulate the thermal stress field in one-dimensional
semi-infinite solid. Therefore, Eq. 2.16 is used to formulate the momentum
equation.
The new initial conditions for Eq. 2.16 are:
At t 0 !; for thestress equation, rx 0

and at t 1 !; for the stress equation, rx 0
2:100
At t 0 !; for the heat transfer equation,T 0
and at t 1 !; for the heattransfer equation, T 0
In the case of zero stress gradient at the surface, the relevant boundary con-
ditions are:
orx orx
At x 0 !; for thestress equation, 0 and 0
ox ot
oT
and at x 0 !; for the heat transferequation, T 0: 0 2:101
ot
Atx 1 !; for the stress equation, rx 0
and at x 1 !; for the heat transferequation, T 0
The solution of thermal stress equation (Eq. 2.16) is possible using the Laplace
transformation method, i.e. Laplace transformation of Eq. 2.16 with respect to
t yields:
o2 rx 1 2 :
2 s rx x; s srx x; 0 rx x; 0
ox2 c1 2:102
h : i
c2 s2 T x; s sTx; 0 T x; 0
where rx x; s and T x; s are the Laplace transform of thermal stress and tem-
perature, respectively. Introducing the initial condition, Eq. 2.102 yields:
o2 rx s2
2 rx x; s c2 s2 T x; s 2:103
ox2 c1
However, it is noted from Eq. 2.94 that T is:

ps
T c1 e ax c2 edx 2:104
where:
p
I1 d d a
c1 p and
k s s d2 as
2:105
I1 d 1
c2
k s d2 as
Substitution of Eq. 2.104 into Eq. 2.103 yields:

" pp ps #
o2 r x s 2 I1 d d a se ax sedx
2 rx x; s c2 2 s 2 s 2:106
ox2 c1 k d a d a
If we define M1 and M2 as:

" pp #
I1 d d a s
M1 s c2 2 s
k d a
" # 2:107
I1 d s
M2 s c2 2 s
k d a
then Eq. 2.106 becomes:

o2 rx s2 ps
2
2 rx x; s M1 se ax M2 sedx 2:108
ox c1
Equation 2.108 has homogenous rx h and particular rx p solutions, i.e.:
rx rx h rx p 2:109
The homogenous solution is:
rx h D1 ec1 x D2 ec1 x
s s
2:110
and the particular solution has two parts. The first part is:
ps
rx p1 G1 e ax 2:111
Substituting Eq. 2.111 into Eq. 2.103 yields:

M1 s
G1 s s2
2:112
a c21
rx p1 G2 edx 2:113
Substituting Eq. 2.113 into Eq. 2.108 yields:

M2 s
G2 2:114
d2 cs2
2
Therefore, Eq. 2.109 becomes:

rx rx h rx p1 rx p2
or
s
x cs x M1 s pasx M2 s dx
rx D1 ec1 D2 e 1 s s2
e 2 s2 e 2:115
a c2
1
d c2
1
where D1 and D2 are constants and they will be calculated through boundary
conditions. The coefficient D1 must be zero, since c1 [ 0 then rx can be finite.
Equation 2.115 becomes:
ps
rx D2 ec1 x f1 se ax f2 sedx
s
2:116
where
2 3
pp
I1 d 4 d a s
f1 s c2 5 2:117
k sd2 s 1 s2
a a c1
or
" #
2 p
I 1 d 1
f1 s c2 c21 a a p 2 2:118
k s s ad2 s ca1
and
" #
I1 d s
f2 s c2 c21 a 2:119
k s ad2 s2 c21 d2
Let us define the constants C10 and C20 as:
I1 d2 p
C10 c2 c21 a a 2:120
k
and
I1 d
C20 c2 c21 a 2:121
k
D2 can be found through taking the derivative of Eq. 2.116 and setting to zero
and knowing that D1 is zero, as explained before, it yields:
r
s s
D2 f1 s df2 s 0 2:122
c1 a
Therefore, D2 becomes:
c1 c1
D2 p f1 s df2 s 2:123
sa s
Substituting D2 and setting D1 as zero into Eq. 2.108, it yields:
2 3
I1 da s
x 1 1
r c2 c31 e c1 4 c2 2
5
x k 2
s s ad s a 1 s ad s c 1 d s c 1 d
2 p 3 2:124
2 p as x cs
I1 d e e 1
c2 c21 a a4p c2 5
k s s ad2 s 1 s ad2 s c1 ds c1 d
a
Let us introduce the followings:

sx
e c1
Term1 c2

s s ad2 s a1
sx
e c1
Term2 2

s ad s c1 ds c1 d
p psx 2:125
ade a
Term3 p c2

s s ad2 s a1
sx
e c1
Term4 2

s ad s c1 ds c1 d
Using partial fraction expansion, the following relation can be obtained for Term1:
sx sx sx
1 e c1 c1 d e c1 c d e c1
Term1 4 2
2
4 1 2:126
c1 d s a2 d c21 d s ad c1 2 2 c2
2 c d s 1
a 1 a
The inversion of Laplace transformation of Term1 yields:

2 3
1 c1 d ad2
tcx
6 7

6 c1 d a2 d4 c2 d2 e
1
7
x 6 1 7
1
Term1 1 t 6 7 2:127
c1 6 6 c2 7
c1 d 1 t x
7
4 c4 e a c1 5
2 2
a2 c1 d
1
Similarly, the inversion of Laplace transformation of Term2 yields:

"
x c1 d
ead tc1
2
1 x
Term2 1 t 2 2
c1 ad c1 d ad c1 d
# 2:128
1 c1 dtcx1 1 c1 dtcx1
e e
2 c1 d ad2 2 c1 d ad2
Using partial fraction expansion, the following relation can be obtained for
Term3:
p " #
d a pasx 1 1
Term3 p e c2 2
2 2 2:129
s a ad
1
s ca1 ad2 ca1 s ad2
or
p " ps ps #
da a e ax e a x
Term3 2 p 2 p 2:130
ad c21 s s ca1 s s ad2
The inversion of Laplace transformation of Term3 yields:

8 2
39
> p x >
>
> e erfc d at p 7 >
dx
>
>
> ad2 t 6 2 at >
>
>
> e 6 7 >
>
>
> p 6
7 >
>
>
> 2d a 4 p x 5 >
>
>
> e dx
erfc d at p >
>
p >< >
=
da a 2 at
1
Term3 2 2 r
3 2:131
ad c21 >>
> xca1 t x >
>
>
>
> c2 6 e erfc c p 7 >
a 2 at 7 >
1
>
> eat 6
1
>
>
>
> p 6 r 7>>
>
> 2c1 a 4 6
7 >
>
>
> xc1 t x 5 >
>
>
: e a erfc c1 p >
;
a 2 at
The Laplace inversion of Term4 yields:
ad2 2 1
1 Term3 4 2 2
ead t 2
ec1 dt ec1 dt edx 2:132
2
a d c1 d 2 c1 d ad
Using the dimensionless quantities and summing the terms Term1, Term2,
Term3, and Term4 gives the Laplace inversion of dimensionless stress, i.e.:
(" c0 s x0 c0 s x00 0 2
#
ec1 sc1 x
0 0
I 1 d2
0
0 2 e 1 c1 e 1 c1
rx c2 c1 0 0 0 1 u
k 2 c01 1 2 c1 1 c1 c1 1

es 0 p x0 0 p x0
ex erfc s p ex erfc s p
2 1 c01 2 2 s 2 s
0 2

ec1 s c01 x0 0
p x0 c01 x0 0
p x0
e erfc c1 s p
e erfc c1 s p

2c01 1 c01 2 2 s 2 s
" #)
es 1 0
x0 c s c 0
s
e 2 0 e 1 e 1
1 c01 2 c1 1
2:133
0
As indicated earlier, when s\ cx0 (or t\cx1 ), the step function is zero (u 0) and
1
0 0
it has a value of one (u 1) when s [ cx0 . Therefore, during the time interval s\ cx0
1 1
the step function in Eq. 2.133 is set to zero (u 0) or else it is set to 1.
Equation 2.133 is used to compute the dimensionless stress distribution for zero
stress gradient at the surface.
2.3 Time Exponentially Varying Laser Pulse Heating
Since the laser power intensity varies with time, temporal variation of the pulse
intensity is incorporated in the stress analysis. In addition, the laser pulse
intensity distribution can be resembled through employing a time exponentially
varying pulse intensity profile. The boundary conditions are important formu-
lating the stress states in the irradiated material. Consequently, the closed form
solutions for the thermal stress generated inside the substrate material are pre-
sented due to different boundary conditions for the time exponentially decaying
laser pulse under the following sub-headings in line with the previous studies
[35].
2.3.1 Stress Free Boundary at the Surface
The Fourier heat transfer equation due to time exponentially decaying laser
heating pulse can be written as:
o2 T I1 d bt dx 1 oT
e e 2:134
ox2 k a ot
where I1 = (1-rf) Io.
It can be assumed that the convection losses from the surface are neglected
during the heating pulse. This implies the insulated boundary at the substrate
surface. In addition, since the substrate material is assumed to be semi-infinite
body, temperature at a depth approaching to infinity becomes the same as the
initial temperature of the substrate material. Therefore, the corresponding
boundary conditions are:
The insulated boundary condition at the surface yields:
At the free surface:

oT
x 0 ) 0 2:135
ox x0
2.3 Time Exponentially Varying Laser Pulse Heating 27
Temperature can be considered as approaching an initial temperature at a depth

infinitely below the surface, i.e.:
At x 1 ) T1; t 0 2:136
Due to the simplicity the initial temperature of the substrate material is set
to zero and temperature is assumed to be uniform inside the solid substrate at
time equal to zero (before heating initiates). Therefore, the initial condition
becomes:
At t 0 ) Tx; 0 0 2:137
The closed form solution of Eq. 2.134 can be obtained through a Laplace
Transformation method after employing the appropriate boundary and initial
conditions. In this case, the Laplace Transformation of Eq. 2.134 with respect to t,
results:
o2 T I1 d edx 1
sTx; s Tx; 0 2:138
ox2 k s b a
o2 T 2 I1 d edx
h T 2:139
ox2 k s b
where h2 = s/a and s is the transform variable. Equation 2.139 has the solution:
I1 dedx
Tx; s Aehx Behx 2:140
k s b d2 h2
where A and B are constants. Introducing the boundary conditions will allow
determining the constants A and B, i.e.:
I1 d2
A 0; and: B 2:141
khs b d2 h2
After substituting the values of A and B in Eq. 2.140, it yields:
" #
s I1 d d exphx expdx
Tx; 2:142
ks b h h2 d2 h2 d2
which gives the solution for temperature in Laplace domain.

The inverse Laplace Transform of Eq. 2.142 gives the temperature distri-
bution inside the substrate material in space (x) and time (t) domain as follows
[35]:
Tx;t
8 2 r!
3 9
>
> b x p >
>
>
> 6 exp ix Erfc p i bt 7 >
>
>
> r 6 a 2 at 7 >
>
>
> a 6 7 >
>
>
> id expbt 66 7 >
>
>
> r !
7 >
>
>
> b 6 p
7 >
>

>
< 4 b x 5 >
=
I1 d a exp ix Erfc p i bt
a 2 at
2k bad >2
> >
>
>
> 2

3>>
>
>
>
> x p p x >
>
>
> 6 expErfc p d at expdxErfc d at p 7 >
>
>
> 2 6 2 at 2 at 7 > >
>
> expad t4 5 >
>
>
> >
>
: ;
2expbt dx
2:143
where Erfc is the complementary error function. Equation 2.143 is the closed form
solution for temperature distribution. The temperature distribution in non-dimen-
sional form is possible by defining dimensionless quantities and substituting in
Eq. 2.143. The dimensionless quantities are:
Tkd b
x xd: t ad2 t: T :b 2 2:144
I1 ad
T x ;t
8 2 r! 3 9
>
x p
>
>
> b >
>
>
> 6 exp ix p i b 7 >
>
>
> r 6 t
2 t 7 >
>
>
> t 6 7 >
>
>
> i exp b
6 7 >
>
>
> b 6 r!
7 >
>

>> 6 b
x p 7 >
>
1 t < 4 exp ix
Erfc p i b 5 =
t
2 t
2 b t >
> >
>
>
> 2

3>>
>
>
>
> x p p x >
>
> exp x Erfc p t expx Erfc t p >
>
> 6 7 >
>
> exp t 6
2 t 2 t 7>
>
>
>
> 4 5 >
>
>
:
>
;
2expb x
2:145
To obtain the closed form solution for the stress distribution inside the semi-
infinite one-dimensional substrate material, equation governing the momentum in
a one-dimensional solid for a linear elastic case can be considered. In this case, the
momentum equation becomes the same as Eq. 2.16.
Since the substrate surface is free to expand during the heating process, a free
stress condition is adopted at the surface of the substrate material. Moreover, as the
depth below the free surface increases to infinity, the thermal stress approaches the
initial stress conditions, which is the stress free condition, i.e. it is assumed that
the substrate material is initially free from any stresses. Therefore, the corre-
sponding boundary conditions are:
At x 0 ) rx 0 2:146
and
At x 1 ) rx 0 2:147
For the temperature, the boundary conditions is:
oT
At x 0 ) 0 2:148
ox
and
At x 1 ) T 0 2:149
Since the substrate material is assumed to be initially stress free and the thermal
stress becomes the same as the initial stress condition as time approaches infinity,
the initial and final conditions for the stress field yield:
At t 0 ) rx 0 2:150
and
At t 1 ) rx 0 2:151
The initial condition for temperature is:
At t 0 ) Tx; 0 0 2:152
Since the laser pulse decays exponentially with time, the final and the initial
temperature of the substrate material becomes the same as the time approaches
infinity, i.e.:
At t 1 ) Tx; 1 0 2:153
Taking the Laplace Transformation of Eq. 2.16 with respect to time yields:
o2 r
x 1 2 :
2
2 s r x x; s srx x; 0 rx x; 0
ox c1 2:154
h : i
s sTx; 0 T x; 0
c2 s2 Tx;
where r s are the Laplace Transforms of thermal stress and tem-

x x; s and Tx;
perature respectively in the x and s domains.
o2 r
x s2 s
r x x; s c2 s2 Tx; 2:155
ox2 c21
Considering the temperature distribution for time exponentially varying pulse
and substituting into Eq. 2.155, and solving for the stress field, yields:

o2 r
x s2 2 I1 d d exphx expdx
r
x x; s c 2 s 2:156
ox2 c21 ks b hh2 d2 h2 d2
which can be arranged further:
o2 r
x s2 I1 d2 c2 s2 I1 dc2 s2
2rx x; s 2
ehx edx 2:157
ox 2 c1 ks bh h2 d ks b h2 d2
Now let W1 and W2 are defined as:
I1 d2 c2 s2
W1 2:158
ks bh h2 d2
and
I1 dc2 s2
W2 2:159
ks b h2 d2
Then Eq. 2.157 becomes:
o2 r
x s2
r x x; s W1 ehx W2 edx 2:160
ox2 c21
The homogeneous and the particular solutions for Eq. 2.160 are:
sx sx
r
x h C1 ec1 C2 e c1 2:161
while the particular solution has two parts, the first part is:
rx p1 Q1 ehx
2:162
Substituting Eq. 2.162 in Eq. 2.160 yields:

W1
Q1 2 2:163
h2 cs2
1
rx p2 Q2 edx
2:164

W2
Q2 22 2:165
d cs2
1
Therefore, the general solution for the stress field becomes:

sx sx
x g C1 ec1 C2 e c1 Q1 ehx Q2 edx
r 2:166
Form the boundary condition (x = ? ) rx = 0), it yields C1 = 0:

Then,
sx W1 hx W2
rx g C2 e c1
2 e 2 s2 edx 2:167
h2 cs2 d c2
1 1
Substituting for W1 and W2 in Eq. 2.167 results:

sx
rx g C2 e c1 g1 sehx g2 sedx
2:168
where:
I1 d2 c2 s2
g1 s 2:169
ks bh h2 d2 h2 cs2
2
and
I1 dc2 s2
g2 s 2:170
ks b h2 d2 d2 cs2
2
Now substituting for h s=a and simplifying the expressions for g1(s) and
g2(s) yields:
p
I1 d2 c2 saac21
g1 s c2
2:171
ks bh s ad2 s a1
and
I1 dc2 s2 c21

g2 s 2:172
ks b s ad2 s2 c21 d2
Rearranging the two expressions as:
" p #
s
g1 s C3 2:173
s bh s ad2 s c21 =a
and
" #
s2
g2 s C4 2:174
s b s ad2 s c1 ds c1 d
where:
p
I1 d2 c2 aac21
C3 2:175
k
and
I1 c2 c21 da
C4 2:176
k
Using partial fraction, the following relations can be obtained:
2 p p 3
s a2 s

6 ad2 b ad2 c2 =a s ad2
6 1 c21 ab c21 a2 d2 s c21 =a 77
g1 s C3 6 p 7
4 s 5
2
2

b ad b c1 =a s b
2:177
and
2 3
ad2 c1
6 2 7
6 ad b a2 d2 c21 s ad2 2c1 d bc1 ads c1 d 7
6
g2 s C4 6 7
c1 b2 7
4 5
2
2 2

2c1 d bc1 ads c1 d 2
b ad b c1 d s b
2:178
Consider the boundary condition at the surface in the stress field, where at
x0)r x 0, the constant in Eq. 2.167 becomes:
C2 g1 s g2 s 2:179
Then:
ps sx sx
x x; s g1 se
r ax g1 se c1 g2 sedx g2 se c1 2:180
Laplace Transform for each term in 2.180. To perform this, the following desig-
nations are introduced:
ps sx
Term1 g1 se ax Term2 g1 se c1 2:181
sx
Term3 g2 sedx Term4 g2 se c1
Consequently, the solution for stress distribution becomes the summation of the
inverse Laplace Transforms of the above terms. Therefore, the Laplace inversion
of Terms (Term1, Term2, Term3, Term4) are:
1 Term1
8 2 0
13 9
>
> dx
p x >
>
>
> 6 B e Erfc d at 2 at C7 > p >
>
> 1 6 2 p B C 7 >
>
>
> x2
6p e 4at d a e B d2
at C7 >>
>
> 6 B
C 7 >
>
>
>
2 2 2
2 ad b ad c1 =a 4 pt @ dx p x A5 >>
>
> p >
>
>
> e Erfc d at >
>
>
> 2 at >
>
>
> >
>
>
> 2 0 r

1 3 >
>
>
> c1 x t x >
>
>
> e Erfc c1
a p >
>
>
< 6 B a 2 at C 7 >
=
a 2 6 2 x 2 c 1
c2 t B C 7
C3 2 2 6p e 4at p e a B 1
C7
> 2
2 c1 ab c1 a d 4 pt 2 6 a B r

C 7 >
>
> @ c1 x t x A5 >
>
>
> e a c1 p >
>
>
> a at >
>
>
> 2 >
>
>
> 2 0 p
1 3 >
>
>
> p
x >
>
>
> xi a b >
>
>
> 6 B e Erfc i bt p C 7 >
>
>
> 1 6 2 p B 2 at C 7 >
>
>
> 6 p

x 2
bt B C 7 >
>
>
> 2 6 e 4at i b e B p
C 7 >
>
>
> 2 bad bc 2 =a 4 pt @ p
x A 5 >
>
>
> 1
e xi b
a Erfc i bt p >
>
: ;
2 at
2:182
and
8 q 9
> p 2 >
>
> p ad e
1 2 ad tx=c 1
Erf 2
ad t x=c1 >
>
>
> ptx=c1 >
>
>
> >
>
>
> 2 2 >
>
>
> ad c 2 =aad b >
>
>
> 0 q
1
q
1 >
>
>
> c2 >
>
>
< p
1 c 2 1
a1 e a tx=c1 Erf c 2
1
t x=c1 >
=
1
Term2 C3 a2 B B p tx=c a C
1
2 C
>
> @ A>>
>
> c1 ab c21 a2 d2 >
>
>
> >
>
>
> >
>
>
> p p
>
>
>
> p
1 i b e btx=c1
Erf i b t x=c >
>
>
> ptx=c1
1 >
>
>
> >
>
: 2 2 ;
b c1 =a ad b

x
1 t
c1
2:183
and
8 2
9
>
>
> a2 d2 ead t c1 ec1 dt >
>
>
< ad2 ba2 d2 c2 2c1 d bc1 ad
> >
=
1 dx 1
Term3 C4 e
>
> c1 ec1 dt b2 ebt >
>
>
> >
>
: 2c d bc ad 2
2 2d ;
2

1 1 b ad b c 1
2:184
and
2 2
3
a2 d2 ead tx=c1 c1 ec1 dtx=c1
6 2 7
6 ad b a2 d2 c21 2c1 d bc1 ad 7
1 Term4 C4 6
6
7
7
4 c1 ec1 dtx=c1 b2 ebtx=c1 5
2
2 2 d2

2c1 d bc1 ad b ad b c

1
x
1 t
c1
2:185
where 1t cx1 is a unit step function and Erf (y) is the error function of the
variable y. It should be noted that the value of a unit step function takes one or
zero, i.e.:

x x x x
t ) 1 t 0 and t [ ) 1 t 1 2:186
c1 c1 c1 c1

-1
where represents the inverse sign of Laplace Transformation.
c1
c1 2:188
ad
and
krx
rx 2 2:189
I1 da2 c
2 c1

U [1] is the unit step function, which is U1 t xc . The values of unit
1
step function are:

x x x x
t ) U1 t
0 and t
[ ) U1 t
1 2:190
c1 c1 c1 c1
Therefore, for the dimensionless stress distribution, the followings are resulted:
8 2 0
13 9
> x
p x >
>
> p
x 2 e t p >
>
>
> 1 6 t e 4t t B 2 t C7 >
>
>
> 6 B C 7 >
>
>
> 6 p B
C 7 >
>
>
> t
b 1 c 2 4 p 2 @ p x A 5 >
>
>
> 1 e x
erfc t p >
>
>
> >
>
>
> 2 t >
>
>
> 2 0
13 >
>
>
> p x >
>
>
> p
x 2 e c1 x
erfc c
t p >
>
< 6 t e 4t c1 t B
2 1
2 t C7 =
1 6 c 1 t e B C7
rx 1 2 2 6 p B
C7
>
> c1 1 c1 t b 4 p 2 @ p x A5 >
>
>
>

ec1 x erfc c1 t p >
>
>
> >
>
>
> 2 t >
>
>
> 2 0 p p

1 3 >
>
>
> b x >
>
>
> 2 p e ix t erfc i b p >
>
>
> t
3=2 6 e
x
i b
e b B
2 t C 7 >
>
>
> 6 4t
B C 7 >
>
>
> 6 p B p
C7 >
>
>
> b
t b c t 4 2 p 2 @ b p x A 5 >
>
>
: 1 e ix
t
erfc i b p >
;
2 t
2:191
8 2 3 9
>
> q >
>
>
> 1 6 t t
x
7 >
>
>
> 4pqt e =c 1 t x =c1 5 >
>
>
>
b 1 c 2
x =c
>
>
>
> t 1 p t >
>
>
>
1 >
>
>
> 2 3 >
>
>
> q
>
>
< 1 1 =
6 2
c1 t c1 x
2 7
rx 2 4pqc1 e Erf c1 t c1 x 5 U1
> c 1
> 2 2
c1 b =t p t x =c1 >
>
>
> 1 >
>
>
> 2 3>>
>
>
>
>
> p q
>>
>
>
> t 3=2 6 t p
b 1x
=c t

7>>
>
>
>
> 4pqi b e 1 erf i b 1x =c t
5 >
>
>
: t b b c 2 t
1 >
;
1 p t x =c1
2:192
8
9
>
> t et x t c1 ec1 t x >
>
> >
>
>
>
> 2 b 2 c1 t b c1 1 >

>
>
> 1 c 1 t >
>
>
> >
>
>
< c
t
x >
=
t c1 e 1
rx 3
2:193
>
> 2 c1 t b c1 1 >
>
>
> >
>
>
> 2 b x >
>
>
> b t e >
>
>
> >
>
>
: b t b 2 c1 t 2 >
;
8 9
t c1 ec1 t x

>
> t et x =c1 t c1 ec1 t x >
>
>
> 2 >
>
>
> 2 c1 t b c1 1 > >
< 1 c1 t b 2 c1 t b c1 1 =
rx 4 U1
>
>

b 2 t eb 1x =c1 t

>
>
>
> >
>
>
>
:
2 >
>
;
b t b 2 c1 t
2:194
Consequently, the dimensionless form of stress equation (Eq. 2.187) is:

rx rx 1 rx 2 rx 3 rx 4 2:195
Equations 2.145 and 2.195 are used to compute the dimensionless temperature
and stress distributions inside the substrate material.
2.3.2 Stress Free Boundary and Convection at the Surface
In practical applications, an inert assisting gas is used to protect the surface from
high temperature exothermic reactions. This intern results in surface cooling due to
convection effect of the assisting gas. Therefore, the convective boundary condi-
tion needs to be incorporated for laser gas assisted heating situations.
The Fourier heat transfer equation for a laser pulse decaying exponentially with
time can be written similar to Eq. 2.134. Therefore, the boundary conditions and
the closed form solution of the heat equation is given lightly.
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:
h i h
At x 0 ) T0; t T0 2:196
ox x0 k
and
At x 1 ) T1; t 0 2:197
Initially substrate material is assumed to be at uniform temperature. Therefore,
the initial condition is:
At t 0 ) Tx; 0 0 2:198
The Laplace transformation of Eq. 2.16 with respect to t, results in:
o2 T I1 d edx 1
sTx; s Tx; 0 2:199
ox2 k s b a
Introducing the initial condition and rearranging the Eq. 2.199, it yields:
o2 T 2 I1 d edx
g T 2:200
ox2 k s b
where g2 = s/a and s is the Laplace transform variable. Equation 2.203 has the
solution:
s A1 egx A2 egx I1 dedx

Tx; 2:201
ks b d2 g2
where A and B are constants. Introducing the boundary conditions will enable
calculation of the constants A1 and A2, i.e.:
A1 = 0, and:
I1 dh dk hT0
A2 2 2:202
2
ks b d g h kg s h kg
After substituting the values of A1 and A2 in Eq. 2.201, it yields:
s I1 dh dkegx hT0 egx I1 dedx

Tx; 2 2 2:203
ks b d g2 h kg sh kg k d g2 s b
which gives the solution for the temperature distribution in the Laplace domain.
The inverse Laplace Transform of Eq. 2.206 provides the temperature distri-
bution within the substrate material in space x and time t. The mathematical
arrangements of the Laplace inversion of Eq. 2.203 are given in the previous study
[35]. Therefore, the equation after the Laplace inversion is given below as:
8 p p 9
ead tdx Erfc 2px ead tdx Erfc 2px
2 2
>
> d at d at >
>
>
> at at >
>
>
> p p p p
>
>
>
> 2 h a 2 h a >
>
>
> 2 b ad k d a 2 b ad k d a >
>
>
> >
>
>
> p
p
p
p
>
>
>
< >
3=2
I1 da h kd bt
e e b=axi
Erfc 2p x
bt i e bt
e b=axi
Erfc p bt i =
x
Tx; t at 2 at
k 2 > p p p p >
>
> 2 b ad2 h k a bi 2 b ad2 h k a bi >
>
>
> >
>
>
> >
>
>
> p >
>
>
> p
hx=k h2 =k2 at p
x h dx ad 2
t bt >
>
>
> h a e e Erfc k at ke e e >
>
>
> h2 a h2 a
2 at
p
>
>
: k 2 2 ;
k2 b k2 ad ab ad h kd
" !
#
x x h p
T0 Erfc p ehx=k eh =k Erfc p
2 2
at
2 2 at k
2:204
solution for temperature distribution. The temperature distribution can be
expressed in a non-dimensional form by introducing dimensionless quantities and
substituting in Eq. 2.204. The dimensionless quantities are:
Tkd h
x xd: t ad2 t: T : h : b bt
I1 dk
The dimensionless temperature distribution for a full pulse is then:

82 p p p p 39
> x t
b t i >
> 6 e e Erfc 2pt t
x x
> x ex et Erfc 2p t eb e b x i Erfc 2p >
>
> t t 7>>
>
> 6 1
1
p 7>>
>
>
> 6 2 b 1 h 2 b 1 h
2b 1 h b i 7 >
>
>
> 6 7 >
>
>
>66 7 >
>
>
> p
p 7 >
>
>
> 6 b t b x i x h x h 2 t x
p
7
7 >
>
> 6
>6 e e Erfc 2 t b t i h e e
p Erfc 2 t h t
p
e x
e t
e b t >
>
>
> 7 >
>
>4 p 5>>
>
>
> b i b
2 2
1 b 1h 1 >
>
>
> 2 b 1 h 2 h h >
>
< =
T x ;t h 1
>
> >
>
> 2
> p p p 3 > >
>
> p >
>
>
> e x t
e Erfc x
p t e x t
e Erfc x
p t e c
e c x i
Erfc x
p c t i >
>
> 6
> 2 t 2 t 2 t 7 >
>
>
> 6 p
7 >
>
> 6
> 2 c 1h 1 2 c 1h 1 2 c 1h c i
7>>
>
> 6 7>>
>
> 7>>
> 6
> p
6 ec t e c x i Erfc p p p >
>
> 7>
2
x t i h x h t x
>
> 6 c h e e Erfc p h t e x
t
e e

c t 7>>
>
> 4 2 t
p
2 t
5>>
>
> >
>
: 2c 1h c i 2 h 2 c h 2 1 c 1h 1 ;
2:205
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. Therefore, Eq. 2.16 is used to formulate thermal stress distri-
bution in the substrate material.
In order to solve the momentum equation (Eq. 2.16) 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. 2.198 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 = 0 ) rx 0 2:206
and
At t = 1 ) rx 0 2:207
and
At x 0 ) rx 0 2:208
and
At x = 1 ) rx 0 2:209
x 1 h 2
o2 r :
i h
2
: i
s r
x x; s srx x; 0 r x; 0 c 2 s Tx; s sTx; 0 T x; 0
ox2 c21 x
2:210
where r s are the Laplace transforms of thermal stress and

x x; s and Tx;
temperature respectively in the x and s domains.
o2 r
x s2 s
r x x; s c2 s2 Tx; 2:211
ox2 c21
Considering the temperature distribution in a Laplace domain for an expo-
nentially decaying pulse with time, equation (Eq. 2.200), and substituting it into
Eq. 2.211, and solving for the stress field, yields:
"
o2 r
x s2 2 I1 dh dkegx
2rx x; s c2 s
ox 2 c1 ks b d2 g2 h kg
# 2:212
hT0 egx I1 dedx

sh kg k d2 g2 s b
Now let M1 and M2 be defined as:
I1 dh dkc2 s2 hT0 c2 s
M1 2 2:213
2
ks b d g h kg h kg
and
I1 dc2 s2
M2 2 2:214
k d g2 s b
Then, Eq. 2.212 becomes:
o2 r
x s2
r x x; s M1 egx M2 edx 2:215
ox2 c21
The complementary and the particular solutions of Eq. 2.215 are:
sx sx
r
x h A3 ec1 A4 e c1 2:216
While the particular solution has two parts, the first part is:
ps
x p1 G1 e ax
r 2:217

M1
G1 2 2:218
g2 cs2
1
x p2 G2 edx
r 2:219

M2
G2 22 2:220
d cs2
1
Hence, the general solution for the stress field becomes:

sx sx
x g A3 ec1 A4 e c1 G1 egx G2 edx
r 2:221
From the boundary condition (x = ? ) rx = 0), this yields A3 = 0.

Then, Eq. 2.21 reduces to:
sx
ps
rx g A4 e c1 G1 e ax G2 edx
2:222
Consider the boundary condition of the stress field at the surface, where at
x 0 ) orx 0, the constant in Eq. 2.222 becomes:
r
c1 s
A4 G1 s dG2 s
s a
Therefore, Eq. 2.222 becomes:
ps c1 sx
x x; s G1 se
r ax
G1 s p e c1
sa
2:223
c1 d sx
G2 sedx G2 s e c1
s
Finding the solution for rx in the x and t domain, one should take the inverse
Laplace Transform for each term in Eq. 2.223. To do this, the following terms are
introduced:
ps sx
Term1 G1 se ax Term2 G1 s pc 1
sa
e c1
sx 2:224
Term3 G2 sedx Term4 G2 s c1sd e c1
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 (Term1, Term2, Term3, Term4) can be stated as follows:
Term1 is composed of the terms:
Term1 Term11 Term 21 2:225
where
2 ps 3
ax
I1 dh dkc2 4 se
Term11
k 2 p 5
d s=a s b h k s=a 1=a s=c2 1
2:226
and
2 p 3
as x
e
Term21 hT0 c2 4 p 5 2:227
h k s=a 1=a s=c21
I1 dh dkC2
Let C10 2:228
k
and
C20 hT0 c2 2:229
Therefore, the Laplace transformation of Term1 can be written as:
1 Term1 1 Term11 1 Term21 2:230

where
1 Term11 1 Term111 1 Term211 1 Term311

1 Term411 1 Term511 2:231
1 1 1
Term611 Term711 Term811
The Laplace transformations of the terms are:
2 p p
3
h a h22athxk h at x
p
6 k e Erfc p
2 at 7
k
c21 h2 k3 a2 aC10 6 k 7
1 Term111
2 6 p p
7
2 2 2 2 2 2 2
2 a h c1 k h a k b h k d 2 4 h a h 2athxk
2 h at x 5
ek Erfc p
k k 2 at
2:232
and
2
3
p btpbx p x
p 6 b e a Erfc bt p
c2 ha3 bC10 6 2 at 7 7
1 Term211 2
1 2 6 p p p
7
2 b ad h2 a k2 b c1 ab 4 b x 5
be axbt Erfc bt p
2 at
2:233
and
p
c31 ha4 aC10
1 Term311
2 a2 h2 c21 k2 c21 ab d2 a2 c21
2 r
3
c1 c21 tc1 x t x
6 p e a a Erfc c1 p 7 2:234
6 a a 2 at 7
6
6 7
r
7
4 c1 1 1c2 t c x
t x 5
p e a a Erfc c1 p
a a 2 at
and
p
1 c21 ha2 adC10
Term411
2 b ad2 d2 a2 c21 h2 k2 d2
2
3
p ad2 tdx p x
6 a de Erfc d at p 2:235
6 2 at 7 7
6
7
4 pad2 tdx p x 5
a Erfc d at p
2 at
and
p
1 c21 h2 k3 a2 aC10
Term511
2 a2 h2 c21 k2 h2 a k2 b h2 k2 d2
2 2
p p
3
2 x h a h22athxk h at x
6 p
exp e Erfc p 7
k
6 pt 4at k k 2 at 7
6 p
p
7
4 h a h2 at hx h at x 5
2 k p
e k Erfc
k k 2 at
2:236
and
p
1 c21 ka2 abC10
Term611
2 c21 ab h2 a k2 b b ad2
2 2
q
pb
3
2 x p
x
6 p exp btx a Erfc bt p 7
6 pt 4at 2 at 7
66 p pb
7
7
4 btx a
p x 5
be Erfc bt p
2 at
2:237
and
p
c21 ka3 aC10
1 Term711
2 a2 h2 c21 k2 c21 ab d2 a2 c21
2 2
r
3
6 p exp p e a a Erfc c1 p 7
6 pt 4at a a 2 at 7
66 7
r
7
4 c1 1 1c2 t c x
t x 5
p e a a Erfc c1 p
a a 2 at
2:238
p
c21 ka2 ad2 C10
1 Term811
2 b ad2 d2 a2 c21 k2 d2 h2
2 2

3
2 x p ad2 tdx p x
6 p
exp
4at
ade Erfc d at p 7
6 pt 2 at 7
6
7
4 p 2 p x 5
adead tdx Erfc d at p
2 at
2:239
The Laplace inversion of Term21 can be written as:
1 Term21 1 Term121 1 Term221

2:240
1 Term321 1 Term421
Therefore, the Laplace inversions of composing terms are:
2 p p
3
h a h22athxk h at x
p e Erfc p 7
c2 hka aC20 6
k
6 k k 2 at 7
1 Term121 1 2 2 6 p
p
7
2h a h c21 k2 4 h a h2 athx h at x 5
ek 2 k Erfc p
k k 2 at
2:241
and
2 r
3
c1 c21 tc1 x t x
p p e a a Erfc c1 p
1 c1 ha2 aC20 6 6 a a 2 at 7 7
Term221 6 r
7
2 a2 h2 c21 k2 4 c1 c21 tc1 x t x 5
p e a a Erfc c1 p
a a 2 at
2:242
and
2 2
p p
3
2 x h a h22 hxk h x
p
6 p
exp e Erfc p
2 at 7
k
c2 ka aC20 6 pt 4 k k 7
1 Term321 12 2 6 p
p
7
2 a h c21 k2 4 h a h22athx h at x 5
e k k Erfc p
k k 2 at
2:243
and
2 2
r
3
p
6 p
exp p
e a a Erfc c p
a 2 at 7
1
c2 ka aC20 6 pt 4 a 7
1 Term421 12 2 6 r

7
2 a h c21 k2 4 c1 c21 tc1 x t x 5
p e a a Erfc c1 p
a a 2 at
2:244
The Laplace transform of Term2 can be written as:
1 Term2 1 Term12 1 Term22 2:245

Therefore, the Laplace inversions of composing terms are:
1 Term12 1 Term112 1 Term212 1 Term312

1 Term412 1 Term512 2:246
1 1 1
Term612 Term712 Term812
Knowing that:
I1 dh dkc2 c1
C30 p 2:247
ak
and
hT0 c2 c1
C40 p 2:248
a
Hence:
c21 hk6 a2 C30

1 Term112
k2 c21 k2 h2 a2 h2 a k2 b h2 k2 d2
2 1 3
p 2:249
6 pt x=c1 7 x

6 7
6 p 71 t
4 h a h2 at hpa p
5 c1
e k Erf
2
t x=c1
k k
and
c21 ha3 C30

1 Term212 2

b ad h2 a k2 b c21 ab
2 3
1
p
2:250
6 pt x=c1 7 x
46 7 1 t
p bt p 5 c1
be Erf bt x=c1
and
c2 ha5 C30
1 Term312 1 2
a2 h2 c21 k2 c1 ab d2 a2 c21
2 1 3
p 2:251
6 pt x=c1 7 x

6 7
6
7 1 t c
4 c1 c21 c1 p 5 1
t
p e Erf p t x=c1
a
a a
and
c21 ha2 C30

1 Term412
b ad d2 a2 c21 h2 k2 d2
2
2 3
1
p
2:252
6 pt x=c1 7 x
46 7 1 t
p ad2 t p p 5 c1
ade Erf ad t x=c1
and
c21 k5 a2 C30
1 Term512
a2 h2 c21 k2 h2 a k2 b h2 k2 d2

2:253
h2 a h22atx=c1 x
Dt x=c1 2 e k 1 t
k c1
and
p h i
1 c21 ka2 aC30 btx=c1
Term612 2 2
D t x=c 1 be
2 2
c1 ab
h a k b b ad
x
1 t
c1
2:254
and
p
c21 ka4 aC30
1 Term712
a2 h2 c21 k2 c21 ab d2 a2 c21

2:255
c21 c21 tx=c1 x
Dt x=c1 e a 1 t
a c1
and
p
1 c21 ka aC30
Term812
b ad2 d2 a2 c21 k2 d2 h2
h i
2:256
2 x
Dt x=c1 ad2 ead tx=c1 1 t
c1
The Laplace transformation of Term22 can be written as:
1 Term22 1 Term122 1 Term222

2:257
1 Term322 1 Term422
Therefore, the Laplace transformations of the composing terms are:

1 aC40 x
Term122 p 1 t 2:258
h pt x=c1 c1
and
c2 k2 aC40
1 Term222 21 2
h a h c21 k2
" p p r
#
1 h a h22at h a x x
p e Erf
k t 1 t
pt x=c1 k k c1 c1
2:259
and
ha3 C40
1 Term322 2 2
a h c21 k2
" r
#
1 c1 c21 t c1 x x
p p e a Erf p t 1 t
pt x=c a a c1 c1
2:260
and
p
k aC40 x
1 Term422 D t x=c 1 1 t 2:261
h2 c1
and
p
c2 k3 aC40 h 2 a h2 a x
1 Term522 12 Dt x=c1 2 e k2 tx=c1 1 t
h2 c 1 k 2 h2 a2 k c1
2:262
and
p
1 ka2 aC40 c21 c21 tx=c1 x

Term622 2 2 Dt x=c1 e a 1 t
c 1 k h2 a2 a c1
2:263
and
" #
I1 dc2 s2 edx
Term3 2 2:264
k d s=a s b d2 s2 =c21
and
8 9
>
> c21 c1 dt dx c31 aec1 dtdx c21 ab2 edx >
>
>
>
2 d2 b ad2 >
>
I dc < 2b c1 dc1 ad 2c1 d bc1 ad b 2
c 1
=
1 1 2
Term3
k > > 2 3
c a d e2 ad 2
tdx >
>
>
> >
>
: 2 21 2 2
;
a d c1 b ad
2:265
and
" #
sx
I 1 d2 c 2 c 1 se c1
Term4 2 2:266
k d s=a s b d2 s2 =c21
and
8 9
>
> c21 aec1 dtx=c1 c21 aec1 dtx=c1 >
>
>
> >
I d2
c c < 2db c1 dc1 ad 2dc1 d bc1 ad > = x

1 1 2 1
Term4 1 t
k >
> c21 abebtx=c1
2
c2 a2 ead tx=c1 >
> c1
>
> 2 1 >
>
: 2 2 2
2 2 2
2 ;
c1 d b b ad c1 a d b ad
2:267
where 1t cx1 is a unit step function, Erf (y) is the error function of the variable y
and Dt x=c1 is the Dirac delta function. The unit step function has the values
of 0 for t cx1 and 1 for t [ cx1 :
The closed form solution of the stress distribution can be written as:

where -1 represents the inverse sign of the Laplace transformation.
The additional dimensionless quantities are defined to present the stress dis-
tribution in the dimensionless form, i.e.:

c1 krx x
c1 : rx and U1 t
2:269
ad I1 da2 c2 c1
where U (1) is the dimensionless unit step function. Therefore, for the dimen-
sionless stress distribution, the followings are resulted:

rx 1 rx 11 rx 21 2:270
where

rx 11 rx 111 rx 211 rx 311 rx 411 rx 511 rx 611 rx 711 rx 811
2:271
In the dimensionless form:
2
3
h2 t h x
p x
h e Erfc h t p
h 1h2 c2 6 2 t 7
1 6 7
rx 111 6
7
2h2 1 h2 c2 2
1 h b =t
4 h2 t h x p x 5
h e Erfc h t p
2 t
2:272
and
2 p p
3
b bt x x
6 e Erfc b p

h 1h c2
1 b

6 2 t 7 7
rx 211
6 p
7
2b =t 1h2 b =t c2
1 b =t 4
b p x 5
eb t x Erfc b p
2 t
2:273
and
2
3
2 c2 p
x
1 t c1 x Erfc
c1
t p 7
h 1h c3 6 c1 e 2 t
1 6 7
rx 311 2 2
2
2
6
7
2 h c1 1 c1 c1 b =t 4 2 p
x 5
c2
1 e
c1 t c1 x
Erfc c1 t p
2 t
2:274
and
2
3
t x
p x

Erfc t p 7
h 1h c2 6e
1 6 2 t 7
rx 411 6
7
2h2 1 1 c2
1 1 b
=t 4
t x p
x 5
e Erfc
t p

2 t
2:275
and
h 1h2 c2
rx 511 1
2h2 1 h2 c2 2
1 h b =t
2 2

3
2 x h2 t h x
p x
p
exp h e Erfc h t p
6 pt 4t 2 t 7
6 7
6
7
4 2 p
x 5
h t h x
h e Erfc h t p
2 t
2:276
and

h 1c2
1 b
rx 611
2t b =t 1h2 b =t c2
1 b =t
2 2
3
2 x
p
6 pt
exp 7
6 4t 7
6 r p p
7 2:277
6
b b b x x 7
6 7
6 e t Erfc b p
7
6 t 2 t 7
6 7
6 r p p
7
4 b b b x x 5

e t Erfc b p
t 2 t
and
h 1c4
rx 711 2 1 2
2 h c1 1 c2
2
1 c1 b =t
2 2

3
2 x c2 t c1 x
p x

6 p
exp c1 e Erfc c1 t p 7
1
6 pt 4t 2 t 7
6
7
4 2 p
x 5
c1 ec1 t c1 x Erfc c1 t p
2 t
2:278
and
h 1c2
rx 811 1
2 1 h2 1 c2
1 1 b =t
2 2

3
2 x t x
p x

6 p
exp
4t
e Erfc t p 7 2:279
6 pt 2 t 7
6
7
4 p
x 5
et x Erfc t p
2 t
The dimensionless form of (r*x)21 is:

rx 21 rx 121 rx 221 rx 321 rx 421 2:280
where
2
3
h2 t h x
p x

Erfc h t p 7
T0 h c2 6h e
1 6 2 t 7
rx 121 2 6
7 2:281
2 h c2 1
4 2
h t h x
p
x 5
h e
Erfc h t p
2 t
and
2
3
c2 p x
c e 1 t c1 x Erfc c t p
T0 h2 c1 6 6
1 1
2 t 7
7
rx 221 2 6
7 2:282
2 h c2 4 c2 t c x p x 5
1
c1 e 1 1 Erfc c t p
1
2 t
and
T0 h c2
1
rx 321 2
2 h c2 1
2 2

3
2 x h2 t h x
p x

6 p
exp
4t
h e Erfc h t p 7
6 pt 2 t 7
6
7
4 2 p
x 5
h eh t h x Erfc h t p
2 t
2:283
and
2 2

3
2 x c2 p x
p
exp c e 1 t c1 x Erfc c t p
T0 c2 6 4t 1 1 7
1 h 6 pt 2 t 7
rx 421 2 6
7
2 h c2 4 2 p x 5
1
c1 ec1 t c1 x Erfc c1 t p
2 t
2:284
The dimensionless form of (r*x)2 is:

rx 2 rx 12 rx 22 2:285
where

rx 12 rx 112 rx 212 rx 312 rx 412
2:286
rx 512 rx 612 rx 712 rx 812
and
h 1h c3
rx 112 1
h2 1 h2 c2 2
1 h b =t
2 3
s! 2:287
1
x 7
6 h2 t
4q
h e Erf h t c 5 U1

p t x =c 1
1
and
h 1c3
1 h
rx 212

b 1h b =t c2
=t 2
1 b =t
2 3
r s!

6 1 b b x 7
4q
eb Erf b 5 U1
p t x =c t c1 t
1
2:288
and
h 1c3 1h

rx 312 2

h c2
1 1 c2
1 c2
1 b =t
2 3
s! 2:289
1
x 7
6 c2 t
4q
c1 e 1 Erf c1 t c 5 U1

p t x =c 1
1
and
h 1c3
1 h

rx 412
1 h2 1 c2
1 1 b =t
2 3
s! 2:290
6 1 x 7
4q
et Erf t 5 U1
p t x =c c1
1
and
h 1c3
rx 512 1
h2 1 h2 c2 2
1 h b =t
2:291

2 h2 x
2 h t c1
D t x =c1 h e U1
and
h 1c3
1
rx 612
b =t 1h2 b =t c2
1 b =t
2:292
b b bcx
D t x =c1 e 1 U1
t
and
h 1c3
rx 712 1 2
h2
c2
1 c2
1 1 c1 b =t 2:293
h i
c2
D t x =c1 c2 1 e
1 t c1 x U1
and

h 1c31

t xc
rx 812 D t x =c1 e 1 U1
1 h2 1 c2
1 1 b =t
2:294
The dimensionless for of (r*x)22 is:

rx 22 rx 122 rx 222 rx 322 rx 422 2:295
where
T0 c1
rx 122 q
U1 2:296
p t x =c1
and
2 3
s!
3
T c 6 1 h2 t

x 7
rx 222 2 0 1 2 4q

h e Erf h t c 5 U1
h c1 p t x =c 1
1
2:297
and
2 3
s!
T c h2 1
x 7
0 1 2 6
2
rx 322 4q
c1 ec1 t Erf c1 t 5 U1
h2 c1 p t x =c1 c1
2:298
and
T c
rx 422 0 1 D t x =c1 U1 2:299
h
and

T0 c3
1

2 h2 x
2 h t c1
rx 522 D t x =c 1 h e U1 2:300
h h2 c21
and
T c h h c2
i
rx 622 20 1 2 D t x =c1 c2
1 e
1 t c1 x U1 2:301
h c1
and

c3 t ec1 t x c3 t ec1 t x
rx 3 1 1
2 b c1 t c1 1 2 c1 t b c1 1
2:302
c2 t b2 eb x c2 t x
1 t e
1 2
b t b c2 1 t
2 b t 1 c2
1
and
2 3
c3
1 t ec1 t x c3
1 t ec1 t x

6 2 c1 t b c1 1 7
6 2 b c1 t c1 1 7
rx 4 6
6 b b x
t x

7 U1
7 2:303
2 c t c
4 c3
1 t b e 1 c3
1 t e 1 5
2 2
2

b t c 1 t b 2
b t c1 1
Consequently, the dimensionless form of the stress equation is:

rx rx 1 rx 2 rx 3 rx 4 2:304
The dimensionless temperature (Eq. 2.205) and stress distributions (Eq. 2.304)
are computed during the heating pulse.
2.3.3 Stress Boundary at the Surface
The Fourier heat transfer equation due to time exponentially decaying laser pulse
can be written similar to Eq. 2.134. Therefore, the boundary conditions and the
solution of the heat equation is given lightly.
In the analysis, no heat convection is considered from the free surface of the
substrate material. The depth well below the surface (x % ?), temperature
remains the same. Therefore, the corresponding boundary conditions are:
At the surface:

oT
0
At x 0 ) 2:305
ox x0
and
At depth infinity:
At x 1 ) T1; t 0 2:306
Initially, substrate material is considered at uniform temperature. Hence, the

initial condition is:
Initially:
At t = 0 ) Tx; 0 0 2:307
The Laplace Transformation of Eq. 2.16 with respect to t, results:
o2 T I1 d edx 1
sTx; s Tx; 0 2:308
ox2 k s b a
o2 T 2 I1 d edx
h T 2:309
ox2 k s b
where h2 = s/a and s is the transform variable. Equation 2.309 has the solution:
I1 dedx
Tx; s Aehx Behx 2:310
k s b d2 h2
where A and B are constants. Introducing the boundary conditions will allow
determining the constants A and B., i.e.:
" #
s I1 d d exphx expdx
Tx; 2:311
k s b h h 2 d 2 h2 d2
which gives the solution for temperature in Laplace domain.

The inverse Laplace Transform of Eq. 2.311 gives the temperature distribution
inside the substrate material in space x and time t domain as follows:

I1 d a
Tx; t
2k b ad2
8 2 r! 3 9
>
> b x p
>
>
>
> 6 exp ix Erfc p i bt 7 >
>
>
> r 6 a 2 at 7 >
>
>
> a 6 7 >
>
>
> id expbt 6 7 >
>
>
> 6 r ! 7

7 >
>
>
> b 6 p
>
>
>
< 4 exp ix b x 5 >
=
Erfc p i bt
a 2 at
>
> >
>
>
> 2

3>>
>
>
>
> x p p x >
>
> 6 expdx Erfc p d at exp Erfc d at p 7 > >
>
>
> 2 6 2 at 2 at 7 >
>
>
> exp ad t 4 5 >
>
>
> >
>
: ;
2 expbt dx
2:312
solution for temperature distribution. The temperature distribution in non-dimen-
sional form is possible by defining dimensionless quantities and substituting in
Eq. 2.312. The dimensionless quantities are:
Tkd b
x xd: t ad2 t: T :b 2 2:313
I1 ad

1 t
T x ; t
2 b t
8 2 r! 3 9
> p
>
>
> b x >
>
>
> 6 exp ix Erfc p i b 7 >
>
>
> r 6 t
2 t 7 >
>
>
> t 6 7 >
>
>
> i expb
6 7 >
>
>
> b 6 r !
7 >
>
>
> 6 b
x p 7 >
>
< 4 exp ix p
i b 5 =
t 2 t
>
> >
>
>
> 2

3>>
>
>
>
> x p
p x
>
>
>
> 6 expx Erfc p
t expx Erfc t p
7 >
>
>
> expt 6 2 t 2 t 7>
>
>
> 4 5 >
>
>
> >
>
: ;
2 expb x
2:314
be considered and Eq. 2.16 can be incorporated in the analysis.
To solve the stress equation (Eq. 2.16), two boundary conditions for the stress
fields should be defined. The first boundary condition is time exponentially
decaying stress at the surface due to recoil pressure developed. It should be noted
that the recoil pressure generated at surface decays almost exponentially with time
and it acts as a stress at the surface as soon as the recoil pressure is generated.
Moreover, as the distance below the surface increases further (extends almost to
infinity), the temperature gradient diminishes; in which case, thermal stress
approaches zero in this region. Therefore, the corresponding boundary conditions
are:
Stress at the surface due to recoil pressure:

At x 0 ) rx kx0 ro ebt eat 2:315
Stress free as depth approaches infinity:
At x 1 ) rx kx1 0 2:316
Initially, the substrate material is assumed to be free from thermal stress and as
the time approaches infinity, heating diminishes, the substrate material also
becomes free from thermal stress. In this case, initial and final conditions for the
stress field are:
Initially stress free substrate material:
At time t 0 ) rx kt0 0 2:317
Finally stress free substrate material:

At time t 1 ) rx kt1 0 2:318
In the case of temperature term in Eq. 2.16, initially substrate material is
assumed at a zero uniform temperature as similar to initial condition for Eq. 2.311.
As time approaches infinity, temperature becomes zero, since heating diminishes.
It should be noted that the laser pulse intensity decays exponentially with time.
Therefore, initial and final conditions for temperature term in Eq. 2.16 are:
Initially:
At t 0 ) Tx; 0 0 2:319
and
Finally:
At t 1 ) Tx; 1 0 2:320
o2 r
x 1 2 : h : i
2
s r
x x; s srx x; 0 r x x; 0 c 2 s Tx; s sTx; 0 T x; 0
ox2 c21
2:321
where r s are the Laplace Transforms of thermal stress and tem-
x x; s and Tx;
perature respectively in the x and s domains.
o2 r
x s2 s
r x x; s c2 s2 Tx; 2:322
ox2 c21
Considering the temperature distribution for time exponentially varying pulse,
Eq. 2.318, and substituting into Eq. 2.322, and solving for the stress field, yields:

o2 r
x s2 2 I1 d d exphx expdx
2rx x; s c2 s 2:323
ox2 c1 ks b hh2 d2 h2 d2
which can be arranged further:
o2 r
x s2 I1 d2 c2 s2
2rx x; s ehx
ox 2 c1 ks bh h2 d2
2:324
I1 dc2 s2
edx
ks b h2 d2
Now let N1 and N2 are defined as:
I1 d2 c2 s2
N1 2:325
ks bh h2 d2
and
I1 dc2 s2
N2 2:326
ks b h2 d2
Equation 2.16 becomes:
o2 r
x s2
r x x; s N1 ehx N2 edx 2:327
ox2 c21
The homogeneous and the particular solutions for 2.328 are:
sx sx

rx h C1 ec1 C2 e c1 2:328
while the particular solution has two parts, the first part is:
rx p1 Q1 ehx
2:329

N1
Q1 2 2:330
h2 cs2
1
rx p2 Q2 edx
2:331

N2
Q2 2 2 2:332
d cs2
1
therefore, the general solution for the stress field becomes:

sx sx
rx g C1 ec1 C2 e c1 Q1 ehx Q2 edx
2:333
Form the boundary condition (x = ? ) rx = 0), it yields C1 = 0:

Then,
sx N1 hx N2
rx g C2 e c1
2 e 2 s2 edx 2:334
h2 cs2 d c2
1 1
Substituting for N1 and N2 in Eq. 2.334 results:

sx
rx g C2 e c1 g1 sehx g2 sedx
2:335
where:
I1 d2 c2 s2
g1 s 2:336
ks bh h2 d2 h2 cs2
2
and
I1 dc2 s2
g2 s 2:337
ks b h2 d2 d2 cs2
2
Now substituting for h s=a and simplifying the expressions for g1(s) and
g2(s) yields:
p
I1 d2 c2 saac21
g1 s c2
2:338
ks bh s ad2 s a1
and
I1 dc2 s2 c21

g2 s 2:339
ks b s ad2 s2 c21 d2
Rearranging the two expressions as:
" p #
s
g1 s C3 2:340
s bh s ad2 s c21 =a
and
" #
s2
g2 s C4 2:341
s b s ad2 s c1 ds c1 d
where:
p
I1 d2 c2 aac21
C3 2:342
k
and
I1 c2 c21 da
C4 2:343
k
Using partial fraction, the following relations can be obtained:

2 p p 3
s a2 s
6 ad2 bad2 c2 =as ad2 c2 abc2 a2 d2 s c2 =a 7
6 1 1 1 1 7
g1 s C3 6
6 p 7
7
4 s 5
2
2

b ad b c1 =a s b
2:344
and
2 3
ad2 c1
6 2 7
6 ad b a2 d2 c21 s ad2 2c1 d bc1 ads c1 d 7
g2 s C4 6
6
7
7
4 c1 b2 5

2c1 d bc1 ads c1 d b ad2 b2 c21 d2 s b
2:345
Consider the boundary condition at the surface, where:
At x 0 ) rx ro ebt eat 2:346

Laplace transformation of the boundary condition gives:

1 1
s ro
r 2:347
sb sa
Then, the constant in Eq. 2.334 becomes:

1 1
C2 ro g1 s g2 s 2:348
sb sa
Then:
ps sx
x x; s g1 se
r ax g1 se c1
2:349
sx 1 1 sx
g2 sedx g2 se c1 ro e c1
sb sa
Laplace Transform for each term in Eq. 2.349. To perform this, the following
designations are introduced:
ps sx
Term1 g1 se ax Term2 g1 se c1
sx
Term3 g2 sedx Term4 g2 se c1 2:350

1 1 sx
Term5 ro e c1
sb sa
Consequently, the solution for stress distribution becomes the summation of the
inverse Laplace Transforms of the above terms.
The details of Laplace Transformation of the terms are given in Appendix 1.
Therefore, the Laplace inversions of Terms (Term1, Term2, Term3, Term4,
Term5) are:
8 2 0
13 9
>
> dx
p x >
>
>
> 6 B e Erfc d at p
2 at C 7>>
>
> 1 6 2 x2 p d2 at B C7 >>
>
>
2 2 2 6
> pe 4at d ae B C7 >>
>
>
> 6 B
C 7 >
>
2 ad b ad c1 =a 4 pt
>
> @ p x A 5 >
>
>
> dx
e Erfc d at p >
>
>
> 2 at >
>
>
> >
>
>
> 2 0 r

1 3 >
>
>
> >
>
>
> c1 x t x >
>
>
> 6 B e a Erfc c
1 p C 7 >
>
< a2 6 2 x 2 c c2 t B a 2 at C 7 =
1 6 1 1
B C 7
Term1C3 2 2 p
e 4at p e a
r

> 2 c ab c a 2 d2 6 pt a B
C7 >
>
> 1 1 4 @ c1 x t x A5 >>
>
> e a Erfc c1 p >
>
>
> a 2 at >
>
>
> >
>
>
>
> 2 0 p
13 >>
>
>
> b p
x >
>
>
>
> 6 B e xi a
Erfc i bt p C7 > >
>
>
> 1 6 2 x p B 2 at C7 >>
> C7 >
2
> 6pe 4at i be B bt >
>
>
> 2 2 =a 6 pt B pb
C7 >>
>
>
> 2 bad bc 1 4 @ p
x A 5 >
>
>
: e xi a
Erfc i bt p >
;
2 at
2:351
and
8 q 9
> p 2 >
>
> p
1 ad e 2 ad tx=c1
Erf 2
ad t x=c1 >
>
>
> p tx=c >
>
>
>
1
>
>
>
> 2 2 >
>
>
> ad c2 =a ad b >
>
>
>
1 >
>
>
> 0 q q
1 >
>
>
> c2
>
>
=
c12 c12
< p a e a
1 1 tx=c1
Erf t x=c 1 x
1
Term2 C3 a2 B B p tx=c a C 1 t
>
1
2 C
>
> @ c1 ab c21 a2 d2 A>>
>
c1
>
> >
>
>
> >
>
>
> >
>
>
> p btx=c1 p >
>
>
> p
1 i be Erf i bt x=c >
>
>
> p
1 >
>
>
>
tx=c 1
>
>
: 2 ;
b c1 =a
2:352
8 2
9
>
a2 d2 ead t
> c1 ec1 dt >
>
>
> >
>
2 < 2
ad ba d c1 2 2 2c d bc ad =
1 dx 1 1
Term3 C4 e
>
> c1 ec1 dt b2 ebt >
>
>
> >
>
: 2c d bc ad 2 2 2 2 ;
1 1 b ad b c1 d
2:353
and
2 2 3
a2 d2 ead tx=c1 c1 ec1 dtx=c1
6
6 ad2 b a2 d2 c21 2c1 d bc1 ad 7 7

1 Term4 C4 6 71 t x
6 7 c1
4 c1 ec1 dtx=c1 b2 ebtx=c1 5
2
2 2

2c1 d bc1 ad b ad b c21 d
2:354
and
!
1 b tcx a tcx x
Term5 ro e 1
e 1
1 t 2:355
c1

where 1 t cx1 is a unit step function and Erf (y) is the error function of the
variable y.
rx x; t 1 Term1 1 Term2 1 Term3 1 Term4

1 Term5 2:356
-1
where represents the inverse sign of Laplace Transformation.
c1
c1 2:357
ad
and
krx
rx 2 2:358
I1 da2 c
2 c1
and
kro
ro 2 2:359
I1 da2 c 2 c1

U [1] is the unit step function, which is U1 t cx
1
Therefore, for the dimensionless stress distribution, the followings are resulted:
8 2 0
13 9
> x
p x >
>
> p

6 x B e erfc t p
C7 >
>
>
> 6
2
B 2 t C7 >
>
>
> 1 6 t e 4t t B C 7 >
>
>
> 6 p B
C 7 >
>
>
> 2
4 p 2 @ p
A 5 >
>
>
> t b 1 c1 x
x >
>
>
> e erfc t p >
>
>
> 2 t >
>
>
> >
>
>
> 2 0
13 >
>
>
> p x >
>
>
> e c
x
erfc c
t p >
>
> 6p x 2
2 B
1
C7 >
< 2 t C7 =
1
1 6 t e 4t c c1 t B
t e

rx 1 2 6 p 1 B C7
6 B
C7
>
>
> c1 1
2
c1 t b 4 p 2 @ p x A5 >
>
>
>
> ec1 x erfc c1 t p >
>
>
> 2 t >
>
>
> >
>
>
> 2 0 p
1 3 >
>
>
> b p x >
>
>
> e ix
t erfc i b p >
>
>
> 6 x 2 p B C 7 >
>
>
> t
3=2 6 e 4t i b e b B 2 t
C 7 >
>
>
> 6
6 p B C 7 >
>
>
> B p
C 7 >
>
>
>
b t b c1 t 4
2 p 2 @ p x A 5 >
>
>
: e ix b
t erfc i

b p >
;
2 t
2:360
and
8 2 3 9
>
> q >
>
>
> 1 6 t 7 >
>
>
>
> 2 4p
qt et x =c1 Erf t x =c1 5 >
>
>
>
t b 1c1 p t x =c >
>
> >
>
>
>
1 >
>
>
> 2 3 >
>
>
> >
>
>
< q
>
=
1 6 1 c 2
t c
x
2 7
rx 2 2 2 4pqc1 e 1 1 c1 t c1 x 5 U1
>
>
c1 1
c1 b =t
p t x =c1 >
>
>
> >
>
>
> >
>
>
> 2 3>>
>
>
> p q
>>
>
>
>
3=2
t 6 t p
b
1x
=c t
7> >
>
>
>
> 4pqi b e 1 erf i b 1x
=c t
5 >
>
>
: 2
1 >
;
t b b c t 1 p t x =c 1
2:361
and
8 9
> t et x t c1 ec1 t x t c1 ec1 t x
>
>
> 2 >
>
>
> b 2 c t b
c 1 2 c t b c 1 > >
>
< 1 c t 1 1 1 1 >
=
1
rx 3
>
> b 2 t eb

x >
>
>
> >
>
>
>
:
2 >
>
;
b t b 2 c1 t
2:362
and
8 9
t et x =c1 t c1 ec1 t x

>
>

t c1 ec1 t x

>
>
> >
>
>
> 2

>
>
>
>
< 1 c t b 2 c 1 t b c 1 1 2 c 1 t b c 1 1 >
=
1
rx 4 U1
> >
b 2 t eb 1x =c1 t

>
> >
>
> >
>
>
>
>
>
>
: b t b 2 c t 2 ;
1
2:363
and
!
b t x
c a t x
c x
rx 5 ro e 1 e 1 1 t 2:364
c1
Consequently, the dimensionless form of stress equation (Eq. 2.356) becomes:

rx rx 1 rx 2 rx 3 rx 4 rx 5 2:365
Equations 2.314 and 2.365 are used to compute the dimensionless temperature
and stress distributions inside the substrate material.
2.4 Entropy Analysis Due to Thermal Stress Field
Entropy analysis provides information on thermodynamic irreversibility in the

thermal system. It also aids to quantify the loss work in the system because of the
heat transfer. Although the formulation of the entropy generation rate for laser
pulse heating is the same for different pulse types, the amount of entropy gener-
ation rate becomes different for different pulses. Consequently, the general for-
mulation of entropy generation rate is given below in line with the previous studies
[68].
Entropy generation in the solid substrate can be written as (2.10):

DS q
r: 2:366
Dt T
or
DS 1 q
r:q 2 rT 2:367
Dt T T
Consider the relation:
h hT; e 2:368
2.4 Entropy Analysis Due to Thermal Stress Field 65
or

oh oh
dh dT de 2:369
ot e oe T

oh
However, the term can be written as [11]:
oe T

oh oCk
q Ck T T T o . . . 2:370
oe T oT To
where Ck is the heat capacity and its value at reference temperature is assumed to
be zero. If T To is the almost the same order of To , then:

oCk
Ck T To 2:371
oT To
It is known that:

oh

Cp 2:372
oT e
Therefore:

oCk
qdh qCp dT To de 2:373
oT To
Furthermore, it was shown that [11]:

oCk 3E
aT 2:374
oT To 1 2m
where E is the Youngs modulus and aT is the thermal expansion coefficient.

Therefore, Eq. 2.369 can be rearranged as:
dh dT 3E de
q qCp aT To 2:375
dt dt 1 2m dt
It was also shown that [11]:
dh
r:q q 2:376
dt
Therefore Eq. 2.375 becomes:
dT 3E de
r:q qCp aT T o 2:377
dt 1 2m dt
oT
Inserting Eq. 2.377 into Eq. 2.367 and knowing that q k , it yields:
ox

DS 1 k dT 3E To de
rT 2 qCp aT 2:378
Dt T T dt 1 2m T dt
However, the strain can be written in terms of displacement (U), i.e.: e oU
ox .

DS 1 k dT 3E To d oU
rT 2 qCp aT 2:379
Dt T T dt 1 2m T dt ox
In the non-dimensional form:

DS 1 1 2 dT

3EaT To d oU
rT 2:380
Dt T T dt qCp 1 2m T dt ox
where
S
U Ud and S 2:381
qCp ad2
The first term on the right hand side of Eq. 2.380 represents the entropy gen-
eration due to temperature field while the second term is entropy generation due to
stress field.
Equation 2.381 is used to compute the entropy generation due to temperature
and stress fields.
The lost work can be evaluated using a Gouy-Stodola theorem [10]. In this case,
the lost work can be written as:
Z
Wlost To dS

2:382

where Wlost is the dimensionless lost work and To is the reference temperature.
Equation 2.383 is used to compute the lost work.
In order to formulate the displacement, the relation between the stress field and
the displacement needs to be established. However, the thermal strain along the
x-axis can be written as [11]:
1 m1 2m 1m
ex rx aT DT 2:383
1 mE 1m
where m, E, and aT are Poissons ratio, elastic module, and thermal expansion
coefficient of a solid material, respectively. The equation of motion in a solid
substrate in the one-dimensional form can be written as:
orx oV
q 2:384
ox ot
2.4 Entropy Analysis Due to Thermal Stress Field 67
where V is the velocity, which can be written in terms of displacement (V) as:
oU
V 2:385
ot
The equation of motion becomes:
orx o2 U
q 2 2:386
ox ot
Therefore, Eq. 2.386 gives the relation between the thermal stress and the
displacement, which is needed for the entropy calculations.
2.5 Findings and Discussions
The findings of the thermal stress developed in the laser irradiated region is
presented below under the appropriate sub-headings in line with the previous
studies [18].
2.5.1 Step Input Pulse Heating
2.5.1.1 Stress Free Boundary Condition at the Surface
A closed form solution for the stress field is obtained using the Laplace transform
method for a single and two successive step input pulses. However, the analytical
solution obtained is limited with the free stress condition at the surface, which is
the most common situation in laser machining applications. The stress field in the
solid phase of the substrate material is modeled although the liquid and gas phases
are developed during the laser high intensity irradiation. This is because of the fact
that the thermal strain generated in these phases is insignificant due to the free
molecular activity in these phases. The elastic stress field is considered because of
minimizing the complications associated with the non-linear equation of motion
when the stress level exceeds the elastic limit of the substrate material. The non-
dimensional closed form solution for the stress field can be applicable to any
material, provided that the laser pulse shape is in the step input form. In line with
the previous study [13], the two pulse modes are considered in the analysis,
which are a single step input pulse and two successive step input pulses as shown
in Figs. (2.1) and (2.2) while Table 2.1 gives the laser properties used in the
simulations.
Figures (2.3) and (2.4) show non-dimensional temperature profile inside the
substrate material for three different heating periods for single and two successive
step input pulses. The differences in temperature profiles appear in the surface
Fig. 2.1 Laser step input 1.2E+00

pulse used in the analysis and
1.0E+00
PULSE INTENSITY
in the simulations
8.0E-01
6.0E-01
4.0E-01
2.0E-01
Heating Period Cooling Period
0.0E+00
0.0E+00 5.0E-05 1.0E-04 1.5E-04 2.0E-04
TIME (t*)
region where, the magnitude of temperature is high and the temperature gradient is
small. The small temperature gradient in the surface region results from the energy
gain in the substrate material from the irradiated field. It should be noted that the
amount of the absorbed laser power varies exponentially with the depth inside the
substrate material (Beer Lamberts law), i.e. it is maximum at the surface and
decays exponentially with increasing depth. Consequently, the energy gain from
the irradiated field dominates the diffusional energy transport from the surface
vicinity to the solid bulk. It should be noted that the diffusional energy transport
enhances as the temperature gradient increases. Moreover, as the depth below the
surface increases, the temperature gradient becomes large due to a sharp change in
temperature in this region and smaller amount of absorbed energy from the irra-
diated field with increasing depth. Therefore, diffusional energy transport domi-
nates internal energy gain of the substrate material through absorption in this
region. Since the time domain selected is close to each other, the temperature
variation along the depth is almost similar, particularly at some depth below the
surface. It should be noted that the selection of a close time domain is due to
examination of stress wave propagation in the substrate material, i.e. a long time
domain obscures the appearance of the stress wave within the surface region, since
it has a considerably high speed. In the case of the heating situation for two
successive pulses, some variation in temperature profiles in the surface region
occurs at different heating durations. As the heating duration increases
(t* = 0.003), the influence of the second pulse on the temperature rise is more
pronounced in the surface region. This, in turn, elevates the temperature rise at the
surface.
Figure (2.5) shows non-dimensional stress inside the substrate material for three
heating periods while Fig. (2.6) shows non-dimensional stress distribution in the
region close to the surface. Thermal stress is compressive in the region of the
surface and becomes tensile with increasing depth from the surface. This situation
can be seen from Fig. (2.6). As the depth increases beyond the absorption depth
(x* = 1), it decays sharply. However, the location at which the thermal stress
reaches its first peak varies with the heating period. This is because of the stress
wave propagation inside the substrate material. The depth at which the wave
2.5 Findings and Discussions 69
1.2E+00
t* = t 1* t* = t1* + dt 1* t* = t 2* t* = t 2* + dt 2* t* = t 3*
1.0E+00
PULSE INTENSITY
8.0E-01
6.0E-01 C1 C2
4.0E-01
2.0E-01 First Pulse Cooling Period Second Pulse Cooling Period
0.0E+00
0.0E+00 5.0E-05 1.0E-04 1.5E-04 2.0E-04
TIME (t*)
Fig. 2.2 Two successive dimensionless laser step input pulses used in the analysis and in the
simulations (dt1* = dt2*)
Table 2.1 Laser pulse First pulse Second pulse

properties used in the
simulations Dimensionless pulse length 0.00006 0.00006
(sad2 , s is the pulse length)
Dimensionless intensity 1 1
Io
( , Io laser peak
kd
power intensity)
Dimensionless cooling period 0.00006 0.00006
between pulses (h1 and h2)
Fig. 2.3 Dimensionless 6.0E-05

temperature profiles inside t* = 0.001
the substrate material t* = 0.0015
4.5E-05
TEMPERATURE (T*)
obtained at three different t* = 0.002

dimensionless heating
periods for one step input
3.0E-05
pulse
1.5E-05
0.0E+00
0 0.4 0.8 1.2
DISTANCE (x*)
reaches its full amplitude depends on the wave speed and the duration of the wave
that travels. Consequently, longer the heating period results in the deeper the
location at which wave reaches its maximum amplitude.

temperature profiles inside
the substrate material
obtained at three different 9.0E-05
TEMPERATURE (T*)
periods for two successive
pulses 6.0E-05
t* = 0.001
3.0E-05
t* = 0.002
t* = 0.003
0.0E+00
0 0.3 0.6 0.9 1.2
DISTANCE (x*)
Fig. 2.5 Dimensionless 1

thermal stress inside the
substrate material for
different dimensionless 0.5
STRESS (*x)
heating periods for single

pulse
0
t* = 0.001
-0.5 t* = 0.0015
t* = 0.002
-1
0 0.4 0.8 1.2
DISTANCE (x*)

thermal stress in the surface
vicinity of the substrate
material for different 0.5
STRESS (*x)
periods for single pulse
0
t* = 0.001
-0.5
t* = 0.0015
t* = 0.002
-1
0 0.05 0.1
DISTANCE (x*)
2.5.1.2 Stress Continuity Boundary at the Surface
Figure (2.7) shows dimensionless stress levels inside the substrate material at
different heating periods. Stress field inside substrate material is compressive. This
is due to the stress boundary condition introduced at the sur-
face oT =ox jx 0 0 . The stress level is high in the surface vicinity and as the
distance increases it reduces sharply. Moreover, the stress level reduces to zero at
some distance below the surface. The depth, where the zero stress occurs,
increases with progressing time. This indicates that the stress wave propagates into
the substrate material with a wave speed c1.
Figure (2.8) shows the temporal variation of dimensionless stress field at dif-
ferent depths inside the substrate material. The stress level is considerably high at
the surface and as the depth increases, the stress level reduces. Moreover, the
magnitude of stress level inside the substrate material rises after certain period
from the pulse beginning. This indicates that the stress wave propagates into the
substrate material. However, the intensity and shape of the stress wave is modified
by the stress boundary condition at the surface, i.e. stress wave is compressive and
its magnitude increases with progressing time.
2.5.2 Time Exponentially Varying Laser Pulse Heating
2.5.2.1 Stress Free Boundary at the Surface
Figure (2.9) shows temporal variation of dimensionless temperature at different

locations inside the substrate material. Temperature rises rapidly in the early
heating period and the rate of temperature rise reduces as the heating progresses.
This occurs because of the energy gain and energy transport mechanisms inside the

stress distribution inside the
DIMENSIONLESS STRESS
-20
different heating periods
-40
t = 0.01
t = 0.04
-60 t = 0.08
t = 0.1
-80
-100
0 1 2 3 4 5
DISTANCE (x*)
Fig. 2.8 Temporal variation 0

of dimensionless stress at
different locations inside the
-20
substrate material
-40
x=0
-60 x=1
x=2
-80 x=3
-100
0 0.03 0.06 0.09 0.12
TIME (t*)
DIMENSIONLESS TEMPERATURE
Fig. 2.9 Temporal variation 0.0016

of dimensionless temperature
at different locations inside
0.0012
0.0008 x* = 0
x* = 0.5
x* = 1
0.0004
x* = 1.5
0
0.0E+00 6.0E-03 1.2E-02 1.8E-02 2.4E-02 3.0E-02
TIME (t*)
substrate material. In this case the internal energy gain by the substrate material
from the irradiated field dominates over the conduction losses in the early heating
period. As the heating period progresses internal energy gain and temperature
gradient in the surface region increase. Consequently, a stage is reached when the
conduction losses from the surface region to the solid bulk becomes important due
to high temperature gradient developed in this region with progressing time. Since
the pulse intensity decreases with time as shown in Fig. (2.10), temperature also
decreases with further progressing heating time. Therefore, temperature curve
decays after reaching its maximum value. The rise of temperature at some depth
below the surface is not as high as that corresponding to the surface. This occurs
because of the energy absorbed by the substrate material at some depth below the
surface, i.e. energy absorbed decreases exponentially with increasing depth
(Lamberts law). As the heating progresses further (t* [ 6 9 10-3) temperature in
the surface region (x* \ 1) reduces while at some depth below the surface
(x* C 1) increases gradually with heating period. This occurs because of the dif-
fusional heating of the substrate material at some depth below the surface due to
high temperature gradient in this region. In the region limited by the absorption
Fig. 2.10 Normalized power 1

intensity distribution
NORMALIZED POWER INTENSITY

0.8
0.6
0.4
0.2
0
0 0.01 0.02 0.03 0.04
TIME (t*)
depth (x* \ 1) energy gain by the substrate material reduces considerably with
further progressing time because of time exponentially decaying pulse intensity.
However, high temperature gradient enhances the energy loss from this region to
solid bulk, i.e. conduction losses dominates over the internal energy gain of the
substrate material. This, in turn, results in reducing temperature in this region with
further progressing time.
Figure (2.11) shows temporal variation of dimensionless stress at different
locations inside substrate material. The occurrence of peak stress at different
periods indicates the propagation of the wave inside the substrate material, i.e.
location x* = 0.1, the peak stress occurs at about t* = 0.3 9 10-2 while at
x* = 1.5 it occurs at about t* = 2.4 9 10-2. The magnitude of maximum stress is
high at x* = 0.5. This is because of the development of the maximum temperature
gradient at this location at t 1:25 102 . The stress level decreases with
increasing time and as the time progresses further it decays gradually.
2.5.2.2 Stress Free Boundary and Convection at the Surface
The closed form solution for the stress distribution due to an exponentially
decaying pulse with time and the convective boundary condition at the surface
have been derived. Steel was employed to simulate the temperature and stress
fields. The laser pulse parameter (b) and substrate material properties are given
Table 2.2. Figure (2.12) shows the laser pulse used in the simulations.
Figure (2.13) shows the dimensionless temperature distribution within the
substrate 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
108 W=m2 K . In this case, the temperature and its gradient in the surface region
are reduced. This can also be seen from Fig. (2.14), in which the temperature
Fig. 2.11 Temporal 6

x* = 0.1
variation of dimensionless
x* = 0.5
stress distribution at different
4 x* = 1
locations inside the substrate
x* = 1.5
material
2
-2
0.0E+00 1.0E-02 2.0E-02 3.0E-02 4.0E-02
TIME (t*)
Table 2.2 Material properties and laser pulse parameter used in the simulations
q d a aT m E Cp k b
(kg/m3) (1/m) (m2/s) (1/K) (Pa) (J/kgK) (W/mK) (1/s)
7930 6.16 9 106 3.7 9 10-5 1.6 9 10-5 0.29 2.10 9 1011 510 150 1.53 9 1011
gradient is shown. 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 substrate 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 irra-
diated 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
x 0:1 is because of: (i) the absorption process, i.e. the absorbed energy
decreases exponentially with increasing depth (Beer Lamberts 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 neighboring
region is low; therefore, the temperature profile is governed by the internal energy
gain in this region.
Figure (2.15) 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
x [ 0:06, 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
Fig. 2.12 Laser pulse used 1

in the simulations
NORMALIZED INTENSITY
0.75
0.5
0.25
0
0.0E+00 1.0E-02 2.0E-02 3.0E-02 4.0E-02
TIME

temperature distributions Time = 0.015
TEMPERATURE
within the substrate material 1.5E-02

h* = 2.02E-4
1.4E-02 h* = 2.02E-3
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
TEMPERATURE
1.8E-02 h* = 2.02E-3
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
TEMPERATURE
2.0E-02
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. 2.14 Dimensionless 0.0E+00

temperature gradient within h* = 2.02E-4
the substrate material -4.0E-03 h* = 2.02E-3
Time = 0.015
h* = 2.02E-2
dT/dx
-8.0E-03 h* = 2.02E-1
-1.2E-02
-1.6E-02
0.0 0.1 0.2 0.3 0.4 0.5
DISTANCE
0.0E+00
h* = 2.02E-4
-4.0E-03 h* = 2.02E-3
Time = 0.018
h* = 2.02E-2
dT/dx
-8.0E-03 h* = 2.02E-1
-1.2E-02
-1.6E-02
0.0 0.1 0.2 0.3 0.4 0.5
DISTANCE
0.0E+00
h* = 2.02E-4
-4.0E-03
h* = 2.02E-3
Time = 0.021
dT/dx
h* = 2.02E-2
-8.0E-03 h* = 2.02E-1
-1.2E-02
-1.6E-02
0.0 0.1 0.2 0.3 0.4 0.5
DISTANCE

h 0:0202 109 W/m2 K . 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 0:0202, increases with an increase in
time, provided that this increase is less than 10 %.
Figure (2.16) shows the change in thermal stress for different dimensionless
heat transfer coefficients with time. Since the substrate material is considered
initially to be stress free, i.e. the magnitude of the stress level is zero at time t = 0.
The stress level is tensile in the in the vicinity of the surface x 0:001. This is
because of the free expansion of the surface, in which case the strain becomes

h = 2.02E-4
stress distribution within the
1.0E-01 h = 2.02E-3
substrate material Time = 0.015
h = 2.02E-2
STRESS
0.0E+00 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
STRESS
0.0E+00 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
STRESS
0.0E+00
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
positive. However, the thermal stress becomes compressive as the depth increases
beyond x 0:01. This occurs as a result of the compression of the substrate
material. The effect of the heat transfer coefficient on the stress distribution
behavior with time is significant for a heat transfer coefficient of h 0:0202. In
this case, the temperature gradient in the surface region differs significantly from
those corresponding to the other heat transfer coefficients. This, in turn, generates a
compressive stress wave propagating into the material. The magnitude of the stress
wave is reduced as it propagates into the substrate material. Moreover, the stress
wave behavior disappears at a depth of 0.1, i.e. the stress behavior beyond this
depth changes significantly. This shows that the stress wave does not attain a depth
of x 0:1 during the time interval considered in the present study.
Fig. 2.16 Change in of 2.0E-01

dimensionless stress 0.0E+00
distribution with time
STRESS
-2.0E-01 h* = 2.02E-4
h* = 2.02E-3
-4.0E-01 x = 0.001
h* = 2.02E-2
-6.0E-01 h* = 2.02E-1
-8.0E-01
0.0E+00 5.0E-03 1.0E-02 1.5E-02 2.0E-02 2.5E-02
TIME
2.0E-01
0.0E+00
STRESS
-2.0E-01 h* = 2.02E-4
h* = 2.02E-3
-4.0E-01
x = 0.01 h* = 2.02E-2
-6.0E-01
h* = 2.02E-1
-8.0E-01
0.0E+00 5.0E-03 1.0E-02 1.5E-02 2.0E-02 2.5E-02
TIME
2.0E-01
0.0E+00
STRESS
-2.0E-01 h* = 2.02E-4
-4.0E-01 h* = 2.02E-3
x = 0.03 h* = 2.02E-2
-6.0E-01
h* = 2.02E-1
-8.0E-01
0.0E+00 5.0E-03 1.0E-02 1.5E-02 2.0E-02 2.5E-02
TIME
2.0E-01
0.0E+00
STRESS
-2.0E-01 h* = 2.02E-4
h* = 2.02E-3
-4.0E-01
h* = 2.02E-2 x = 0.1
-6.0E-01
h* = 2.02E-1
-8.0E-01
0.0E+00 5.0E-03 1.0E-02 1.5E-02 2.0E-02 2.5E-02
TIME
2.5.2.3 Stress Boundary at the Surface
Figure (2.17) shows dimensionless temperature distribution inside the substrate

material at different heating periods. The decay of temperature in the surface
region is slower as compared to some depth below the surface. This occurs
because of the insulated boundary condition at the surface as well as the internal
energy gain from the irradiated field in this region. In this case, the amount of
energy absorbed from the irradiated field in the surface region is higher than that

temperature distributions
inside the substrate material t* = 0.01
TEMPERATURE (T*)
for different time 1.2E-03 t* = 0.02
t* = 0.03
8.0E-04
4.0E-04
0.0E+00
0 1 2 3 4
DISTANCE (x*)
corresponding to some depth below the surface due to the Lamberts law of
absorption, i.e., energy absorbed by the substrate reduces exponentially with
distance away from the surface. Consequently, energy absorption enhances the
internal energy gain in the surface region, which dominates over the conduction
energy transfer from this region to the solid bulk. However, as the depth below the
surface increases further, amount of energy absorbed reduces and conduction
losses dominates the internal energy gain of the substrate material. Hence, the
temperature gradient changes considerably in this region.
Figure (2.18) shows stress levels inside the substrate material at different
periods. It is evident that the stress wave propagates into the substrate material
with a wave speed. The magnitude of peak stress reduces as the depth below the
surface increases towards the solid bulk. This indicates that material dumps
gradually the magnitude of stress level as the wave propagates towards the solid
bulk. Moreover, the time occurrence of peak stress inside the substrate material
differs for the stress free condition and stress condition at the substrate surface.
This is because of the temporal variation of the stress distribution at the surface
and the temporal variation of laser pulse intensity distribution, which differ, i.e.
pulse intensity decays with a progressing time while stress distribution at the
substrate increases to reach its peak, then decreases with increasing time as shown
in Fig. (2.18).
2.5.3 Entropy Analysis Due to Thermal Stress Field
2.5.3.1 Step Input Pulse
Figure (2.19) shows dimensionless entropy generation due to temperature field at

different heating periods. Entropy generation is low in the surface region and it
increases steadily with increasing depth from the surface. The increase in entropy

stress distribution inside the t* = 0.01
substrate material at different 10
t* = 0.02
times for stress free and stress
t* = 0.03
conditions at the surface 8
STRESS
6
No, Stress Sur.
4
-2
0 0.6 1.2 1.8 2.4 3
DISTANCE (x*)

entropy generation profiles Time = 0.4
due to temperature field Time = 0.8
inside the substrate material 3 Time = 1.2
at different heating periods
ENTROPY (S*)
Time = 1.6
0
0 0.5 1 1.5 2 2.5
DISTANCE (x*)
can be explained in terms of entropy equation. In this case, term rT 2 increases

in magnitude in the region next to the surface vicinity (Fig. (2.20)). However,
temperature is less in this region next to the surface vicinity. Consequently,
increase in entropy due to term rT 2 is enhanced by the term (T12 ), i.e. T12 rT 2
increases with depth. Moreover, in the region further away from the surface
rT 2 does not attain high values as much as in the surface region, but term (T12 )
increases considerably due to low temperature in this region. Similarly, the term
(T1 dTdt ) increases with increasing depth, since (T12 ) attains higher values as compared
to term (T1 dTdt ). Therefore, entropy generation due to temperature field increases
with depth. Moreover, entropy generation in the early heating period is high. In
this case, rapid rise of temperature in the surface region results in high rate of
entropy generation in the early heating period.

temperature gradient inside
the substrate material at
-0.06
different dimensionless
heating periods
Time = 0.4
-0.12
dT*/dx*
Time = 0.8
Time = 1.2
-0.18 Time = 1.6
-0.24
-0.3
0 1 2 3 4 5
DISTANCE (x*)
Figure (2.21) shows dimensionless entropy generation due to stress field at

different heating periods. Entropy generation in the surface region is high and as
the distance increases from the surface it reduces, particularly sharply for the long
heating period (t 1:6). Attainment of high entropy generation in the surface
region is because of high magnitude of surface displacement in this region.
Entropy contribution of displacement is presented by the term dtd oU ox in entropy
equation. High entropy generation in the surface region is high values of the term
d oU
dt ox in this region. When comparing the magnitude of entropy generation due
to stress field with its counterpart corresponding to temperature field, it is evident
that entropy generation due to temperature field is considerably high, i.e. the ratio
of entropy is in the order of 10-5. This is because of the small displacement of the
surface during the heating process.

entropy generation profiles Time = 0.4
due to stress field inside the Time = 0.8

Time = 1.2
substrate material at different
Time = 1.6
ENTROPY (S*)
heating periods 5.0E-06
0.0E+00
-5.0E-06
0 0.5 1 1.5 2 2.5
DISTANCE (x*)
2.5.3.2 Time Exponentially Decaying Pulse
Figure (2.22) shows the dimensionless volumetric entropy generation inside the
substrate material due to temperature field. Entropy generation is high in the
surface region and reduces sharply as the depth increases from the surface. The
sharp decay of the entropy generation in the region next to the surface vicinity is
because of the temperature gradient, which varies significantly in this region.
Moreover, entropy generation
due to temperature field is influenced considerably

by the energy diffusion Tk rT 2 as well as by the energy storage oT ot .
Consequently, entropy profiles do not exactly follow the square of temperature
gradient, i.e. the influence of energy storage on entropy generation is none zero. It
was shown that at some depth below the surface energy balance attains among the
energy absorbed from the irradiated field, internal energy gain of the substrate
material and the diffusional energy transport from the surface region to the solid
bulk. Consequently, entropy generation reduces to minimum in the region close to
the location of the equilibrium point, i.e. an equilibrium state is reached. As the
distance increases from the point of the equilibrium state, entropy generation
increases sharply. In this case, the diffusional energy transport dominates the
internal energy gain due to the energy absorbed from the irradiated field. Conse-
quently, the decay of temperature gradient changes in this region. As the depth
increases further towards to solid bulk, entropy generation attains a steady value,
i.e. a steady state is reached. As the heating period progresses, the magnitude of
entropy generation reduces; in which case, the laser power intensity reduces with
progressing time. Moreover, the location of minimum entropy generation moves
away from the surface as heating period progresses.
Figure (2.23) shows the dimensionless volumetric entropy generation inside the
substrate material due to stress field at different heating periods. Although the
entropy generation does not follow the stress distribution (Fig. 2.24), the location
of peak entropy generation coincides with the location of the peak stress inside the
substrate material. The negative sign of the entropy generation rate at some depth
below the surface is because of the negative magnitude of stress field (compressive
Fig. 2.22 Dimensionless 5.0

DIMENSIONLESS ENTROPY
entropy generation inside the Dimensionless Time

material due to temperature 4.0 t = 0.01
GENERATION
field at different heating t = 0.02

periods 3.0 t = 0.03
2.0
1.0
0.0
0.0 0.5 1.0 1.5 2.0
DIMENSIONLESS DISTANCE
Fig. 2.23 Dimensionless 0.6
DIMENSIONLESS ENTROPY
entropy generation inside the Dimensionless Time
material due to stress field at t = 0.01
0.3 t = 0.02
different heating periods
GENERATION
t = 0.03
0.0
-0.3
-0.6
0.0 0.5 1.0 1.5 2.0
Fig. 2.24 Dimensionless

Dimensionless Time
thermal stress inside the 4.0
t = 0.01
material at different heating t = 0.02
periods t = 0.03
2.0
0.0
-2.0
0.0 0.5 1.0 1.5 2.0
stress field), which is due to the propagation of the stress wave and, in all cases, the
entropy generation is positive. It should be noted that the negative sign in the
entropy generation curve is the indication of direction of the stress field and, in all
cases, the entropy generation is positive. The entropy generation inside the sub-
strate material appears like a cyclic with time. This is because of the stress
behavior, which occurs in the wave form. It should be noted that since the sub-
strate material is assumed to be stress free at initial state and elastic in nature, once
the thermal field diminishes it returns the initial state. Consequently, the final net
balance of entropy generation should be zero. This situation can be seen from Fig.
(2.23), i.e. the final net balance of entropy generation is zero.
Figure (2.25) shows the dimensionless total entropy generation inside the
substrate material, due to temperature and stress fields, at different heating periods.
Entropy profile in the surface region follows the entropy profile corresponding to
the temperature field. This indicates that entropy generation in this region is
dominated by the temperature field and the entropy generation due to stress field is
negligibly small. As the depth increases further towards the solid bulk, entropy
generation due to stress fields become important. In this case, the magnitude of
DIMENSIONLESS TOTAL ENTROPY

total entropy generation 5.0
Dimensionless Time
inside the material at different
4.0 t = 0.01
heating periods t = 0.02
GENERATION
3.0 t = 0.03
2.0
1.0
0.0
0.0 0.5 1.0 1.5 2.0
entropy generation due to temperature field becomes less than the peak values of
the entropy generation due to stress field. This, in turn, results in cyclic appearance
on the entropy curve in the region at some depth below the surface.
References
1. M. Kalyon, B.S. Yilbas, Analytical solution for thermal stresses during laser pulse heating
process. Proc. Inst. Mech. Eng. Part C, J. Mech. Eng. Sci. 215, 14291445 (2001)
2. B.S. Yilbas, M. Kalyon, Repetitive laser pulse heating with a convective boundary condition
at the surface. J Phys D Appl Phys 34, 222231 (2001)
3. B.S. Yilbas, N. Ageeli, Thermal stress development due to laser step input pulse intensity
heating. J Therm Stresses 29(8), 721751 (2006)
4. B.S. Yilbas, N. Ageeli, M. Kalyon, Laser induced thermal stresses in solids: exponentially
time decaying pulse case. Lasers Eng 14(1), 81101 (2004)
5. B.S. Yilbas, N. Al-Ageeli, Formulation of laser induced thermal stresses: stress boundary at
the surface. Proc. Inst. Mech. Eng. Part C: J. Mech. Eng. Sci. 217, 423434 (2003)
6. B.S. Yilbas, N. Ageeli, Thermal stresses due to exponentially decaying laser pulse and a
convection boundary at the surface. Lasers Eng 16, 235265 (2006)
7. B.S. Yilbas, M. Kalyon, Analytical approach for entropy generation during a laser pulse
heating process. AIChE J 52, 19411950 (2006)
8. B.S. Yilbas, Entropy analysis due to temperature and stress fields in the solid irradiated by a
time exponentially varying laser pulse. Heat Transfer Eng J 26(8), 8089 (2005)
9. A.D. Kovalenko, Thermoelaticity: basic theory and applications. (Wolters-Noordhoff
Publishing, Groningen, 1969), p. 190
10. A. Bejan, Entropy generation minimization (CRC press, New York, 1995)
11. D.Y. Tzou, Macro-to-microscale heat transfer. (Taylor and Francis, Washington, 1997),
pp. 231234
Chapter 3
Analytical Solution of Cattaneo
and Thermal Stress Equations
Abstract Laser short pulse heating of metallic surfaces initiates non-equilibrium

energy transport in the irradiated region. In this case, thermal separation of elec-
tron and lattice sub-systems takes place. The thermal communication of these sub-
systems occurs through the collisional process and the electrons transfer some of
their excess energy during this process. Although electron temperature attains
significantly high values due to the energy gain from the irradiated field through
absorption, lattice site temperature remains low. Since the heated region is limited
within a small volume, temperature gradients remain high across the irradiated
region despite the attainment of low temperature field. Consequently, high tem-
perature gradients cause the development of high thermal stress field in the small
region. This limits the practical applications of the laser treatment process at
microscopic scales. In this chapter, heat transfer at micro-scale is formulated and
temperature field is presented analytically. The closed for solutions for the
temperature and stress fields are obtained for various heating situations.
3.1 Introduction
Short-pulse laser heating of solid surfaces causes the heat wave generation in the
irradiated region and the energy transport to the solid bulk takes place through the
heat wave at a finite speed. In this case, heat is conducted in the solid due to the near
neighborhood excitation via changing of momentum and energy on a microscopic
scale in a wave form. The average communication time between the neighborhoods
is associated with the phonon relaxation time. The thermal communication takes
place in a dissipative nature resulting in the thermal resistance in the solid medium.
If the heating duration is longer than the relaxation time, the speed of the heat wave
propagation approaches to infinity and the heat wave equation reduces to the
classical Fourier heat equation. However, the Fourier heating model fails to predict
the temperature propagation speed and suffers anomalies when the heating duration

86 3 Analytical Solution of Cattaneo and Thermal Stress Equations
is comparable or less than the phonon relaxation time. Therefore, the heat equation
incorporating the wave nature of the energy transport in solids becomes unavoid-
able to formulate correct temperature filed in the heated region. Moreover, the
heating of sub-micron sized solid devices during the short heating period results in
non-equilibrium heating situation in the solid; in which case, the governing equa-
tion takes the hyperbolic form. The heating situation can be modeled using the
Cattaneos equation, since Cattaneos heating model is governed by the hyperbolic
temperature equations. Hence, Cattaneos heating model becomes appropriate to
account for the temperature propagation speed while eliminating the anomalies
arise from the Fourier heating model.
Thermal energy deposited into the substrate material causes internal energy
gain of the substrate material and thermal expansion because of the presence of
temperature gradients in the heated region. The work done during thermal
expansion of the substrate material can be associated with the energy loss in terms
of the internal energy gain. This is because of the fact that not all the deposited
energy is transformed into the internal energy gain of the substrate material, but
some fraction of the deposited energy is dissipated through the mechanical work
done during the thermal expansion. In reality, the temperature field developed in
the solid is responsible for the thermal expansion of the substrate. This, in turn,
generates the thermal stress field within the heated region. Since the heat equation
associated with the energy deposition, diffusion, and internal energy gain is in a
hyperbolic form for the short heating durations, investigation into the solution of
the hyperbolic heat and the thermal stress equations becomes necessary to asses
the thermal and mechanical responses of the substrate material. Although the
numerical solution is possible for such heating situation, the analytical solution of
Cattaneos heating equation is fruitful to generalize the temperature field in terms
of the heat source parameters and the material properties. The application of the
Laplace transformation method provides the exact solution to the Cattaneos
heating equation.
In the current chapter, the model study in relation to the short-pulse laser heating
of surfaces is presented and the analytical solutions for the heat equation, in the
form of Cattaneos equation, and thermal stress equation are provided for surface
and volumetric heat sources in line with the previous studies [15]. The influence of
laser pulse shape, namely step input and the exponential pulses, on temperature and
stress fields is also considered in the analytical treatment of the problem.
3.2 Surface Heat Source Consideration
In order to simplify the laser heating process, energy deposition from the irradiated
source can be considered as a surface heat source. The incident irradiated energy is
absorbed within the surface skin of the substrate material; however, consideration
of small absorption depth of the substrate material for the incident radiation can
simplify the necessity for the volumetric heat source consideration to the surface
3.2 Surface Heat Source Consideration 87
heat source energy deposition. This simplification minimizes the mathematical

complications associated with the solution of the temperature and stress fields.
Two pulses, including the step input and exponential pulses, are considered in the
below analysis.
The model presented considers the Cattaneo heating in line with the previous study
[1]. The Cattaneo equation can be written as:
oq
q KrT 3:1
ot
where q krT, e is the relaxation time, K is the conductivity tensor, q is the
heat flux vector, and T is the temperature. The Cattaneo equation yields to the
hyperbolic heat conduction equation for the temperature field, which is propaga-
q
tive with a speed c qCkp e, where q is the density and Cp is the specific heat
capacity.

o2 T oT o2 T
K 2 qCp s 2 3:2
ox ot ot
The heat equation can be nondimensionalized through introducing the fol-
lowing dimensionless parameters:
t x T j
t x p T p 3:3
2e 4ej=qCp P 4ej=qCp
where the P is the pulse intensity.
The resulting dimensionless equation is thus:
o2 T oT o2 T
2 2 3:4
ox2 ot ot
with dimensionless initial and boundary conditions:
T0 at t0 3:5
q f t at x0 3:6
The closed form solution of Eq. 3.4 can be possible by adapting the Laplace
transformation method. Taking the Laplace transformation of the governing
Eq. 3.4, initial and boundary conditions (Eqs. 3.5 and 3.6) results in:
b
o2 T b 2s T
s2 T b 0 3:7
ox2
b
oT
b
q b Fs
sT at x0 3:8
ox
The last equation can be rewritten as:
oT 1
Fss 1 at x 0; 3:9
ox s
The step input pulse intensity can be written as:
f t Put s ut 3:10
Since the model is linear, one can use the superposition principle to compute
the temperature distribution due to the two terms of the pulse function separately.
It is convenient to get the solution due to the second term of the f(t) first and then
use the mathematical formula:
L1 ess Fs ut sf t s 3:11

where L1 is the inverse Laplace Transform. Solution of the governing equation in
the s-domain (Laplace domain) can be represented as:
p p
2 2
Tx; s C1 ex s 2s C2 ex s 2s 3:12
For the solution to be bounded C2 must vanish and therefore,
p
2
Tx; s C1 ex s 2s 3:13
By applying the transformed boundary condition, the constant C1 can be
determined and the solution in the transformed domain can be finally written as:
p
2

1 2s ex s 2s
Tx; s p 3:14
ss 2
The first term of the solution can be inversely transformed as,
!
1 p hpi
1 x ss2
L p e et Io t2 x2 ut x 3:15
ss 2
where I0 is the modified Bessel function of the first kind and u is the unit step
function.
The second term can be inversed using the rule:
Z t
1
L1 Fs f k dk 3:16
s 0
Therefore, the final solution in the time domain can be expressed as:
hpi Z t hpi
Tx; t ut x et Io t2 x2 2 ek Io k2 x2 dk 3:17
0
be considered [6], i.e.:
o2 r 1 o2 r o2 T
2
2 2 c2 2 3:18
ox c1 ot ot
The preceding equation is rewritten using the following dimensionless quan-
tities along with the previously described ones (Eq. 3.3):
r
r eqCp 2c2 Pj
r c1 c1 c2 p 3:19
E j Cp Eq2 ej=Cp q
2
Therefore,
o2 r 1 o2 r o2 T
2
2 2 c2 2 3:20
ox c1 ot ot
By applying the Laplace transformation, the dimensionless stress field equation
in the s-domain is given by:
s2 r
^x p
^00 x
r M 2 e ss2x
3:21
c21
where M2 cp2 ss2

s2 2s
and c1 and c2 are the dimensionless wave speed and the
dimensionless thermal modulus, respectively. Therefore, the solution of stress
equation in the Laplace domain becomes:
p
2 ss2x
sx
sx c M 2 e
^ B1 ec1 B2 e c1 1 2
r 3:22
s c1 s 2 s
For boundness of the solution B1 0, and the condition of free stress (i.e.
r 0) on the surface is imposed as this stage to calculate the other coefficient B2
as:
c2 M2
B2 2 1 2 3:23
s c1 s 2c1 s
^x; s r
r ^h x; s r
^p x; s 3:24
where
sx
c21 c2 s 2e c1
^h p
r
s2 2s c21 s 2c21 s
p
c21 c2 s 2e ss2x
^p p 2
r 3:25
s2 2s c1 s 2 s
The stress component r
^h can be written as a multiplication of two subcom-
ponents as:
1 2
^h x; s r
r ^h x; s^
rh x; s 3:26
where
1 c2 c2 s 2 sx
r
^h 21 ec1
c1 s 2c21 s
and
2 1
r
^h p 3:27
2
s 2s
The inverse Laplace transform of the two subcomponents is carried out as the
following:

1 1
rh x; t L1 r^h

2c1 xc1 t
!
2 1 x 2 c2 1
x
c1 c2 2 d t 2 e 1 U t 3:28
c1 1 c1 c2 1 c1
1
and

2 2
rh x; t L1 r^h et J0 it 3:29
The Laplace inversion of the terms r

^h are:

1 1 1 1 c21 c2 s 2 csx
rh x; t L ^h L
r 2 e 1
c1 s 2 s
!
2 2
2c c 2 sx c c 2 sx
L1 2 1 e c1 2 1 e c1
c1 1 c21 s 2c21 s c1 1
! 2
1 2c21 c2 csx 1 c1 c2 csx
L e 1 L e 1
c21 1 c21 s 2c21 s c21 1

2c1 xc1 t
!
c21 c2 x 2c21 c2 c2 1
x
2 d t 2 e 1 U t
c1 1 c1 c21 1 c1

2c1 xc1 t
!
2 1 x 2 c2 1
x
c1 c2 2 d t 2 e 1 U t
c1 1 c1 c2 1 c1
1
and
!
1
2 1 2 1
rh x; t L r
^h L p et J0 it 3:30
ss 2
1 2
Having obtained rh x; t and rh x; t one can get rh x; t by employing the
convolution theorem:
Z t
1 2
rh x; t ^h s^
r rh t s ds 3:31
0
Similarly, the second component of the stresses r

^p can be written can be written
as a multiplication of two subcomponents:
r ^1
^p x; s r r2
p x; s^ p x; s 3:32
where
c21 c2 s 2
^1
rp
c21 s 2 s
and
p
e ss2x
^2
r p p 3:33
s2 2s
The inverse Laplace transform is carried out for both subcomponents:
0 2c2 t 1
21
c 1
B dt 2e 1 C
r1
p x; t L
1
^1
r p c21 c2 @ 2 2 A 3:34
c1 1 c2 1 1
p
r2
p x; t L 1
r
^ 2
p e t
I 0 t2 x2 Ut x 3:35
Having obtained rp1 x; t and rp2 x; t one can get rp x; t by employing the
convolution theorem:
Z t
rp x; t ^1
r p s^r2
p t s ds 3:36
0
The Laplace inversions of terms r ^p are:

2
c1 c2 s 2
r1
p x; t L1
r
^ 1
p L1
c21 s 2 s
!
1c21 c2 2c21 c2
L 2
c21 1 c1 1 c21 s 2c21 s
2 !
1 c1 c2 1 2c21 c2
L L
c21 1 c21 1 c21 s 2c21 s
2 2c2 t
!
c1 c2 2c21 c2 21
c 1
2 dt 2 e 1
c1 1 c21 1
and
p !
e ss2x hpi
r2
p x; t
1
L ^2
r p L 1
p et Ut xI0 t2 x2 3:37
ss 2
All components of the stresses have been transformed to (x, t) domain, and we
are ready to get the stresses as,
rx; t rh x; t rp x; t 3:38
Equations 3.17 and 3.38 can be used to compute the temperature and thermal
stress distributions inside the substrate material.
3.2.2 Exponential Pulse Heating
Laser heating pulse temporal behavior can be formulated through the time varying
exponential function. In this case, the formulation of the heat equation and the
relevant boundary conditions become identical to those presented in Sect. 2.2.1.
The analysis presented below is line with the previous study [2]. However, the
surface heat source can be formulated through the exponential pulse as shown in
Fig. 3.1 and it takes the form:
f t eb1 t eb2 t 3:39

the temperature distribution due to the two terms of the pulse function separately
and add them to get the final solution. In the sequel, one term namely, eb1 t is
considered as a pulse source and then the total solution is simply:
T Tjeb1 t Tjeb2 t 3:40
Fig. 3.1 Temporal variation

of the exponential pulse
The solution of the governing heat equation (Eq. 3.9) in the s-domain can be
represented as:
p p
b x; s C1 epa
x px
es2 s es2 s
T C2 e a 3:41
For the solution to be bounded C2 must vanish.
determined and the solution in the transformed domain can be finally written as:
p p
px ss1=e
ea
b x; s
T q e a=e 3:42
Ks b ss 1=e
The temperature distribution can be written as a multiplication of two functions

as:
b x; s T^1 x; sT^2 x; s
T 3:43
The inverse of the two terms can individually be written as:
p !

1 ^
1 1 px ss1=e
L T1 x; s L p e a=e
ss 1=e
s! " #
t
2e 1 x2 x
e Io 2
t U t p 3:44
2e a=e a=e
and
p p
ae ae bt
L T^2 x; s L1
1
e 3:45
Ks b K
where Io is the modified Bessel function of the first kind, U is the unit step
function.
The overall temperature distribution is obtained using the convolution:

Z t
Tx; t T^1 sT^2 t sds 3:46
0
The thermoelasticity is a temperature rate dependent by including temperature

rate among the constitutive variables [7]. However, most of the materials, the
strain rate is of the same order of the temperature rate; in which case, the thermo-
mechanical coupling becomes small for the small displacements and it can be
neglected [7, 8]. This simplifies the physical problem to yield the closed form
solution for the thermal stress field in the solid substrate. Consequently, after
assuming the homogenous isotropic structure, the temperature distribution of the
semi-infinite media in the s-domain can be represented in a dimensionless form:
p
s2 2sx
^ s s 2e
Tx; p 3:47
s b s2 2s
The corresponding dimensionless stress field equation in the s-domain is given
by:
s2 b
r x c2 s2 s 2 p
ss2x
b
r x p e 3:48
c21 s b s2 2s
c1 and c2 are the dimensionless wave speed and the dimensionless thermal
modulus, respectively.
p
2 2 ss2x
sx
sx c 2 s s 2 c 1 e
b
r B1 ec1 B2 e c1 p 3:49
s b s2 2s sc21 s 2 s
For boundness of the solution B2 0, and the condition of free stress (i.e.
b
r 0) on the surface is imposed as this stage to calculate the other coefficient B2
as:
c21 c2 s2 s 2
B2 p 3:50
sc21 s 2c21 s s b s2 2s
Therefore,
p
sx
c21 c2 ss 2e c1 c21 c2 ss 2e ss2s
^x; s p 2
r p 2
s2 2s c1 s 2c21 s s b s2 2s c1 s 2 s s b
3:51
where r^x; s r
^h s; x r
^p x; s. The stress component r ^h can be written as a
1 2
multiplication of two subcomponents as r ^h x; s r
^h s; x^
rh x; s. The inverse
Laplace transform of the two subcomponents is carried out as the following:

1 1
rh L1 r
^h
!

2c1 xc1 t
c21 c2 xb
tb 4c2 c2 1 x c2 c2 x
2
bb 2ec1 2 1 2 e 1 U t 21 d t
c1 b 2 b c1 1 c1 c1 1 c1
3:52

2 2
rh x; t L1 r
^h et J0 it 3:53
One can get rh x; t by employing the convolution theorem:

Z 1 2
t
rh x; t 0
^h s^
r rh t sds 3:54
Similarly, the second component of the stresses r

^p can be written can be written
as a multiplication of two subcomponents:
r ^1
^p x; s r r2
p x; s^ p x; s 3:55
The inverse Laplace transform is carried out for both subcomponents,
r1 1 1
p x; t L ^
rp
2c2
!
1 2b b2 bt 4c21 1
c2 1
c21 c2 d t e e 1
c21 1 c21 b 2c21 b c21 12 c21 b 2c21 b
3:56
p
r2
p x; t L1 2
^
r p e t
I 0 t 2 x2 U t x 3:57
^p x; t by employing the convolution theorem:

One can get r
Z t
rp x; t ^1
r r2
p s^ p t sds 3:58
0
All components of the stresses have been transformed to (x, t) domain, and we
are ready to get the stresses as,
The temperature and stress distributions inside the substrate material can be
computed from Eqs. (3.46) and (3.59), respectively.
3.3 Volumetric Source Consideration
The incident irradiated energy is absorbed in the skin of the substrate material
surface according with the Beer Lamberts law. In the case, a volumetric heat
source should be incorporated in the Cattaneo equation to account for the
absorption process. The solutions of the governing equations of heat conduction

and thermal stress are presented in line with the previous studies [3, 4].
In line with the previous study [4], the Cattaneo heat equation incorporating the
volumetric heat source can be written as:
o2 T oT o2 T
a F 0 f tgx e 3:60
ox2 ot ot2

k
where Fo f tgx is the source term and a is the thermal diffusivity a qCe .
Fo
Moreover, F0 Fo qCe is the normalized laser peak intensity Fo is the laser
peak intensity, Ce is the specific heat, f(t) is temporal function representing the
laser pulse intensity variation, g(x) is the absorption term, and e is the relaxation
time. The initial and boundary conditions are:
oT
T0 and 0 at t0 3:61
ot
oT
0 at x0 3:62
ox
The step input pulse intensity is considered as shown in Fig. 3.2, which can be
written as:
f t U x s U x 3:63

of laser step input pulse
3.3 Volumetric Source Consideration 97
where U x is the unit step function and s is the width of the step function. The
heating due to absorption of the laser beam is assumed to be exponential along the
depth of the substrate to resemble the actual heating process:
gx edx 3:64
the temperature distribution due to the two terms of the temporal part of the pulse
(i.e. f t) separately. It is convenient to get the solution due to the second term of
the first and then use the mathematical formula:
L1 ess Fs ut sf t s 3:65

Taking the Laplace transformation of the governing Eqs. 3.60 and the boundary
conditions (Eq. 3.62) results in:
o2 T^ F0
a sse 1T^ edx 3:66
ox2 s
and
oT^
0 at x0 3:67
ox
The solution of the governing equation in the s-domain can be represented as:
p p
^ x ps aes
2
p
x s2 es F0 exd
Tx; s C1 e C2 e a 3:68
ses2 s ad2
For the solution to be bounded, C2 must vanish and therefore,
p
^ s C1 e pa
x s2 es F0 exd
Tx; 3:69
ses2 s ad2
determined as:
p
F0 d a
C1 p
s ses 1es2 s ad2
Therefore, the temperature distribution in the s-domain can be obtained as:
p
xd
p x ps2es
F 0 e F 0 a de a
T^ 2
p 3:70
ses s ad s ses 1es s ad2
2 2
For convenience, the above expression is rewritten as:
T^ T^p T^h 3:71

where the homogeneous and particular solutions of the temperature are:
T^p F0 exd ^
hs 3:72
and
p
^
T^h F0 ad /s 3:73
where
^ 1
hs 3:74
s s e s ad2
2
and
^
/s ^ ^
hsws 3:75
where
p
x ps aes
2
^ e
ws p 3:76
sse 1
To invert the particular part T^p into the time domain, partial fraction and
standard inverse Laplace tables are employed to invert hs and for sake of brevity,
i.e.:
h tpc p p
p
1 p t c1 p t ct
ht p 2 e e c 12 ce 2e c 1 e 2e 3:77
2 c ad
Therefore, the particular solution of the temperature in the space-time domain
is:

p
F0 edx h tpc p
p
p t c1 p t ct
Tp x; t p 2 e e c 12 ce 2e c 1 e 2e 3:78
2 c ad
Similarly, the Laplace inversion of the function wt can be written as:
" s# " #
1 x 2 x
2et
wt e Io t2 U t p 3:79
2e a=e a=e
where I0 is the modified Bessel function of the first kind and U is the unit step
function
Having obtained the inversion of ^ ^
hs and ws, one can use the convolution
^
technique along with Eqs. 3.773.79 to calculate the inverse Laplace of /s as:
Zt
/t hswt sds 3:80
0
Therefore, the homogenous term of the temperature distribution is expressed as:
Th x; t F0 edx /t 3:81
By combining the homogeneous and the particular solutions of the overall
temperature distribution is obtained as:
T x; t Tp x; t Th x; t 3:82
The stress distribution inside the substrate material due to step volumetric pulse
is obtained by solving the equation that governs the momentum in an elastic linear
solid, which is given in a dimensionless form in the s-domain (Laplace domain) as:
s2 r
^x p
^00 x
r 2
M1 edx M2 e ss2x 3:83
c1
where
F0 c 2 s 2
M1 and
s b s2 2s d2
3:84
F0 c2 s2 d
M2 p
s b s2 2s d2 s2 2s
and C1 and C2 are the dimensionless wave speed and the dimensionless thermal
Solution of Eq. 3.83 can be written as:
sx c21 M1 dx c2 M p
csx
^ B1 ec1 B2 e
r 1 2
e 2 1 2 e ss2x 3:85
c21 d s2 s c1 s 2 s
For the boundness of the solution B1 0, and the condition of free stress (i.e.
^ 0) on the surface is imposed as this stage to calculate the other coefficient B2
r
as:
!
M 1 M 2
B2 c21 2 2 3:86
c1 d s2 s c21 s 2 s
Therefore, the stress can be written as:

r ^1 x; s r
^x; s r ^2 x; s 3:87
where
sx

c21 c2 F0 s2 e c1 edx
^1
r 3:88
s2 2s d2 s2 c21 d2
sx p

dc21 c2 F0 s e c1 e ss2x
^2 2
r p 3:89
c1 s 2 s ss 2 d2 ss 2
^1 and r
The inversions of r ^2 will be carried out next.
For convenience, the first part is divided into two parts,
^1 x; s r
r ^1a x; s r
^1b x; s 3:90
where
c21 c2 F0 s2 edx
^1a
r 3:91
s2 2s d2 s2 c21 d2
and
sx
c21 c2 F0 s2 e c1
^1b
r 3:92
s 2s d2 s2 c21 d2
2
Using standard partial fraction techniques and Laplace tables, the inverse
Laplace transform of the a-part can be evaluated as:
p p
c2 c21 F0 c21 d2 1 1 d2 1 1 p
t d2 11 xd
r1a x; t
p 2 e
2 d2 1 c21 1 d2 4c21
p p
c2 c21 F0 c21 d2 1 1 d2 1 1 p
t d2 11 xd
p 2 e
2 d2 1 c21 1 d2 4c21
c2 c21 F0 c2 c2 F0
ec1 tdxd 2 1 ec1 tdxd
2dc1 2 c1 d d 2d c1 1 d 2c1
3:93
csx
It is clear that r ^1a e where A exd and therefore, the Laplace s-
^1b A r 1
shift theorem is conveniently employed here which states that:
L1 eas Fs Ut af t a 3:94

Here again, U is the unit step function. Thus, the b-part can be inverted as:

xd sx sx
r1b x; t e U t r1a x; t 3:95
c1 c1
where r1a has already been calculated in Eq. 3.93.
Consequently, the first part of the total stress is explicitly determined as:
r1 x; t r1a x; t r1b x; t 3:96
Analogous to the first part of stress, this part of stress can be decomposed into:
^2 x; s r
r ^2a x; s r
^2b x; s 3:97
where
sx
dc2 c2 F0 e c1
^2a x; s 2
r 1 p 3:98
c1 s 2 s ss 2 d2 ss 2
and
p
dc21 c2 F0 e ss2x
^2b x; s 2
r p 3:99
c1 s 2 s ss 2 d2 ss 2
The convolution theorem is employed to get the inverse Laplace transform of
both parts. First, the a-part can be written as a multiplication of two functions:
^ a x; s N
^2a x; s R
r ^ a x; s 3:100
where
^ a x; s c21 c2 F0 d
R 3:101
c21 s 2 s ss 2 d2
sx
e c1
b a x; s p
N 3:102
ss 2
Therefore, the inversion of r^2a is obtained as:
Z t
1
r2a x; t L r^2a x; s Ra x; s Na x; t sds 3:103
0
where
2 2c2 t
c1 1 c2 c21 F0 d c2 1
1
1 ^
Ra x; t L Ra x; s e 1
2
c21 1 d2 4c21
p p
c2 c21 F0 d c21 d2 1 1 d2 1 1 p
t d2 11
p 2 e
2 d2 1 c21 1 d2 4c21
p p
c2 c21 F0 d c21 d2 1 1 d2 1 1 p
t d2 11
p 2
e
2 d2 1 c21 1 d2 4c21
3:104
and

x x cx t
Na x; t L 1 ^
Na x; s U t I0 t e1 3:105
c1 c1
Similarly, the b-part can be expressed as a product of two functions as:
^ b x; s N
^2b x; s R
r ^ b x; s 3:106
where
^ b x; s dc21 c2 F0
R 3:107
c21 s 2 s ss 2 d2
and
p
^ e ss2x
Nb x; s p 3:108
ss 2
As before, the inversion of r ^2b is obtained using the convolution theorem:
Z t
r2b x; t L1 r
^2b x; s Rb x; s Nb x; t sds 3:109
0
where
2 2c2 t
c1 1 c2 c21 F0 d c2 1
1
1 ^
Rb x; t L Rb x; s e 1
2
c2 1 d2 4c21
1p p
c2 c21 F0 d c21 d2 1 1 d2 1 1 p
t d2 11
p 2 e
2 d2 1 c21 1 d2 4c21
p p
c2 c21 F0 d c21 d2 1 1 d2 1 1 p
t d2 11
p 2
e
2 d2 1 c21 1 d2 4c21
3:110
and
hpi
Nb x; t L1 N^ b x; s et Ut xI0 t2 x2 3:111
Having obtained r2a x; t and r2b x; s using Eqs. 3.103 and Eq. 3.109
respectively, the second part of stress is completely determined in the time
domain:
r2 x; t r2a x; t r2b x; t 3:112
The total stress is computed by adding the two parts:

rx; t r1 x; t r2 x; t 3:113
where r1 and r2 have already been computed as in Eqs. 3.96 and 3.112,
respectively.
The temperature and thermal stresses fields can be obtained from Eqs. 3.82 and
3.113.
3.3.2 Exponential Pulse Heating
The temporal variation of the volumetric source term in the heat conduction
equation can be represented by an exponential function, which can be constructed
through subtraction of two exponential functions. The analysis presented below in
in line with the previous study [4]. However, Catteneo equation pertinent to the
conduction heating including the volumetric heat source is similar to Eq. 3.60, i.e.:
o2 T oT o2 T
a F 0 f tgx e 3:114
ox2 ot ot2
where F0 F0 =qCe with initial and boundary conditions:
oT
T0 and 0 at t0 3:115
ot
and
oT
0 at x0 3:116
ox
The laser pulses are not exactly in a step pulses, but having decaying intensity
tail. Similarly, the rise of the pulse intensity is not exactly step, but with increasing
rate. This can be described as an exponential pulse. In addition, introducing the
function f t eb1 t eb2 t provides a convenience in mathematical arrange-
ments for obtaining the closed form solution. The heating pulse takes the expo-
nential form:
f t eb1 t eb2 t 3:117

and
gx edx 3:118
the temperature distribution due to the first term of f t, i.e. eb1 t only, and then the
solution due to the second term is achieved by replacing b1 with b2 . In the sequel a
generic b will be used in the formulation to refer to either b1 or b2 .
Taking the Laplace transformation of the governing Eq. 3.114 and the boundary
conditions (Eqs. 3.115 and 3.116) results in:
o2 T^ F0 dx
a 2
sse 1T^ e 3:119
ox sb
and
oT^
q K
^ es^
q0 at x0 3:120
ox
The last equation can be rewritten as:
oT^
at x = 0: 3:121
ox
The solution of the governing equation in the s-domain can be represented as:
p p
^ x ps aes
2
p
x s2 es I0exd
Tx; s C1 e C2 e a 3:122
s b es2 s ad2
For the solution to be bounded, C2 must vanish and therefore:
p
^
px ss1=e I0exd
Tx; s C1 e a=e
3:123
s b es2 s ad2
determined as:
p
F0 d a
C1 p
s b ses 1 es2 s ad2
Therefore, the temperature distribution in the s-domain can be obtained as:
p
xd
p x ps2es
b I0e I0 a de a
T 2
p 3:124
2
s b s e s ad s b sse 1 s e s ad2
2
For convenience, the above expression is rewritten as:
T^ T^p T^h 3:125

where the homogeneous and particular solutions of the temperature are:
T^p I0 exd ^
hs 3:126
and
p
^
T^h I0 ad /s 3:127
where
^ 1
hs 3:128
s b s e s ad2
2
and
^
/s ^ ^
hsws 3:129
where
p
x ps aes
2
^ e
ws p 3:130
sse 1
To invert the particular part T^p into the time domain, partial fraction and
standard inverse Laplace tables are employed to invert hs and for sake of brevity,
tedious intermediate steps are glossed over here.
1 h tpc p
ht p 2 e e c 1 2be
2 c ad eb2 b
p
p 3:131
p t2be c1 p t ct
2 ce 2e c 1 2be e 2e
Therefore, the particular solution of the temperature in the space-time domain

is:
Fo edx h tpc p
Tx; t p 2 e e c 1 2be
2 c ad eb2 b
p
p 3:132
p t2be c1 p t ct
2 ce 2e c 1 2be e 2e
Similarly, the Laplace inversion of the function ws can be written as:

" s# " #
1 x 2 x
2et
ut e Io t2 U t p 3:133
2e a=e a=e
where Io is the modified Bessel function of the first kind and U is the unit step
function.. Having obtained the inversion of ^ ^ s, one can use the con-
hs and w
volution technique to calculate the inverse Laplace of /s as:
Zt
;t hsut sds 3:134
0
Therefore, the homogenous term of the temperature distribution is expressed as:
Th x; t Fo edx ;t 3:135
By combining the homogeneous and the particular solutions of the overall
temperature distribution is obtained as:
Tx; t Tp x; t Th x; t 3:136
The stress distribution inside the substrate material due to exponentially
decaying volumetric pulse is obtained by solving the equation that governs the
momentum in an elastic linear solid, which is given in a dimensionless form in the
s-domain as:
s2 r
^x p
^00
r 2
M1 edx M2 e ss2x 3:137
c1
where
F0 c 2 s 2
M1 and
s b s2 2s d2
3:138
F0 c2 s2 d
M2 p
s b s2 2s d2 s2 2s
and c1 and c2 are the dimensionless wave speed and the dimensionless thermal
Solution of Eq. 3.137 can be written as:
sx c21 M1 c21 M2 p
csx dx ss2x
^ B1 ec1 B2 e
r 1 2
e 2
e 3:139
c21 d s2 sc1 s 2 s
For the boundness of the solution B1 0, and the condition of free stress (i.e.
^ 0) on the surface is imposed as this stage to calculate the other coefficient B2
r
as:
" #
M 1 M 2
B2 c21 2 2 3:140
c1 d s2 sc21 s 2 s
Therefore, the stress can be written as:

^x; s r
r ^1 x; s r
^2 x; s 3:141
where
sx

c21 c2 F0 s2 e c1 edx
^1
r 3:142
s b s2 2s d2 s2 c21 d2
and
sx p

dc21 c2 F0 s e c1 e ss2x
^2
r p 3:143
s b c21 s 2 s ss 2 d2 ss 2
The inversions of r1 and r2 will be carried out next.
Utilizing t-shifting rule of Laplace transform, the partial fraction technique and
the standard Laplace tables, the inverse of stress component r ^1 can be written as:

x x
r1 x; t kx; t U t k x; t 3:144
c1 c1
where
8 9
>
> 2b2 etb 2et m sinhtv v d2 c21 2b d2 coshtv >>
>
> >
>
1 2 < b2 2b d2 b2 c2 d2 v d2 l2 4c21 d2 bg =
xd 1
k c1 c2 F0 e
2 >
> c1 ec1 td c1 ec1 td >
>
>
> >
>
: ;
c1 d bdl 2c1 c1 d b2c1 dl
2

et v coshtv 2c21 b d l m sinhtv
h21 c21 c2 F0 d
v d2 l2 4c21 d2 bg
2c2 t
1
!
2c21 le l betb
2 2
c1 g b d l2 4c21 c21 g b bg d2

x
t x x
w21 ec1 U t I0 t
c1 c1
h22 h21
hpi
w22 et Ut xI0 t2 x2
3:145
p
v d2 1
gb2
3:146
l c21 1

m d2 c21 b 1 b 1 2c21 b
The second part of the stress is now written as a summation of two sub-
functions as:
^2 x; s r
r ^21 x; s r
^22 x; s 3:147
where
sx
c21 c2 F0 d s e c1
^21 p 2
r 3:148
ss 2 c1 s 2 s s b ss 2 d2
and
p
c21 c2 F0 d s e ss2x
r
^22 p 2 3:149
ss 2 c1 s 2 s s b ss 2 d2
To invert r
^21 , the convolution theory is employed here as well:
Z t
r21 x; t h21 sw21 t sds 3:150
0
where h21 and w21 are given in Eq. 3.145

Similarly, the inversion of r ^21 is carried out using the convolution theory:
Z t
r22 x; t h22 sw22 t sds 3:151
0
where h22 and w22 are given in Eq. 3.145 Therefore, the inverse Laplace transform
of r2 is simply written as:
r2 x; t r21 x; t r22 x; t 3:152
Hence, the two components of the transformed stress have been inverted and as
a final step, one has to sum both terms to get the total stress distribution in space-
time domain:
rx; t r1 x; t r2 x; t 3:153
3.4 Entropy Analysis
Laser short-pulse heating of metallic surfaces triggers thermal separation of

electron and lattice sub-systems. The collisional process occurring between elec-
trons and lattice site governs the energy transfer in the irradiated region. Since
electrons gain energy from the irradiated field through the absorption process, their
excess energy increases rapidly and electrons attain higher temperature than that of
the lattice phonons. As the irradiated pulse ends, this process continues until the
temperature equilibrium between electrons and lattice phonons are achieved. Since
the energy transfer takes place in a short time, a hyperbolic heating model may
represent the heat transfer taking place in the irradiated material. Heat transfer due
to conduction and the work done during the formation of thermal stresses are
irreversible and cause entropy generation in the heated region. The entropy gen-
eration rate can be considered one of the measures of the energy dissipated due to
the thermodynamic irreversibility during the laser heating process. However,
3.4 Entropy Analysis 109
analytical formulation of the entropy generation rate provides useful information

for the thermodynamic irreversibility taking place in the temperature and stress
fields. In addition, it facilitates the parametric analysis for improved understanding
of thermodynamics of the laser irradiation process. The analytical formulation of
the entropy generation rate is given in line with the previous work [5].
The analytical solutions presented for the temperature and stress fields can be
used to predict the entropy generation rate during heating process. Yilbas [9]
formulated the entropy generation due to temperature and stress fields. Following
his approach, the entropy rate can be written as:

DS 1 dT k 2 3E T0 de
qCp rT aT 3:154
Dt T dt T 1 2m T dt

The strain de dt can be calculated from the stress field. For instance, for the
exponential laser pulses, the following equations can be incorporated in Eqs. 3.155
and 3.158 to compute the entropy generation rate. The temperature distribution can
be evaluated from:
Z
Tx; t t
0
T^1 sT^2 t sds 3:155
where
p !
1
1 1 px ss1=e
L T^1 x; s L p e a=e
ss 1=e
s! " #
t 1 x 2 x
e2e Io t2 U t p 3:156
2e a=e a=e
and

c
L1 T^2 x; s L1 cetb 3:157
sb
here I0 is the modified Bessel function of the first kind and U is the unit step
function.
The stress distribution can be obtained from:
where
Z t
1 2
rh x; t ^h s^
r rh t sds 3:159
0
and

1 1
rh x; t L1 r
^h
!
2c1 xc1 t
c21 c2 xb
tb 4c2 c2 1
x
2 bb 2ec1 2 1 2 e 1 U t
c1 b 2 b c1 1 c1

c21 c2 x
2 d t
c1 1 c1
3:160
and

2 2
rh x; t L1 r
^h et J0 it 3:161
and
2c2
!
1 2b b2 bt 4c21 21t
r1 1 1
rp c21 c2
p x; t L ^ 2
dt 2 2
e 2 2
e c1 1
c1 1 c1 b 2c1 b 2 2
c1 1 c1 b 2c1 b
3:162
and
p
r2 1 2
rp et I0
p x; t L ^ t 2 x2 U t x 3:163
The similar arguments can be used for the step input laser pulses for the entropy
calculations.
3.5 Findings and Discussion
The findings, form the solution of Cattaneos equation, are presented for two
different heat source considerations and two different laser pulse types. In addition,
thermodynamic irreversibility and entropy analysis due to temperature and stress
fields are given in details. The results obtained from the simulations of the ana-
lytical solutions are presented under the appropriate sub-headings in line with the
previous studies [15].
3.5.1 Surface Heat Source Consideration
The surface heat source consideration is divided into two categories including the
step input laser pulse and time exponentially decaying laser pule.
3.5 Findings and Discussion 111
Step Input Pulse Heating:

The results obtained from the analytical solutions of Cattaneo and the thermal
stress equations are presented below in line with the previous study [1].
Figure 3.3 shows temporal variation of dimensionless stress distribution at
different locations below the surface. Thermal stress developed appears to be in
wave form such that it decays first sharply and later the decay becomes gradual in
the amplitude. The stress wave propagates into the substrate material with a
constant speed and with almost constant peak amplitude, provided that the
amplitude decays incrementally due to the damping effect of the substrate material.
During the heating cycle, the wave generated is tensile with positive amplitude.
However, in the cooling cycle, it becomes compressive with negative amplitude.
The generation of the tensile wave in the heating cycle is attributed to the thermal
expansion of the substrate material in the surface region. This generates a positive
thermal displacement at the surface. Consequently, the stress wave generated in
the surface vicinity becomes tensile. Once the thermal stress wave is generated, it
propagates into the substrate material with the speed of sound, c, [19]. Conse-
quently, at depth below the surface and at the time it reaches any location, the
wave appears as repeating at this particular location. In the case of the cooling
cycle (t C 400), the sudden cooling of the surface results in contraction in the
surface region and generates a compression wave propagating into the substrate
material. Since the tensile wave is generated in an earlier stage than the com-
pressive wave, both waves do not meet at any location in the substrate material.
Consequently, canceling of the amplitudes of the tensile and the compression
waves in the substrate is less likely to occur. The compression wave behaves
similar to the tensile wave; in which case, the wave propagates at constant speed
and the wave amplitude decays slowly as it propagates into the substrate material
due to the damping effect of the substrate material.
Figure 3.4 shows dimensionless thermal stresses developed inside the substrate
material for different times in the heating period. The thermal stress developed
during the heating cycle appears as the stress wave propagating at a constant wave

stresses versus time at
different locations.
(x1 = 100; x2 = 200;
x3 = 300; x4 = 400)

stresses distribution inside the
different times. (t1 = 70;
t2 = 124; t3 = 179;
t4 = 233)
speed. The wave front is sharp and a tail is formed at the back of the wave. The tail
of the wave is due to the temperature gradient developed inside the substrate
material. In this case, the thermal strain developed inside the substrate material
during the heating cycle is responsible for the formation of the wave tail, which
decays gradually towards the surface. As the wave further propagates into the
substrate material the tail of the wave extends into the substrate material. More-
over, the amplitude of the tail of the wave reduces in the substrate material due to
the change in the temperature gradient along the depth below the surface.
Exponential Pulse Heating:
The findings of the closed form solutions of Cattaneo and thermal stress equations
are presented in line with the previous study [2].
Figure 3.5 shows temporal variation of dimensionless temperature at different
locations inside the substrate material. The peak temperature attains the maximum
at the surface and as the locations move at some distance below the surface, the
peak temperature reduces. Moreover, the location of the peak temperature below

at different locations inside
the surface changes in time due to the finite speed of temperature in the substrate
material. In general, the material response to the heating pulse at the surface is
slow, since the maximum peak pulse intensity occurs at dimensionless time t = 10
while the peak temperature occurs at around dimensionless time t & 25 at the
surface. This is associated with the pulse profile and the energy transfer taking
place in the surface region during the heating cycle of the short pulse. In this case,
the energy deposited to the substrate material during the early heating duration is
not sufficient to rise temperature at the same rate as the rate of pulse rise. As the
heating period progresses, the rate of temperature rise becomes high, particularly
at the surface. The internal energy gain in the surface vicinity from the heating
pulse is responsible for the high rate of temperature rise at the surface. However,
as the heating period progresses further, temperature decays from its peak value.
The rate of temperature decay is slower than that corresponding to the heating
pulse. This is particularly true at some depth below the surface. In this case, the
internal energy gain in the surface vicinity results in the attainment of high tem-
perature and diffusional energy transfer from the surface region to the solid bulk
takes place in a finite speed, which is less than the decay rate of the pulse intensity.
As the pulse intensity decays beyond 5 % of its peak intensity, temperature decay
becomes gradual. Consequently, the temporal gradient of temperature becomes
small; the Cattaneo equation tends to reduce the Fourier equation. Therefore,
diffusional energy transport with infinite speed governs the energy transfer in the
substrate material. This is more pronounced in the region below the surface.
Figure 3.6 shows temporal behavior of dimensionless thermal stress at different
locations inside the substrate material. The compressive stress waves are formed
first in the surface vicinity and they moved into the substrate material as the
heating period progresses. The compression due to initial heating results in the
formation of the compressive waves. However, the expansion of the initially
compressed surface results in tensile wave formation as the time progresses. This
appears as positive wave amplitude in the tail of the stress wave. This situation is
true for all the locations below the surface. The magnitude of the compressive
wave is larger than the amplitude of the tensile part of the wave. This indicates the

of dimensionless thermal
stress at different locations
inside the substrate material
initially formed wave propagates in compression form into the substrate material.
Moreover, the rate of rise of the amplitude of the compressive wave is higher than
that of the decay rate of the wave. This behavior is similar to temperature response
of the heated substrate material. Consequently, rapid heating and gradual cooling
of the substrate material in the surface region generates a compressive wave in a
similar fashion, provided that the rates of temperature rise and decay are not the
same as the rates of rise and decay of the thermal stresses.
3.5.2 Volumetric Heat Source Consideration
Step Input Pulse Heating

The step input pulse intensity resembling the laser pulse is incorporated in the
analysis and the absorption of the laser beam is modeled after considering the vol-
umetric heat source. The findings are presented in line with the previous study [3].
Figure 3.7 shows the temporal variation of temperature distribution at different
depths inside the substrate material. Temperature rises sharply at the surface
reaching its maximum prior to the step input pulse ends. However, the rise of
temperature is slightly less towards the end of the pulse as compared to its
counterpart corresponding to the mid-pulse-length. The high rate of rise at the mid-
pulse-length is associated with the internal energy gain of the substrate material
from the irradiated field. In addition, the rate of heat diffusion from surface region
towards the solid bulk is less in the early heating period. This enhances the rise of
temperature at the surface. Once the laser pulse ends, temperature first decays
sharply at surface and as the heating period progresses, temperature decay
becomes gradual. The wave nature of heat transfer becomes important during the
rapid decay of temperature at the initiation of the cooling period. This is because
2
of ootT2 , which becomes significantly large during this period. However, once

at different depths below the
surface
2
temperature decay becomes gradual the term ootT2 becomes small and the wave
nature of the heating replaces with the diffusional heating in the late cooling
period. The temporal behavior of temperature changes as the depth below the
surface increases. In this case, the rise and decay rates of temperature reduce,
which are more pronounced at dimensionless depths x3 and x4. Temporal behavior
of temperature at different depths reveals that the energy transfer from the irradiant
surface as well as the internal energy gain due to absorption do not result in the
rapid rise of temperature at some depth below the surface, which opposes to its
counterpart observed at the surface. This is attributed to the internal energy gain
and energy loss from this region towards the solid bulk through the wave motion

and diffusion. Since the temporal gradient of temperature oT ot becomes small at
some depth below the surface, the second time derivative of temperature also
becomes small. This, in turn, results in diffusional heat transfer from the surface
region to the solid bulk of the material with progressing time. Consequently, the
wave nature of heat transfer replaces with the diffusional heat transfer at some
depth below the surface during both heating and cooling periods.
Figure 3.8 shows temporal variation of thermal stress at different depths below
the surface. Thermal stress is compressive during the heating cycle irrespective of
the locations inside the substrate material. The stress waves compressive and they
have similar rise and fall times. The stress wave generated in the early heating
period has smaller amplitude as compared to those generated at different periods.
This is associated with the temperature gradient, which is low in the early heating
period because of the temperature rise. It should be noted that the temperature
gradient becomes high once the heating period progresses as observed from the
temperature distribution inside the substrate material (Fig. 3.3). Consequently the
amplitude of the stress wave becomes high as the heating cycle progresses towards
the pulse ending. Moreover, once the heating pulse ceases, stress wave becomes
tensile. Consequently the expansion of the free surface after the compression is
responsible for the formation of the tensile stress waves.

of dimensionless thermal
stress at different depths
below the surface
Exponential Pulse Heating

The closed form solutions of Cattaneo and thermal stress equations are presented
for time exponentially varying short pulse. The pulse is treated as a volumetric
source to incorporate the absorption of the incident laser irradiation. The findings
are presented below in line with the previous study [4].
Figure 3.9 shows temporal variation of dimensionless temperature at different
locations inside the substrate material. The peak temperature attains the maximum
at the surface and as the locations move at some distance below the surface, the
peak temperature reduces. Moreover, the location of the peak temperature below
the surface changes in time due to the finite speed of temperature in the substrate
material. In general, the material response to the heating pulse at the surface is
slow, since the maximum peak pulse intensity occurs at dimensionless time
t = 2.5 while the peak temperature occurs at around dimensionless time t 7:5 at
the surface. This is associated with the pulse profile and the energy transfer taking
place in the surface region during the heating cycle of the short pulse. In this case,
the energy deposited to the substrate material during the early heating duration is
not sufficient to rise temperature at the same rate as the rate of pulse rise. In
addition, the temperature gradient developed in the surface vicinity enhances the
heat diffusion from the surface to the solid bulk as the heating cycle progresses.
This contributes to temperature rise at the surface such that the delay in the peak
temperature is observed at the surface. Moreover, the rate of rise of temperature in
the early heating period is slow, which is more pronounced at locations some depth
below the surface. However, as the heating period progresses, the rate of tem-
perature rise becomes high, particularly at the surface. The internal energy gain in
the surface vicinity from the heating pulse is responsible for the high rate of
temperature rise at the surface. However, as the heating period progresses further,
temperature decays from its peak value. The rate of temperature decay is slower
than that corresponding to the heating pulse. This is particularly true at some depth
below the surface. In this case, the internal energy gain in the surface vicinity
results in the attainment of high temperature and diffusional energy transfer from

temperature distribution
versus time at different
depths,
x1 0; x2 1; x3 3,
and x4 5
the surface region to the solid bulk takes place in a finite speed, which is less than
the decay rate of the pulse intensity. As the pulse intensity decays beyond 5 % of
its peak intensity, temperature decay becomes gradual. Consequently, the temporal
gradient of temperature becomes small; the Cattaneo equation tends to reduce the
Fourier equation. Therefore, diffusional energy transport with infinite speed gov-
erns the energy transfer in the substrate material. This is more pronounced in the
region below the surface.
Figure 3.10 shows temporal behavior of dimensionless thermal stress at dif-
ferent locations inside the substrate material. The compressive stress waves are
formed first in the surface vicinity and they moved into the substrate material as
the heating period progresses. The compression due to initial heating results in the
formation of the compressive waves. However, the expansion of the initially
compressed surface results in tensile wave formation as the time progresses. This
appears as positive wave amplitude in the tail of the stress wave. This situation is
true for all the locations below the surface. The magnitude of the compressive
wave is larger than the amplitude of the tensile part of the wave. This indicates the
initially formed wave propagates in compression form into the substrate material.
Moreover, the rate of rise of the amplitude of the compressive wave is higher than
that of the decay rate of the wave. This behavior is similar to temperature response
of the heated substrate material. Consequently, rapid heating and gradual cooling
of the substrate material in the surface region generates a compressive wave in a
similar fashion, provided that the rates of temperature rise and decay are not the
same as the rates of rise and decay of the thermal stresses.
3.5.3 Entropy Generation Rate
Entropy generation rate due to temperature and stress fields are evaluated for time
exponentially varying laser short-pulse. The findings are presented below in line
with the previous findings.

thermal stresses distribution
with time at different depths,
x1 0; x2 1; x3 3,
and x4 5

entropy generation due to
heat transfer and stress
developed with depth at
different times
Figure 3.11 shows dimensionless entropy generation rate due to heat transfer
and thermal stress developed inside the substrate material for different heating
periods. Entropy generation rate is low in the surface region as compared to at
some depth below the surface. This is attributed to 1/T term in the entropy
equation, which becomes high when temperature reduces. Although temperature
gradient is high at some depth below the surface, its contribution to entropy
generation rate is lower than that of the term 1/T in the entropy equation. This
situation can also be observed when examining the entropy curves at different
heating periods. In this case, entropy generation rate reduces with progressing time
in the surface region despite the fact that the temperature gradient in the region
next to the surface vicinity increases with progressing time.
Figure 3.12 shows entropy generation rate due to thermal stress developed
inside the substrate material. Entropy generation rate in the surface region is
negligibly small and increases slightly as the depth below the surface increases.
This behavior is associated with the propagation of the thermal stress waves; in

entropy generation due to
stress developed inside the
material for different times
which case, the maximum magnitude of stress wave increases with progressing
time. When comparing the entropy generation rate due to heat transfer and thermal
stress developed, it is evident that entropy generation rate due to heat transfer is
significantly higher than that of thermal stress developed. This is because of the
energy dissipated by heat transfer, which is significantly higher than the work done
during the thermal expansion of the substrate material.
References
1. H. Al-Qahtani, B.S. Yilbas, The closed form solutions for Cattaneo and stress equations due to
step input pulse heating. Physica B 405(18), 38693874 (2010)
2. H. Al-Qahtani, B.S. Yilbas, Temperature and stress fields for short pulse heating of solids.
J. Thermophys. Heat Transf. 25(1), 173176 (2011)
3. H. Al-Qahtani, B.S. Yilbas, Closed form solution of Cattaneo equation including volumetric
source in relation to laser short-pulse heating. Can. J.Phys. 89(7), 761767 (2011)
4. H. Al-Qahtani, A closed form solution of temperature and stress fields for laser short-pulse
heating of a solid: exponentially decaying volumetric source. Lasers Eng. 22(12), 109124
(2011)
5. H. Al-Qahtani, B.S. Yilbas, Entropy generation rate during laser short-pulse heating:
contribution of heat transfer and thermal stress. Lasers Eng. (2012) (in print)
6. A.D. Kovalenko, Thermoelaticity: Basic Theory and Applications (Wolters-Noordhoff
Publishing, Groningen, 1969), p. 190
7. B.A. Boley, J.H. Weiner, Theory of Thermal Stresses (Robert E. Krieger Publishing Company,
Malabar, 1985)
8. T. Darabseha, M. Najia, M.A. Al-Nimr, Transient thermal stresses in an orthotropic cylinder
under the hyperbolic heat conduction model. Heat Transf. Eng. 29, 632642 (2008)
9. H. Al-Qahtani, B.S. Yilbas, Entropy generation rate during laser pulse heating: effect of laser
pulse parameters on entropy generation rate. Opt. Lasers Eng. 46, 2733 (2008)
Chapter 4
Analytical Treatment of Hyperbolic
Equations for Stress Analysis
Abstract Laser ultra-short pulse heating of metallic surfaces causes the hyperbolic
behavior of energy transport in the heated region. The consideration of the parabolic
nature of the non-equilibrium heating situation fails to formulate the correct heating
process. Although heating duration is ultra-short, material response to the heating
pulse is not limited to only heat transfer and the mechanical response of the heated
surface also becomes important. Consequently, mechanical response of the surface
under ultra-short thermal loading becomes critical in terms of the generation of the
high stress levels. In this chapter, hyperbolic behavior of heat transfer is introduced
in the laser heated region. The closed for solutions for the temperature and stress
fields are obtained for various heating situations. Two-dimensional effect of heating
on temperature rise is also considered for nano-scale applications.
4.1 Introduction
When the heating duration becomes less than the thermalization time of the substrate
material and the size of the irradiated region becomes comparable to the lattice
phonon mean free path, the equation governing the laser heating becomes hyperbolic
in nature. The analytical solution of hyperbolic heat equation can be achieved using
the Lie symmetry, or Laplace transformation technique, or perturbation method.
However, the formulation of the physical problem requires deep understanding of
electron and phonon behavior in the metallic substrates and microscopic modeling
of the process. When the short-pulse laser irradiation interacts with the metallic
surface, thermal separation of the electron and the lattice sub-systems takes place in
the irradiated region during the short heating period. The thermal communication
between electron and lattice sub-systems results in non-equilibrium energy transport
in the heated region. The collisional process taking place between excited electrons
and the lattice sub-system governs the energy transfer from the electron sub-system
to the lattice sub-system. This process continues until the thermal equilibrium is
established between the sub-systems. However, for the heating duration, which is

122 4 Analytical Treatment of Hyperbolic Equations for Stress Analysis
comparable to electron relaxation time, non-equilibrium energy transfer takes place

through the collisional process while dominating over the diffusional energy transfer
in the solid. In this case, the parabolic heating model fails to describe the physical
insight into heat transfer in the substrate material. Consequently, electron kinetic
theory approach incorporating the electron-lattice site collisions between the lattice
and electrons sub-systems at microscopic level becomes essential to account for the
formulation of the energy transport in the solids. Moreover, the closed form solution
for the governing equation of the physical problem becomes fruitful, since it pro-
vides the functional relation between the independent variables, such as time and
space, and the dependent variable, such as temperature. In order to obtain the ana-
lytical solution to the heat equation, some useful assumptions can be made such as;
electronphonon collision is elastic and the energy transport process based on the
collisions process can be reduced to one-dimensional form. In addition, if the
heating process is limited with the solid heating while excluding the phase change
process, elastic collisions between electrons and lattice site are justifiable. The size
of the absorption depth is considerably smaller than the irradiated spot size, which
enables to simply the problem into one-dimensional space. In this chapter, analytical
solution for the laser short-pulse heating of the metallic substrate is presented in line
with the previous studies [14]. The study covers the one-dimensional and two-
dimensional hyperbolic equations and the thermal stress development during the
short time period.
4.2 Formulation of Energy Transport in Metallic

Substrates at Microscopic Level
The formulation of the heat equation from the electron kinetic approach at micro-
scopic level is given below in the line with the previous study [1]. During the laser
heating of the metallic substrates, electrons within the absorption depth of the
substrate material gain energy from the irradiated field through the absorption
process. This, in turn, increases the electron energy and results in transferring of their
excess energy to lattice site through scattering, which depends upon the duration of
the interaction. In the case of short-pulses (slightly higher than the electronphonon
interaction time), electrons undergo few collisions with lattice site, since the elec-
tronphonon collision time is in the order of 0.02 ps. Electrons in the surface region
continuously gain energy from the irradiated field, which in turn results in energy
differences between the electrons and the lattice site in this region. The specific heat
capacity of electron is much smaller than its counterpart corresponding to the lattice
site; consequently, electron temperature increases rapidly while lattice site tem-
perature increase is gradual during the short heating duration. The temperature
differential between electron and lattice sub-systems results in non-equilibrium
energy transport in the substrate material. However, energy distribution of the
excited electrons may not be uniform in the surface region and also varies with time.
This causes temperature differential occurring in the electron sub-system.
4.2 Formulation of Energy Transport in Metallic Substrates at Microscopic Level 123
Fig. 4.1 Electron movement

in the surface region (x = 0
is the free surface)
In order to formulate the electron kinetic theory approach for the laser short-pulse
heating process, the investigation into the electron motion in the surface region of
the metallic substrates is necessary. The electron motion in the surface region of the
metals due to the irradiated field can be shown schematically in Fig. 4.1.
Electrons travel from surface to solid bulk as well as bulk to solid. Electrons
reaching the surface can escape once their energy exceeds the surface potential
barrier. Consequently, the number of electrons with X fraction reaching the surface
can manage to escape. In order to account for the reflected electrons from the surface,
a mirror image at the surface is considered [5]. Consider the location A in Fig. 4.1, the
net flow of Nsx ( N6 ) number of electrons towards the surface occurs and X fraction
of these electrons may escape from the surface. The situation, which occurs at
location A in Fig. 4.1 is an exact mirror image of that corresponding to point B, i.e.
(1 w) Nsx number electrons still flow towards the surface from location B, i.e.
(1 w) Nsx electrons flow to the right and N6 to the left at location B (Fig. 4.1). In the
case of location A approaches to location B, the number of electrons Nsx , which flow
from s to x changes discontinuously at s 0 and at s x. These changes can be
summarized as follows:
In the negative direction of x
N
1\s\x: Nxs
6
N
x\s\0: Nxs 1 w 4:1
6
N
0\s\1: Nxs
6
and
In the positive direction of x
N
1\s\0: Nsx
6
N
0\s\x: Nsx 1 w 4:2
6
N
x\s\1: Nsx
6
Moreover, at all locations:

N
Nsx Nxs 2 w 4:3
6
where Nxs is the number of electrons, which flow from x to s. It should be
noted that changes in Nsx follows a distribution, which can be described by a
rectangle function of unit height and width x centered on the positive s 2x.
Consequently, the rectangle function can be written as:
Y s x Y 2s x
2
4:4
x 2x
where
Y 2s x 1
0 for j2s xj [ 4:5
2x 2
and
Y 2s x
1 for j2s xj\ x 4:6
2x
Therefore, the electron distribution can be described as:
Y 2s x
N N
Nsx 1w with Nsx Nxs 1 w 4:7
6 2x 6
It should be noted that electron energy, which is characterized by temperature
Te s; t, is augmented from the initial Te s; t by an amount equal to that absorbed
in travelling from s to x. The total amount of energy, which is absorbed in an
element dn, area A in time dt is:
Io A f 0 n dt dn 4:8
where Io and f 0 n are the laser peak power intensity and the intensity distribution
function in the absorption depth of the solid material, since all the beam energy is
absorbed in the x-axis. The electron density can vary along the x-axis, in partic-
ular, the number of electrons travelling from ds to dx may not be the same as that
from dx to ds. Therefore, the portion of energy which is absorbed by electrons
which travel from ds to dx in dt is:
Nsx
Io A dt dn f 0 n 4:9
Nsx Nxs
where Nsx and Nxs are the number of electrons which travel from s to x and from x
to s, respectively. The total number of electrons which travel from ds to dx in this
time is:
Nsx AV dt 4:10
where V is electron mean velocity. Hence, the average energy absorbed by one
electron in dn in time dt is:
f 0 n
Io dn 4:11
Nsx Nxs V
and the total amount of energy absorbed by this electron from dx to ds is:
Z s
f 0 n
Io dn 4:12
x Nsx Nxs V
This expression gives the extra energy gain by the electrons in travelling from
ds to dx.
Electrons receive energy from the irradiated field and make collisions among
themselves as well as lattice site ions and they transfer some fraction of their
excess energy through the collisional process, i.e. electrons after the first collision
scatter and make further collisions with less energetic electrons and lattice site
ions. The energy exchange between energetic electrons, due to absorption of the
irradiated field, and other species can be formulated after considering the collision
probability of energetic electrons.
Consider the probability of electrons travelling a distance x without making a
collision is [6]:
x
exp 4:13
k
where x\2k and k is the mean free path of the electrons. Consider Fig. 4.1), the
probability of electrons, which make collision in B can be written as:
x
1 exp 4:14
k
or

dx
1 1 4:15
k
or
dx
4:16
k
provided that x\2k. The probability of electrons which last collided in B now
colliding in A is:

ds jx sj dx
exp 4:17
k k k
However, the number of electrons (Nsx ) leaving the location A in Fig. 4.1, area A in
time dt is Nsx AV dt where Nsx is the number density of electrons which transfer
energy from dx to ds, and V is the mean electron velocity. The number these
electrons which have just collided in location A is:

ds dx jx sj
Nsx AV dt exp 4:18
k k k
where dsk dxk expjxs j
k is the probability of electrons just collided in location A. If the
temperature of lattice site in dx is Tl x; t and the temperature of the electrons
when they arrive at dx is Te s; t (allowing for absorption on the way), then the
energy transfer to the lattice site in dx from collisions with electrons in which the
electrons give up a fraction f of their excess energy is:

ds dx j x sj
Nsx AV dt exp f Ee El 4:19
k k k
where Ee and El are the energy of electron and lattice ion, respectively. The
analysis related to f is given below:
The fraction of electron excess energy transfer during the time comparable or
slightly greater than the electronphonon collision time (sp ) can be written in
terms of the energy balance across the section dx in the substrate material, i.e.:
Electron energy)in (Electron energy)out Te in Te out
f or f ;
Electron energy)in (Phonon Energy) Te in Tl
where (Te ) in is the temperature of an electron entering the section, (Te ) out tem-
perature of the an electron leaving the section, and Tl is the phonon temperature. f
takes the values 0 f 1.
Integrating the contributions from all such infinitely small strips as to the
energy in location B (Fig. 4.1) gives:
Z 1
ds dx j x sj
Nsx AVdt exp f Ee El ds 4:20
1 k k k
In this case, energy transfer during Dt (Dt sp , where sp is the electronphonon
collision time) due to absorption of irradiated field and the collisional process can
be written after incorporating electron distribution function [7]:
Z 1 Y 2s x
DEtrans Vf kB j x sj
2
Nsx 1 w exp Te s; tds
AdxDt 1 k 2x k
Z 1 Y 2s x
Vf kB j x sj
2
Nsx 1 w exp Tl x; tds
1 k 2x k
Z 1 Z s
Io f Nsx jx sj
2 N N
exp f 0 ndnds
1 k sx xs k x
4:21
where
Q
Nsx 1 w 2sx
x
4:22
Nsx Nxs 2w
where f is the fraction of electron excess energy, which transfers to lattice site
during a single electron lattice site collision. The first term on the left hand side of
Eq. 4.21 is energy gain by the substrate material through the collisional process,
first and second terms on the right hand side represent electron and lattice energies,
and third term on the right hand site is the energy gain of the electrons due to the
irradiated field.
The final temperature of the electrons in dx after the collisional process can be
readily found from the conservation of energy, i.e.:
Total electron energy after collision = Total electron energy in during dt-
Change of lattice site energy
Total electron energy after collision is:
Z 1 Y 2s x
VkB j x sj
2
N sx 1 w exp Te s; t f Tl x; tds
1 k 2x k
4:23
Total electron energy carried into dx during dt is:
Z 1 Y 2s x
VkB 1 f j x sj
N sx 1 w exp Te s; tds
1 k2 2x k
Z 1 Q Z s 4:24
Io 1 f 1 w 2sx
2x j x sj
exp f 0 ndnds
1 k2 2w k x
Therefore, the conservation of energy yields:
Z 1 Y 2s x
VkB j x sj
2
N sx 1 w exp Te s; t f Tl x; tds
1 k 2x k
Z 1 Y 2s x
VkB 1 f jx sj
Nsx 1 w exp Te s; tds
1 k2 2x k
Z 1 Q 2sx Z s
Io 1 f 1 w 2x jx sj
2
exp f 0 ndnds
1 k 2 w k x
4:25
Equations 4.21 and 4.25 can be re-written after considering the electron distri-
butions for the lattice element dx apart and for electrons passing an area A, i.e.:
Z 1 Z 1
DEtrans fk jx sj fk jx sj
3
exp Te s; tds 3
exp Tl x; tds
AdxDt 1 k k 1 k k
Z j xj Z j xj
fk jx s j fk jx sj
w 3 exp Tl x; tds w 3 exp Te s; tds
0 k k 0 k k
Z 1 Z s
Io f 1 jx sj
2 2 w
exp f 0 ndnds
1 k k x
Z 1 Z s
Io f w jx sj
2
exp f 0 ndnds
1 k 2 w k x
4:26
and
Z 1
k j x sj
2
exp Te s; t f Tl x; tds
1 k k
Z j xj
wk j x sj
exp Te s; t f Tl x; tds
0 k2 k
Z 1 Z j xj
k1 f j x sj k1 f j x sj
exp Te s; tds exp Te s; tds
1 k2 k 0 k2 k
Z 1 Z s
Io 1 f j x sj
2
exp f 0 ndnds
1 k 2 w k x
Z j xj Z s
Io 1 f j x sj
exp f 0 ndnds
0 k2 2 w k x
4:27
where k is the thermal conductivity, which makes use of the simple kinetic theory
result for the electron thermal conductivity [8]:
NVkB k
k : 4:28
3
The energy content of the small lattice site element dx apart can be written as:
DE A dx Cl Tl x; t 4:29
where Cl q Cp. The energy gain of the small lattice element during the small
time interval dt is DE DE
dt . The expansion of dt yields:
" #
DE 1 0 dt2 00
Et dt E t E t Et 4:30
dt dt 2!
or
DE dt 00
E0 t E t . . . 4:31
dt 2!
Combining Eqs. 4.29 and 4.31 yields:

DE o dt o2
A dx Cl Tl x; t Cl Tl x; t 4:32
dt ot 2! ot2
When the time increment approaches to thermal relaxation time (dt ! ss ),
Eq. 4.32 reduces to:
oe DE o o
Cl Tl x; t ss Tl x; t 4:33
ot dt A dx ot ot
where e is the volumetric energy content of lattice site. The energy gain of the
small lattice element through collisional energy transport can also be written as:

DE 1 DEtrans o DEtrans
ss 4:34
A dx Dt A dx Dt ot Dt
Substituting Eq. 4.26 into Eq. 4.34 yields the change of lattice site energy
which is:
0Z 1 Z 1 1
fk jx sj fk jx sj
3 exp Te s; tds 3 exp Tl x; tds
B 1 k k 1 k k C
B C
B Z Z C
B j xj
jx sj
j xj
jx sj C
B
fk
w 3 exp Tl x; tds
fk
w 3 exp Te s; tds C
DE B k k k k C
B 0 0 C
B Z 1 Z s C
AdxDt B jx sj
C
B
Io f 1
exp 0
f ndnds C
B 2 2 w k C
B 1 k x C
B Z 1 s Z C
@ Io f w jx sj
A
0
2 2 w
exp f ndnds
1 k k x
0Z 1 Z 1 1
fk jx sj fk jx sj
exp T e s; tds exp T l x; tds
B 1 k3 k 1 k
3 k C
B C
B Z Z C
B j xj
jx sj
jxj
jx sj C
B w
fk
exp T x; tds w
fk
exp T s; tds C
oB C
3 l 3 e
B 0 k k 0 k k C
ss B Z 1 Z s C
B j j
C
B I o f 1
exp
x s
f 0
ndnds C
B 2 2 w k C
B 1 k x C
B Z s Z C
@ 1
Io f w jx sj
A
0
2
exp f ndnds
1 k 2 w k x
4:35
Equations 4.27 and 4.35 are the energy transport equations of interest for laser
short pulse heating process. However, for small rise of electron temperature during
the low intensity laser heating pulse prevents electron escape from the surface.
Consequently, the term w in Eqs. 4.27 and 4.35 becomes zero.
Equations 4.27 and 4.35 can be transformed into differential equations. The
method of solution to be used in the following analysis is the transformation of the
simultaneous differentialintegral Eqs. 4.27 and 4.35 using the Fourier integral
transformation, with respect to x [7]. This is due to the fact that the resultant ordinary
differential equations may then be handled much more conveniently. Consider first
reduction of the set of equations to the differential equation of heat conduction.
The Fourier transformation of a function f x is defined by:
Z 1
Ff x expixxf xdx Fx 4:36
1
and the Fourier inversion by:

Z 1
1
f x Fxexpixxdx 4:37
2p 1
The Fourier transformation of the convolution integral:

Z 1
f ngx sds 4:38
1
is the product of the transforms:
f x:gx 4:39
and the transform of function expjkxj is:
2k
4:40
1 x 2 k2
Therefore, the Fourier transform of the function:
Z 1
k j x sj
IX 3
exp Tl x; tds 4:41
1 k k
will be a constant factor (the value of integral) multiplying the transform of the
function Tl x; t, i.e.:
Z 1
kf jx sj
FIX 3 Tl F exp ds 4:42
k 1 k
or
Z 1
kf jx sj
FIX Tl F exp H jsjds 4:43
k3 1 k
where H jsj 1 for 1\s\1.Therefore:
Z 1
kf jx sj
FIX 3 Tl F exp F fH jsjdsg
k 1 k
4:44
kf 2k
3 Tl dx
k x k2 1
2
where dx is the Dirac delta function. Since this function only has a value of 1 at
x 0, then the transform is:
kf
Tl 4:45a
k2
Using these results, the Eqs. 4.27 and 4.35 can be Fourier transformed, the
result of which is:

oe k f 2k kf Io df 2k 2d
Te 2 Tl
ot k3 x2 k2 1 k 2k x2 k2 1 d2 x2
4:45b
o kf 2k kf Io df 2k 2d
sp Te 2 Tl
ot k3 x2 k2 1 k 2k x2 k2 1 d2 x2
and

k k1 f 2k Io d1 f 2k 2d
Te f Tl Te
k2 k3 x 2 k2 1 2k x 2 k2 1 d2 x 2
4:46
If the transform function Te is obtained from Eqs. 4.45a, 4.45b using 4.46, the
result is:

oe o 2d
f x2 k2 x2 k f Tl sp x2 k f Tl Io df 2
ot ot d x2
o 2d
sp Io df 2 4:47
ot d x2
Insertion of oeot in terms of Tl and multiplication of Eq. 4.47 which is in the
transform domain, by ix2 corresponds to second order differential in the real
plane. Hence the inversion of the above equation gives:

o k2 o2 oTl o2 Tl o o2 Tl
1 ss C l k s p
ot f ox2 ot dx2 ot dx2
o
Io df expdj xj sp Io df expdj xj
ot
4:48
Equation 4.48 is similar to that obtained from the quasi-ballistic

o2 Tapproach
[9]. It
should be noted that the time derivative of diffusion sp ot dx2 o l
and source
(oto Io df expdj xj) terms are included in Eq. (4.48) as similar to the that obtained
from the quasi-ballistic approach [9].
Parabolic Heating Model:
First, let us consider the two-equation model presented previously [10]. The two-
equation model representing the energy exchange mechanism during phonon
absorption and electronphonon coupling after one-dimensional consideration can

be written as [10]:
oTe s; t
Ce r : krTe s; t GTe s; t Tl x; t S
ot 4:49
oTl x; t
Cl GTe s; t Tl x; t
ot
Te s; t and Tl x; t are the electron and lattice site temperatures, S is the laser
source term (Io d expdx), and Ce and Cl are the electron and lattice heat
capacities, respectively. G is the electronphonon coupling factor, given by [11]:
2
p2 me N V
G 4:50
6sp Te s; t
where me , N, V, and sp are electron mass, electron number density, electron drift
velocity, and the electronphonon collision time, respectively.
Now, let us consider the equation
2 derived from the electron kinetic theory
approach (Eq. 4.48). If the terms sp oto odxT2l and sp oto Io df expdj xj are neglected in
Eq. (4.48) for all t values, Eq. 4.48 becomes:

o k2 o2 oTl o2 Tl
1 ss C l k Io df expdj xj 4:51
ot f ox2 ot dx2
which is the same as previously formulated kinetic theory model [12], i.e. the time
derivative of diffusion and source terms are omitted in the previously formulated
kinetic theory approach. Eq. 4.51 can be re-written as:

oTl o2 Tl k2 o2 oTl o2 T l
Cl k 2 f 2 q Cp Cl ss 2 Io d expdj xj 4:52
ot dx ox ot ot
Equation 4.52 is a third order partial differential equation, which can be decom-
posed into second and first order two differential equations, i.e., when Eq. 4.52 is
decomposed into two equations, the resulting probable differential equations are:
oTe o2 Te
A B 2 CTe Tl Io d expdj xj
ot dx 4:53
oTl
D CTe Tl
ot
where A, B, C, and D are the coefficients. To find the values of A, B, C, and D, the
following procedure is adopted, i.e.:

o2 Tl oTe oTl
D 2 C 4:54
ot ot ot
or:
oTe D o2 Tl oTl
4:55
ot C ot2 ot
Similarly:
2
o2 oTl o Te o2 Tl
D C 4:56
ox2 ot ox2 ox2
or:

o2 Te D o2 oTl o2 Tl
4:57
ox2 C ox2 ot ox2
Substitution of Eqs. 4.57 and 4.55 into Eq. 4.53, and inserting CTe Tl DoTotl
into Eq. 4.53, it yields:

oTl BD o2 oTl o2 Tl AD o2 Tl
D A B Io d expdj xj 4:58
ot C ox2 ot ox2 C ot2
After equating Eqs. 4.52 and 4.58, the coefficients A, B, C, and D can be calcu-
lated, i.e.:
f kss
A
k2
Bk
fk f k ss 4:59
C 2 1
k q Cp k2
f k ss
D q Cp 2
k
Equation 4.53 is identical to Eq. 4.49 given in the two-equation model. Conse-
quently, setting the coefficients of equations Eqs. 4.53 and 4.49, it yields:
f k ss
Ce
k2
fk f k ss
2
1 G 4:60
k q Cp k2
f kss
q Cp 2 Cl
k
where ts Ce =G [10] and Ce cTe (where c is constant; for example, c 96:6 m3JK 2
for copper). Moreover, the electron phonon coupling factor is temperature depen-
dent, which can be written as [11]:
2
p2 m N V
G 4:61
6 sp Te Te
Equation 4.61 can be used for lattice site temperature less than the Debye tem-
perature. Since the electronphonon collision time sp T1e , the electron phonon
coupling factor becomes constant as temperature increases [13, 14], i.e. it becomes
independent of temperature.
The electron mean free path (k) can be written as [15]:
k Vs 4:62
where s is the electron relaxation time, which is [16]:
3m k
s
p2 N kB2 Te
where N and m are electron number density and effective mass of free electrons,
respectively. kB is the Boltzmanns constant.
4.3 Thermal Stress Field: Consideration of Surface

and Volumetric Sources
The surface and volumetric source term is associated with the laser output pulse
intensity. In the case of surface heat source consideration, it is assumed that the
laser pulse energy is deposited at the surface of the irradiated material. However,
the volumetric heat source considers the absorption of the incident laser beam
within the skin of the substrate surface. Since the absorption takes place according
to the Beer Lamberts law, exponential decay of the laser intensity along the
absorption depth defines the volumetric heat source in the heat equation.
Equation 4.48 represents the temperature field in the lattice sub-system. The
analysis is presented in line with the previous study [2] and the one-dimensional
form of Eq. 4.48 can be written as:
2
o k2 o2 oTL o2 TL o o TL
1 ss CL k 2 sp k 2
ot f ox2 ot ox ot ox 4:63
f dI t expdjx j
where ss is the electronphonon characteristic time (ss CE =G), G is the elec-
tronphonon coupling factor, k is the mean free path of electrons, f is the fraction
4.3 Thermal Stress Field: Consideration of Surface and Volumetric Sources 135
of excess energy change, CL and CE are the lattice and electron heat capacities,
respectively, k is the thermal conductivity, sp is the electron mean free time
between electronphonon coupling, I t I expd t where I is the laser
peak power intensity, expd t is the temporal distribution function of laser
pulse, d is the absorption coefficient. x is the distance along the x-axis and t is the
time variable. TL and TE are the lattice site and electron temperatures, respectively.
Introducing the following equalities and dimensionless variables:
fk CE
k2 ; ss
G G
4:64
TE TL t k d2 CE
hE ; hL ; x x d; t ; a ;
T T CE =G G CL
to Eq. 4.63 yields finally:
o3 hL o2 hL o2 hL ohL
a 1 l 2
a 2 b expx expc t 0 4:65
ox ot ox2 ot ot
sp G f I d l c
where l CL ; b To G ; c d GCE :
This model provides the improved energy transport equation including ballistic
effects with volumetric source in dimensionless form. Once the lattice site tem-
perature is determined, the electron temperature can be found from:
ohL
hE hL 4:66
ot
Equations 4.65 and 4.66 describe temperature distribution in the lattice and
electron sub-systems.
Moreover, the coefficients of Eq. 4.66 are considered to be independent of
temperature. Hence:
o 3 hL o2 hL o2 hL ohL
a 1 l a 2 b expx expct 0;
ox2 ot ox2 ot ot
ohL
hE hL 4:67
ot
In the absence of the volumetric source, one can get that b 0. So the system
(Eq. 4.67) reduces to the following system:
o3 hL o2 hL o2 hL ohL
a1 l a 2 0;
ox2 ot ox2 ot ot
ohL
hE hL 4:68
ot
Since the absorption depth of the substrate material is significantly smaller than the
thickness of the substrate material, the assumption of a semi-infinite solid heated
by a time decaying source from the surface can be justifiable. The boundary
conditions for the problem can be written as follows:
ohE ohE
0; t I ek1 t ; 1; t 0
ox ox 4:69
hE x; 1 h ; hL x; 1 h
The dimensionless source amplitude I is related to the dimensional one I

through the relationI k dI T , and the dimensionless power k1 is related to the
dimensional one d through the relationk1 d GCE c.
Recently, Ylbas et al. [17] constructed the closed form solutions for temper-
ature distribution with surface heat sources. The surface heat source corresponds to
the short duration contact of the gold film with a heat source at the surface. The
solution of the boundary value problem can be given, when:

k1 [ 1 & k1 \ or k1 \1 & k1 [ 4:70
1l 1l
as follow:
I
hL x; t h Expk1 t xx
xk1 k1
4:71
I
hE x; t h Expk1 t xx
xk1
q
where x a k1 k1k 1
1 1l
:
It should n for hL x; t and hEo

n be noted that the solution
o x; t in Eq. 4.71 is only

valid for k1 [ 1 & k1 \ 1l or k1 \1 & k1 [ 1l . In which case, e is of
2
order 10 for metals and k1 \1for short pulse heating situation.
The equation governing the momentum in one-dimensional solid for a linear
elastic case can be written as [18]:
o2 rx 1 o2 rx o2 T L
2
2 2 c2 2 4:72
ox c1 ot ot
where c1 is the wave speed in the solid.
q
c1 Eq and c2 1m1m q aTL where mis Poissons ratio, qis the density of the
solid and aTL is the thermal expansion coefficient of the solid.
Our goal in this paper will be solving Eq. 4.72 to find the exact solution for
thermal stresses corresponding to the closed form solutions for temperature dis-
tribution hL with surface heat sources. So by using the dimensionless variables in
2
Eq. 4.65, the dimensionless stress isrx c2 dd2 T0 rx . Equation 4.72 can be
re-written after incorporating Eq. 4.71, in which temperature is presented.

o2 rx 2 o rx
2
A B2 Expc t x x 4:73
ox2 ot2
q
I
where A c1dc d and B c xc :
The initial conditions for Eq. 4.73 are as follows:

rx x; 0 0; rx x; 1 0 4:74
Now, applying the Fourier sine transform for Eq. 4.73 with respect to t, by
requiring that, for a physically meaningful system, rx x; t and its first derivatives
in t vanish as t ! 1. To obtain the analytical solution for Eq. 4.73, Fourier sine
and Fourier cosine transformations are used. In this case, Eq. 4.73 yields:
r r
o2 2 2 2 2 2 2 Expc x x
2
U x; p A p rx x; 0 A p U x; p B p 4:75
ox p p c 2 p2
where p is the Fourier sine variable and U x; p is the Fourier sine of rx x; t with
q R
1
respect to t, which is define by Ux; p p2 0 rx x; t sinp tdt, and the
q R
1
inverse Fourier sine of rx x; t is given by rx x; t p2 0 Ux; p sinp tdp.
To apply the Fourier sine transform, with respect to t, in Eq. 4.75, one can use
the following formulas [19]:
2 r
o 2 2
Fs ux; t; t; p p Fs fux; t; t; pg p ux; 0:
ot2 p
r 4:76
at 2 p
Fs fe ; t; pg ; a [ 0:
p a2 p2
By applying the initial condition for Eq. 4.75, one can get
r
o2 2 2 2 2 Expc x x
U x; p A p U x; p B p 4:77
ox2 p c 2 p2
For the boundary conditions, two cases are considered.
Case 1 (stress-free boundary condition):
The thermal stress at the surface is set to zero. The relevant boundary conditions
are:
rx 0; t 0; rx 1; t 0 4:78
Now, applying the Fourier sine transform for Eq. 4.77 with respect to x, by
in x vanish as x ! 1, as well as using the formulas [19]:
r
o2 2 2 o
Fc 2 ux; t; x; s s Fc fux; t; x; sg u0; t:
p ox
2 r
o 2
Fs 2
ux; t; x; s s2 Fs fux; t; x; sg s u0; t:
ox p
r 4:79
ax 2 a
Fc fe ; x; sg ; a [ 0:
p a2 s 2
r
2 s
Fs feax ; x; sg ; a [ 0:
p a2 s 2
The resulting equation is:
r
2 2 1
sU 0; p s2 V s; p A2 p2 V s; p B2 p s 2
p p c p c2 x2 s2
2
4:80
where s is Fourier sine variable and V s; p is the Fourier sine of U x; p with
q R
1
respect to x; which is define by V s; p p2 0 U x; p sins xdx, and the
q R
1
inverse Fourier sine of rx x; t is given by U x; p p2 0 V s; p sins xds:
Applying the boundary conditions by using the formula Ux; p

q R
2 1
p 0 rx x; t sinp tdt gives:
U0; p 0 4:81
So, Eq. 4.80 reduces to:
2 2 ps
V s; p B 4:82
p c2 p2 s2 c2 x2 s2 A2 p2
Applying the inverse Fourier sine transform for Eq. 4.82 with respect to x, and
using tables of Fourier sine transform [19, 20], one can obtain:
r Z 1
2
U x; p V s; p sins xds
p 0
r Z 1
2 B2 p 2 s
sins xds
p c p p 0 s c x s2 A2 p2
2 2 2 2 2
r Z 1
2 B2 p 2 s s
sins xds
p c2 p2 c2 x2 A2 p2 p 0 s2 A2 p2 s2 c2 x2

2 B2 p s s
F s ; s; x Fs ; s; x
p c2 p2 c2 x2 A2 p2 s2 A2 p2 s2 c2 x2
r
2 B2 pcosA p x exp( c x x

p 2 2
A2 p2 c2 p2 c x2A
Therefore:
r
2 B2 p cosA p x exp( c x x
U x; p 4:83
p 2 2
A2 p2 c2 p2 c x2 A
Applying the inverse Fourier sine transform for Eq. 4.83 with respect to t, one can
have:
8 cx
c t
>
> e coshc
> A x e A t coshc x x
>
> ; tAx
>
>
cx
< ec x x e c t e A t
2
rx x; t c2 xB2 A2 cx
>
> e c A x sinhc t e c x x sinh t
>
> A
>
> ; t \ Ax
> cx
: ec x x e c t e A t
4:84
Case 2 (stress gradient-free boundary condition):
The thermal stress gradient at the surface is set to zero. The relevant boundary
conditions are:
orx
0; t 0; rx 1; t 0 4:85
ox
Now, applying the Fourier cosine transform for Eq. 4.77 with respect to x, by
in x vanish as x ! 1 and the re-arrangement results in:
r
2 o 2 p
U 0; p s2 W s; p A2 p2 W s; p B2 c x 2
p ox p c p2 c2 x2 s2
4:86
where s is Fourier cosine variable and W s; p is the Fourier cosine of U x; p with

q R
1
respect to x; which is define by W s; p p2 0 U x; p coss xdx, and the
q R
1
inverse Fourier cosine of rx x; t is given by U x; p p2 0 W s; p coss xds:

q R
2 1
o
U 0; p 0 4:87
ox
2 2 p
W s; p B cx 2 4:88
p c p2 s2 c2 x2 s2 A2 p2
Applying the inverse Fourier cosine transform for Eq. 4.88 with respect to x,
one can have:
r
2 B2 A exp( c x x p c x sinA x p
U x; p 4:89
p A3 2 2
p2 c2 p2 c x2 A
Applying the inverse Fourier sine transform for Eq. 4.24 with respect to t, one can
obtain by using Fourier cosine transform [19, 20]. Therefore,
8 x cx
>
> ec t sinh c A x e A t sinhc x x
>
> A
>
> ;tAx
2
> cx
< exp( c x x ec t e A t
B
rx x; t 2 2 x c A x c x
c x A2 >
> e sinhc t e c x x
sinh t
>
> A A
>
> ; t\A x
>
: exp( c x x ec t ecAx t
4:90
Equations 4.84 and 4.90 can be used to compute the stress distribution in accor-
dance with the physical properties.
Equation 6.67 is applicable for the volumetric heat source heating. Since the
irradiated spot size is small, one can assume a semi-infinite substrate material
being subjected to laser heated. The analysis presented below is in line with the
previous study [3]. The boundary conditions for the problem can be written as
follows:
ohE ohE
0; t 0; 1; t 0
ox ox 4:91
hE x; 1 h ; hL x; 1 h
Recently, Ylbas et al. [17] constructed the closed form solutions for temperature
distribution with volumetric heat sources. The solution of the boundary value
problem
n can be given, owhen:n o

c [ 1 & c \ 1l or c\1 & c [ 1l as follow:

hL x; t h H Expct x x Expx ct
xc
4:92
c
hE x; t h H Expct x x cH Expx ct
xc
q
1c b
where x acc1l 2 and H c2 cal11a :
It should
n be noted that theo solution
n for hL x; t and
o hE x; t in Eq. 4.92 is only

valid for c [ 1 and c\ 1l or c\1 and c [ 1l . In which case, e is of order
102 for metals and c\1for short pulse heating situation.
Now, the equation governing the momentum in one dimensional solid for a
linear elastic case can be written as [18]:
o2 rx 1 o2 rx o2 TL
2
2 2 c2 2 4:93
ox c1 ot ot
q
E
where c1 is the wave speed in the solid, which is c1 q, c2 1m
1m q aTL , m is
Poissons ratio, qis the density of the solid, and aTL is the thermal expansion
coefficient of the solid.
Our aim is solving Eq. 4.93 to find the exact solution for thermal stresses corre-
sponding to the closed form solutions for temperature distribution hL with volumetric
heat sources. So by using the dimensionless variables Eq. 4.64, the dimensionless
2
stress is rx c2 dd2 T0 rx . Equation 4.93 can be re-written after incorporation
Eq. 4.92, in which temperature is presented. Therefore Eq. 4.93 becomes:

o2 rx 2
2 o rx 2 1
A B Expct c x x Expc t x 4:94
ox2 ot2 cx
p
where A c1dc d and B H :
The initial conditions for Eq. 4.94 are as follows:
rx x; 0 0; rx x; 1 0 4:95
Now, applying the Fourier sine transform for Eq. (4.94) with respect to t, by
in t vanish as t ! 1, To apply the Fourier sine transform, with respect to t, in
Eq. 4.96. One can use the following formulas [19, 20]:
2 r
o 2 2
Fs 2 ux; t; t; p p Fs fux; t; t; pg p ux; 0:
p
r 4:96
at 2 p
Fs fe ; t; pg ; a [ 0:
p a2 p2
Therefore:
r
o2 2 2
U x; p A p rx x; 0 A2 p2 U x; p
ox2 r p
2 B2 Expc x x c x Expx
p 4:97
p cx c 2 p2
where p is the Fourier sine variable and U x; p is the Fourier sine of rx x; t with
q R
1
respect to t, which is define by Ux; p p2 0 rx x; t sinp tdt, and the inverse
q R
1
Fourier sine of rx x; t is given by rx x; t p2 0 Ux; p sinp tdp:
By applying the initial conditions for Eq. 4.97, one can get:
r
o2 2 2 2 B2 Expc x x c x Expx
2
U x; p A p U x; p p 4:98
ox p cx c 2 p2
For boundary conditions, two cases are considered. These are given below:
Case 1 (stress-free boundary condition):
The thermal stress at the surface is set to zero. The relevant boundary conditions
are:
rx 0; t 0; rx 1; t 0 4:99
Now, applying the Fourier sine transform for Eq. 4.98 with respect to x, by
in x vanish as x ! 1, Therefore:
r
2
s U 0; p s2 V s; p A2 p2 V s; p
p
4:100
2 B2 p s s
cx 2
p c x p2 c 2 s 2 c 2 x 2 s 1
where s is Fourier sine variable and V s; p is the Fourier sine of U x; p with

q R
1
respect to x; which is define by V s; p p2 0 U x; p sins xdx, and the
q R
1
inverse Fourier sine of rx x; t is given by U x; p p2 0 V s; p sins xds:
q
Applying the boundary conditions by using the formula Ux; p p2
R1
0 rx x; t sinp tdt, it gives:
U0; p 0 4:101

2 B2 p s s
V s; p c x 4:102
p c x p2 c2 s2 A2 p2 s2 c2 x2 s2 1
Applying the inverse Fourier sine transform [19, 20] for Eq. 4.107 with respect
to x, one can have:
r Z 1
2
U x; p V s; p sins xds
p 0
r Z 1
2B2 p 2 1 s s
c x sins xds
p c x p2 c2 p 0 s2 A2 p2 s2 c2 x2 s2 1
r Z 1
2B2 p 2 s s
sins xds
p c x p2 c2 c2 x2 A2 p2 p 0 s2 A2 p2 s2 c2 x2
r Z 1
2B2 p 2 s s
sins xds
p p2 c2 1 A2 p2 p 0 s2 A2 p2 s2 1

2B2 p s s
F s ; s; x F s ; s; x
p c x p2 c2 c2 x2 A2 p2 s2 A2 p2 s2 c2 x 2
2

2B p s s
F s ; s; x F s ; s; x
p p2 c2 1 A2 p2 s2 A2 p2 s2 1
r r
2
2 B pcosA p x exp( c x x 2 2 pcosA p x exp( x
B

p c x A2 p2 c2 p2 c2 x2 2 p A2 p2 c2 p2 A12
A
4:103
or
r
2 B2 p cosA p x exp( c x x
U x; p
p c x A2 p2 c2 p2 c2 x2 2
A
r
2 2 p cosA p x exp( x
B
4:104
p A2 p2 c2 p2 A12
Applying the inverse Fourier sine transform [20, 21] for Eq. 4.104 with respect to
t, it yields:
r r Z 1 0 1
2 B2 2 @ 1 p cos A p x exp( c x x p cos A p x exp( x
rx x; t
A sinp tdp
p A2 p 0 cx p2 c2 p2 cAx2
2 2
p2 c2 p2 A12
r 0 1 r 0 1
2 B2 p cos A x p 2 B2
p
Fs @ ; p; tA e c x x
Fs @ ; p; tA
p A2 c x p2 c2 p2 cAx2
2 2
2 2
r ! r !
2 B2 p cosA x p 2 B2 x p
Fs
; p; t e Fs
; p; t
p A2 p2 c2 p2 A12 p A2 p2 c2 p2 A12
4:105
where
0 1 0 8 91
< =
p 1 p p
Fs @ ; p; tA 2 2 @Fs ; p; t Fs ; p; t A
c 2 x2
p2 c2 p2 A2 c x 2 2 2
p c : p2 c 2 x2
;
A2 c A2
r r !
1 p c t cx
p 2 ec t e A t
cx
2 2 e e A t A
c x
2 c
2 2 2 c2 x2 A2
A
4:106
and
! ( )!
p 1 p p
Fs
; p; t
1 Fs ; p; t Fs
2 1 ; p; t
p2 c2 p2 A12 A2 c
2 p2 c 2 p A2
r r !
1 p ct 1
p 2 ect eAt
1
t

1 e e A A
A2 c
2 2 2 1 A 2 c2
4:107
and
0 1 0 8 91
<p cosA x p =
p cosA x p 1 p cosA x p
Fs @ A
; p; t 2 2 @
Fs ; p; t Fs ; p; t A
2 2 2
p c p A2c 2 x2 c x 2 p 2 c2 : 2 c2 x 2 ;
A2 c p A2
8 p p cx
> p c t p At
> 2e cosh c A x 2e
>
<
coshc x x; t [ A x
1 pp e2 c t e2cAxt

2 2 ; tAx
c2 >
c x 2 2
>
A2 : ppec A x sinhc t pp ec x x sinh
c x t; t \ A x
>
2 2 A
4:108
and
! ( )!
p cosA x p 1 p cosA x p p cosA x p
Fs
; p; t
1 Fs ; p; t Fs
2 ; p; t
p2 c2 p2 A12 A2 c
2 p2 c 2 p A12
8 p p 1
p ct
>
> 2e coshcAx
> p2eAt coshx; t [ A x
1 < pp e2ct e 2 A1 t

1
2 > 2 2 ; t Ax
c >
: ppecAx sinhct ppex sinh 1 t; t\A x
>
A 2
2 2 A
4:109
Therefore, the solution becomes:
8 B2

c t cx
c x
>
> c3 x x2 A2 e coshc A x e A t coshc x x ec x x ec t e A t
>
>
>
< 1A B2 1
ec t cosh c A x eA t cosh x ex e c t eAt
1
; t Ax
2 c2
rx x; t B2

c A x

c x
>
>
> c3 x x2 A2 e sinhc t ec x x sinh cAx t ec x x ec t e A t
>
>

: 1A B2 1
ec A x sinhc t ex sinh A1 t ex ec t eA t ; t\Ax
2 c2
4:110
Case 2 (stress continuity boundary condition at the surface):
The thermal stress gradient at the surface is set to zero. The relevant boundary
conditions are:
orx
0; t 0; rx 1; t 0 4:111
ox
Now, applying the Fourier cosine transform for Eq. 4.98 with respect to x, by
in x vanish as x ! 1, as well as using the formulas in appendix, results in:
r
2 o
U 0; p s2 W s; p A2 p2 W s; p
p ox
2 2 p 1 1
B 4:112
p p2 c 2 s 2 c 2 x 2 s 2 1
where s is Fourier cosine variable and W s; p is the Fourier cosine of U x; p with
q R
1
respect to x; which is define by W s; p p2 0 U x; p coss xdx, and the
q R
1
inverse Fourier cosine of rx x; t is given by U x; p p2 0 W s; p coss xds:

q R
2 1
o
U 0; p 0 4:113
ox
So, Eq. 4.112 will reduce to:

2 2 p 1 1
W s; p B 4:114
p p2 c 2 s 2 A2 p2 s 2 c 2 x 2 s 2 1
Applying the inverse Fouriercos transform [20, 21] for Eq. 4.114 with respect
to x, it yields:
r Z 1
2
U x; p W s; p coss xds
p 0
r Z 1
2 2 p 2 1 1 1
B 2 coss xds
p p c2 p 0 s2 A2 p2 s2 c2 x2 s2 1
r Z 1
2 2 p 2 1 1
B 2 coss xds
p p c2 c2 x2 A2 p2 p 0 s2 A2 p2 s2 c2 x2
r Z 1
2 p 2 1 1
B2 2 coss xds
p p c2 1 A2 p2 p 0 s2 A2 p2 s2 1

2 p 1 1
B2 2 F c ; s; x F c ; s; x
p p c2 c2 x2 A2 p2 s 2 A2 p 2 s2 c2 x2

2 p 1 1
B2 2 F c ; s; x F c ; s; x
p p c2 1 A2 p2 s2 A2 p2 s2 1
r

p c x A3 2 2
p2 c2 p2 c Ax2
r
2 B2 A exp( x p sinA x p

p A3 p2 c2 p2 A12
Therefore,
r
U x; p
p c x A3 2 2
p2 c2 p2 c Ax2
r
2 B2 A exp( x p sinA x p

4:115
p A3 p2 c2 p2 A12
Applying the inverse Fourier sine transform for Eq. 4.115 with respect to t, it
yields:
r r Z 1 0 1
2 B2 2 @ 1 A exp( c x x p c x sin A x p A exp( x p sin A x p A sinp tdp
rx x; t

p A3 p 0 cx 2 2
p2 c2 p2 c Ax2 p2 c2 p2 A12
r 0 1 r 0 1
2 B2 @ p A 2 B Fs @
2
sinA x p A
exp( c x x Fs ; p; t ; p; t
2 2
p A3 p2 c2 p2 cAx2
2 2
r ! r !
2 B2 p 2 B2 sinA x p
2
exp( x Fs
1
; p; t 3
Fs
1
; p; t
pA 2 2 2
p c p A2 pA 2 2 2
p c p A2
4:116
or
8 x
> B2 cx cx
>
>
> 3 2 2
ec t sinh c A x e A t sinh c x x exp( c x x ec t e A t
>
> c x x A A
>
> ;tAx
>
>
> B2 1 c t
>
> e
1 1
sinh c A x eA t sinh x exp( x e c t e A t
< 2 2
1 A c A c
rx x; t x c x
>
> B2 cx
>
> ec A x sinhc t ec x x sinh t exp( c x x ec t e A t
>
> c3 xx2 A2 A A
>
>
> ; t\A x
>
>
> B2 1 c A x 1
>
: e sinhc t e x sinh t exp( x ec t e A t
1
1 A 2 c2 A c A
4:117
Equations 4.11 and 4.117 can be used to compute the stress fields in the substrate
material subjected to the laser heating pulse.
4.4 Thermal Stress Field: Two-Dimensional Consideration
Two-dimensional axisymmetric form of lattice site temperature equation due to

laser short-pulse heating can be written in line with the previous study [4];
therefore, the heat equation in two-dimensional cylindrical coordinate system
yields:

o k2 1 o o2 o2 oTL o
1 ss 2 2 C L 1 sp
ot f r or or oy ot ot
2 2 4:118
1o o o
k TL f d Tr I t expdjy j
r or or 2 oy2
where ss is the electronphonon characteristic time (ss CE =G), G is the elec-
tronphonon coupling factor, k is the mean free path of the electrons, f is the
fraction of excess energy change, CL and CE are the lattice and electron heat
capacities, respectively, k is the thermal conductivity, sp is the electron mean free
time between electronphonon coupling, I t I expd t where I is the
laser peak power intensity, expd t is the temporal distribution function of
laser pulse, d is the absorption coefficient, Tr 1 rf where rf is surface
reflectivity, r is the distance along r-axis, y is the distance along y -axis, t is the
time variable and TL is the lattice site temperatures. Introducing the following
equalities and dimensionless variables:
fk CE
k2 ; ss
G G
4:119
TL T t k d2 CE
hL ; x r d; y y d; t ; a ;
T CE =G G CL
to Eq. 4.118 yields finally:
3 2
o hL o3 hL 1 o2 hL o hL o2 hL 1 ohL o2 hL ohL
a 1 l 2 a
ox2 ot oy2 ot x oxot ox2 oy2 x ox ot2 ot
b Tr expy expc t
4:120
s G
where l Cp L ; b f I dTl
oG
c
; c d GCE :
This model is the improved energy transport equation including ballistic effects
with the presence of the source term in dimensionless form.
By using Maple program, one can find that Eq. 4.120 admits the following Lie
point symmetries:
1o o o 1o o
X1 hL : X2 : XF Fx; y; t 4:121
c ot ohL oy c ot ohL
where Fx; y; t is a solution for the following homogenous equation:
3 2
a 1 l 2
2
a 2
2
2 0
ox ot oy ot x oxot ox oy x ox ot ot
4:122
And since the invariants of X1 are:
x; y; ec t hL 4:123
Then the corresponding similarity solution can be given as:
hL ux; y ec t 4:124

This similarity solution reduces the boundary conditions, so the boundary value
problems in both of the following two cases can be reduced.
Case 1: Volumetric source:
Now, consider a semi-infinite cylindrical nano-sized wire subjected to a laser
short-pulse heating (Fig. 4.2).
The boundary conditions for the problem can be written as follows:
4.4 Thermal Stress Field: Two-Dimensional Consideration 149
Fig. 4.2 A schematic view Laser Beam

of the irradiated element and
the coordinate system
x=L
Symmetry Axis
oTL
0; y ; t 0 L L; y

; t T
or
oTL
r; 0; t 0 TL r; 1; t T 4:125
oy
oTL
r; y ; 1 0 TL r; y ; 1 T
ot
Using the dimensionless variables Eq. 4.109 yields the dimensionless boundary
conditions:
ohL
0; y; t 0 hL L d; y; t 0
ox
ohL
x; 0; t 0 hL x; 1; t 0 4:126
oy
ohL
x; y; 1 0 hL x; y; 1 0
ot
The similarity solution hL ux; y ec t transforms Eq. 4.120 to the second
order partial differential equation (PDE):
o2 u o2 u 1 ou
A2 ux; y B2 Tr ey 4:127
ox2 oy2 x ox
p p p
c 1c b
where A r ; B r and r a l 1c and transforms the
dimensionless boundary conditions to the following boundary conditions:
ou
0; y 0; uL d; y 0
ox 4:128
ou
x; 0 0; ux; 1 0
oy
Now, applying the Fourier cosine transform for Eq. 4.127 with respect to y, by
requiring that, for a physically meaningful system, ux; y and its first derivatives
in y vanish as y ! 1, which results in:
r r
o2
2 2
2 ou 1o 2 B2 Tr
U x; s s A U x; s x; 0 U x; s
ox2 p oy x ox p s 2 1
4:129
where s is Fourier cosine variable and U x; s is the Fourier cosine of ux; y with
q R
1
respect to y; which is define by U x; s p2 0 ux; y coss ydy, and the inverse
q R
1
Fourier cosine of U x; s is given by ux; y p2 0 U x; s coss yds.
Applying the boundary conditions on y; Eq. 4.128, it gives:
r
o2 1o
2 2
2 B2 Tr
2
U x; s U x; s s A U x; s 4:130
ox x ox p s 2 1
p
The transformation z s2 A2 xtransforms Eq. 4.130 to the inhomogeneous
modified Bessel equation:
r
o2 1o 2 B2 Tr
U x; s U z; s U z; s 4:131
oz2 z oz p s2 1s2 A2
It should be noted that Eq. 4.131 is the inhomogeneous
n modified oBessel

equation for real value of A: i.e. when c [ 1 & c \ 1l or
n o
c\1 & c [ 1l
. In which case, e is of order 102 for metals and c\1:
So, the general solution of Eq. 4.131 can be given as:
p
U x; s F1 sBesselI 0; s2 A2 x
p r
2 B2 Tr
4:132
F2 sBesselK 0; s2 A2 x
p s2 1s2 A2
where BesselI and BesselK are the modified Bessel functions of the first and
second kinds, respectively.
Applying the Fourier cosine to the boundary conditions on x, Eq. 4.128, it
gives:
oU
0; s 0 U L d; s 0 4:133
ox
Therefore, the solution of Eq. 4.131 can be given finally as:

r
p !
2 B2 Tr BesselI 0; s2 A2 x
U x; s 1
p 4:134
p s 2 1 s 2 A2 BesselI 0; L d s2 A2
Then the final solution is the inverse Fourier cosine transform to U x; s, i.e.:
r Z
c t 2 c t 1
hL x; y; t ux; y e e U x; s coss yds 4:135
p 0
where U x; s is given by 4.135.

It should be noted that U x; s doesnt have singularity and although the range of
integration is from 0 to 1, the improper integral is convergent and can be evaluated
numerically at given x; y by using adaptive GaussKronrod quadrature [21]. This
method is included in MATLAB, where the function quadgk(fun,a,b,0 RelTol0 ,1e-
3,0 AbsTol0 ,1e-3) attempts to approximate the integral of a scalar-valued function,
fun, from a to b using high-order global adaptive quadrature and take both of
relative error tolerance and absolute error tolerance as 10-3. Moreover, the function
of quadgk attempts to satisfy that error bound \= max(AbsTol,RelTol*|Q|) so
that the results have an excellent accuracy.
Also, since the value of the modified Bessel function of the first kind exceeds
the range of MATLAB environment for large s. One can avoid any undefined
numerical results by scaling the modified Bessel function of the first kind using the
function besseli(nu,z,1),this function is included in MATLAB, where the function
besseli(nu,z,1),scales the modified Bessel function of the first kind, besseli(nu,z),
by exp(-abs(real(z))).
Case 2: Surface source:
In the absence of the volumetric source, we get that b 0. Hence, the Eq. 4.122
reduces to the following equation:
3 2
a 1 l 2
2 a 2
2 2 0
ox ot oy ot x oxot ox x ox ot ot
4:136
Now, assume a semi-infinite substrate material heated with a time decaying source
from the surface (Fig. 4.2). The boundary conditions for the problem can be
written as follows:
oTL
0; y ; t 0 TL L; y ; t T
or
oTL I Tr

r; 0; t expd t TL r; 1; t T 4:137
oy k
oTL
r; y ; 1 0 TL r; y ; 1 T
ot
Using the dimensionless variables Eq. 4.119 yields the dimensionless boundary
conditions
ohL
ox 0; y; t 0 hL L d; y; t 0
ohL
oy x; 0; t I Tr expc t hL x; 1; t 0 4:138
ohL
ot x; y; 1 0 hL x; y; 1 0
where the dimensionless source amplitude I is related to the dimensional one I
through the relation:
I I
4:139
k d T
The similarity solution hL ux; y ec t transforms Eq. 4.136 to the second order
partial differential equation (PDE):
o2 u o2 u 1 ou
A2 ux; y 0 4:140
ox2 oy2 x ox
p p
c 1c
where A r and r a l 1c and transforms the dimension-
less boundary conditions to the following boundary conditions:
ou
0; y 0 uL d; y 0
ox
4:141
ou
x; 0 I Tr ux; 1 0
oy
Now, applying the Fourier cosine transform for Eq. 4.140 with respect to y, by
requiring that, for a physically meaningful system, ux; y and its first derivatives
in y vanish as y ! 1, as well as using the formulas in appendix, results in:
r
o2
2 2
2 ou 1o
U x; s s A U x; s x; 0 U x; s 0 4:142
ox2 p oy x ox
where s is Fourier cosine variable and U x; s is the Fourier cosine of ux; y with
q R
1
respect to y; which is define by U x; s p2 0 ux; y coss ydy, and the inverse
q R
1
Fourier cosine of U x; s is given by ux; y p2 0 U x; s coss yds:
Applying the boundary conditions on y; Eq. 4.138, gives:
r
o2 1o
2 2
2
2
U x; s U x; s s A U x; s I Tr 4:143
ox x ox p
p
Similarly, the transformation z s2 A2 x transforms Eq. 4.143 to the inho-
mogeneous modified Bessel equation:
r
o2 1o 2 I Tr
U x; s U z; s U z; s 4:144
oz2 z oz p s 2 A2
So, the general solution of Eq. 4.144 can be given as:
p p
U x; s F1 sBesselI 0; s2 A2 x F2 sBesselK 0; s2 A2 x
r
2 I Tr

p s2 A2
4:145
where BesselI and BesselK are the modified Bessel functions of the first and
second kinds, respectively.
Applying the Fourier cosine to the boundary conditions on x; Eq. 4.138, gives:
oU
0; s 0 U L d; s 0 4:146
ox
Therefore, the solution of Eq. 4.144 can be given finally written as:
r
p !
2 I Tr BesselI 0; s2 A2 x
U x; s 1
p 4:147
p s2 A 2 BesselI 0; L d s2 A2
Then, the final solution is the inverse Fourier cosine transform to U x; s, i.e.:
r Z
c t 2 c t 1
hL x; y; t ux; y e e U x; s coss yds 4:148
p 0
where U x; s is given by Eq. 4.148. Similarly, this improper integral, in

Eq. 4.148, is convergent and it can be evaluated numerically at given x; y.
Therefore, Eqs. 4.117 and 4.148 are used to compute the stress field in the
substrate material subjected to the laser heating pulse.
4.5 Findings and Discussions
The findings of the stress equations are presented under the appropriate sub-
heading and in line with the previous studies [24]. Two different stress boundary
conditions are incorporated to account for the stress continuity at the surface and
the stress free boundary conditions. In addition, the surface and volumetric heat
sources are considered in the analysis. The surface heat source assumes the
deposition of laser energy at the surface of the irradiated material while the vol-
umetric heat source considers the absorption of the incident irradiated energy in
the absorption depth of the material according to the Beer Lamberts law.
Two different boundary conditions are incorporated in the thermal stress analysis,
which are stress free boundary at the surface and stress continuity at the surface.
The stress free boundary condition represents the free expansions of the surface
while stress continuity corresponds to the presence of optically transparent film at
the surface. The discussions will be presented in line with the previous study [2].
Table 4.1 gives the parameters used in the simulations.
Figure 4.3 shows dimensionless lattice site temperature distribution inside the
substrate material for different dimensionless heating times. Lattice temperature
increases along the x-axis due to energy gain from the excited electrons through
the collisional process. Since electrons undergo several collisions with increasing
distance, lattice site temperature increases along the x-axis accordingly. As the
heating period increases further, lattice site temperature increase becomes almost
steady inside the substrate material. Temperature gradient is high in the early
heating period due to high rate of energy gain of lattice site from electrons through
the collisional process in the early heating period. Temperature gradient reduces
gradually along the x-axis as the heating time progresses.
Case I: Stress free boundary
Figure 4.4 shows temporal variation of dimensionless thermal stress at different
dimensionless locations inside the substrate material. Temporal behavior of ther-
mal stress is in the wave form and it propagates at a constant speed into the
substrate material. The maximum amplitude of the thermal stress increases slightly
with progressing time, which is attributed to the temperature gradient developed in
the substrate material, which modifies the stress field in the substrate material. The
rise of stress amplitude is higher than its decay in the stress wave. This indicates
that stress wave dies gradually with progressing time while it propagates into the
substrate material. In addition, the thermal stress generated is compressive in the
surface region, i.e. stress value is negative. As the heating period progresses, it
becomes tensile, i.e. stress value becomes positive. The compressive stress initi-
ation at some depth below the surface x [ 0 reveals that material undergoes
contractions due to sudden thermal expansion of the surface. As the heating period
progresses, temperature field inside the substrate increases resulting thermal
expansion of the substrate material where it is initially compressed by the thermal
expansion of the surface. Consequently, compressive stress replaces with tensile
stress in the surface region with progressing time. The magnitude of compressive
stress reduces with progressing time, which is attributed to the high temperature
gradient (dT/dx) dtjdx developed in the solid bulk with progressing time.
Figure 4.5 shows dimensionless stress distribution inside the substrate material
for different dimensionless heating periods. Thermal stress is tensile in the surface
region and it becomes compressive as the distance from the surface increases
towards the solid bulk. The tensile behavior of thermal stress is attributed to
thermal expansion of the surface, since stress free boundary condition is in con-
sidered at the surface. As the heating progresses, the region of tensile stress
Table 4.1 Properties used in Property Numerical value

the solution of temperature
filed of gold CL 2:8
106 J m3 K1
CE 2:1
104 J m3 K1
G 2:6
1016 W m3 K1
k 315 W m1 K1
d 109 m1
d 0:03
1013 s1
sp 0:024
1012 s
T 300 K
h 1
I 1013 W/m2
E 207
109 Pa
q 7930 kg/m3
aTL 16
106 K1
m 0.3

lattice site temperature
distribution inside the
heating periods
extends further inside the substrate material. This is because of the extension of
heated region inside the substrate material with progressing time. Therefore,
thermally expanded region extends further inside the substrate material resulting in
tensile stress in this region. The maximum magnitude of dimensionless tensile
stress is on the order of 30, which is about 20 MPa in a dimensional form. The
maximum tensile stress is less than the yielding limit of the substrate material and
crack does not form at the surface vicinity. On the other hand, some distance
below the surface, material cannot expands freely which results in the formation of
compressive stress in this region as observed from Fig. 4.5. The magnitude of
tensile stress increases with progressing time due to the increasing temperature
gradient below the surface. This also suppresses the compressive stress magnitude
inside the substrate material due to extension of the heated region below the
surface.

material for free stress at the
surface case (Case 1)

heating periods and for free
stress at the surface case
(Case 1)
Case II: Stress continuity at the surface

Figure 4.6 shows temporal variation of dimensionless thermal stress at different
dimensionless locations inside the substrate material. The magnitude of thermal
stress wave increases sharply below the surface and it decays gradually with
progressing time. The time corresponding to the maximum stress becomes longer
as the depth below the surface increases. This is because of the propagation of the
stress wave; in which case, time taken for the wave to reach at some depth below
the surface becomes long. The consideration of stress continuity at the surface
modifies the temporal stress distribution inside the substrate material resulting in
decreasing peak stress in the substrate material. The stress continuity condition at
the surface results in compressive stress waves inside the workpiece. In this case,
the surface of the workpiece is not free to expand during the heating process
because of the presence of stress boundary at the surface.

material for stress continuity
at the surface case (Case 2)
Figure 4.7 shows dimensionless stress distribution inside the substrate material
for different dimensionless times. Thermal stress remains high in the region close
to the surface and as the depth below the surface increases; it forms a peak first
and, then, decays gradually with increasing distance towards the solid bulk. The
attainment of high thermal stress levels in the surface region is associated with
the presence of stress boundary at the surface; in which case, a region close to the
surface cannot expand freely during the heating process. When comparing
Figs. 4.4 and 4.7, it can be observed that the behavior of the stress wave due to
stress free surface condition and stress continuity at the surface is different. In this
case, tensile stress region totally replaces with the compressive stress with high
amplitude for the stress continuity at the surface case. In addition, the rise and fall
of the stress waves also differ significantly. The maximum magnitude of dimen-
sionless compressive stress is on the order of 80, which is about 50 MPa in a

heating periods and for stress
continuity at the surface case
(Case 2)
dimensional form. Consequently, the maximum compressive stress is less than the
yielding limit of the substrate material and crack does not initiate at the surface
vicinity.
The discussions for the findings are presented in line with the previous studies [3].
The properties used in the simulations are similar to those given in Table 4.1.
Figure 4.8 shows dimensionless lattice site temperature distribution along the
substrate material for different heating periods. Lattice site temperature increases
with increasing depth below the surface, which is more pronounced with
increasing heating periods. This is associated with electron excess energy transfer
to the lattice site through the collisional process with progressing time. In this case,
the number of electron lattice site collisions increases with progressing time
resulting in energy gain of the lattice site. Since the electron excess energy transfer
to lattice is considerably small during a single collision, successive collisions with
progressing time enhances lattice site temperature increase inside the substrate
material. In addition, the number of collisions increases with increasing depth;
which in turn elevates lattice site temperature rise with increasing depth below the
surface. Lattice temperature increase is sharp in the surface region in the early
heating period and it becomes gradual as the depth below the surface increases.
This is associated with electron energy transfer through the collisional process,
which increases with increasing depth below the surface [1]. As the heating period
increases, lattice temperature along the depth below the surface becomes gradual.
Figure 4.9 shows dimensionless thermal stress developed inside the substrate
material for different heating periods for stress free surface boundary condition.
Thermal stress is tensile in the surface region for all heating periods and it
becomes compressive at same depth below the surface. The tensile stress behavior
is attributed to the free expansion of the surface during the heating period.

lattice temperature 1.000000
distribution for different
heating durations
0.999999
t=10
L
0.999998 t=20
t=30
0.999997
0.999996
0 50 100 150 200 250 300
x

workpiece for stress free t=10
2.0E-03
boundary condition at the t=20
surface t=30
0.0E+00

-2.0E-03
-4.0E-03
-6.0E-03
0 50 100 150 200 250 300
x
However, thermal expansion replaces to thermal contraction at same depth below

the surface; in this case, material in this region is not free to expand and free
expansion of the surface results in contraction in this region. However, as the
heating period progresses, temperature gradient in the surface region reduces
causing less thermally induced strain in the surface region. This lowers the
maximum tensile stress magnitude in the surface region. The wave nature of
heating results in wave nature of thermal stress fields in the irradiated region. As
the thermal stress reduces at the surface region, compressive stress magnitude
increases below the surface. Moreover, the depth of the tensile stress region
increases below the surface as the heating period progresses. In the case of stress
continuity at the surface (Fig. 4.10), stress wave is modified in the surface region
such that peak stress reduces with progressing time. Moreover, stress field inside
the substrate material becomes compressive for all heating periods; which is
because of the constraint introduced at the surface. The peak stress moves further
into the substrate material with progressing time.
Fig. 4.10 Temporal 4.0E-03

2.0E-03
locations inside the
workpiece for stress free
boundary condition at the 0.0E+00
surface

-2.0E-03
x=100
x=200
-4.0E-03 x=300
-6.0E-03
0 30 60 90 120 150
t
Figure 4.11 shows temporal behavior of the thermal stress at different dimen-
sionless depths below the surface. Thermal stress is compressive at dimensionless
depths x 100 in the early heating period. However, as the heating period pro-
gresses it becomes tensile in these regions. This is attributed to wave nature of
thermal stress, which propagates into the substrate material with progressing time.
The attainment of the high compressive stress is balanced with low magnitude
tensile stress with progressing time at a certain location inside the substrate
material. When comparing the time occurrence of the peak stress, it is evident that
thermal stress wave propagates at a constant speed inside the substrate material.
This is because of the assumption of constant acoustic wave speed in the stress
analysis. In the case of stress continuity at the surface Fig. 4.12, thermal stress
remains compressive inside the substrate material and peak stress decays with
increasing depth below the surface. In this case, temporal distribution of stress
becomes more extended over the time as compared to that corresponding to close
to the surface.
4.5.3 Two-Dimensional Analysis
The volumetric heat source is incorporated in the analysis to account for

absorption of the laser irradiated energy. The findings are discussed in line with the
previous study [4]. Tables 4.2 and 4.3 gives the properties used in the simulations.
Figure 4.13 shows dimensionless temperature distribution along the x-axis for
different dimensionless time. It should be noted that the y-axis location is y = 0,
which corresponds to the free surface of the irradiated element. Temperature
decays along the x-axis towards the outer radius of the wire due to the consid-
eration of low temperature boundary at the outer edge of the wire. However,
temperature decay in the central region of the wire x 0:025 is gradual and it
becomes sharp as the distance increases towards the wire edge. In this case, energy
Fig. 4.11 Temporal 0.0E+00

locations inside the -2.0E-03
workpiece for stress
continuity boundary
condition at the surface
-4.0E-03

x=100
-6.0E-03 x=200
x=300
-8.0E-03
0 30 60 90 120 150
t
Fig. 4.12 Dimensionless 0.0E+00

workpiece for stress t=10
continuity boundary t=20

-2.0E-03
condition at the surface t=30
-4.0E-03

-6.0E-03
-8.0E-03
0 50 100 150 200 250 300
x
absorbed in the surface region of the workpiece enhances the internal energy gain
at the substrate material in the irradiated region. Consequently, radial heat transfer
along the x-axis is much less than that of along the y-axis due to the absorption,
which takes place along the y-axis. Moreover, the presence of low temperature
boundary at the wire edge acts like a heat sink while lowering temperature sharply
towards the nano-sized wire edge. This is more pronounced for silver, then, fol-
lows chromium, and copper. Since temperature rise for chromium and copper
along the x-axis is much lower than that of silver, temperature decay at central and
towards of the wire edge appears to be gradual. As the heating period progresses,
temperature reduces because of the intensity decay with progressing time (expo-
nential heating pulse). When comparing temperature distribution due to silver,
chromium and copper, it can be seen that the maximum temperature occurs for
silver, then follows chromium, and copper.
Table 4.2 Physical properties of the materials used in the simulations

d
107 CL
106 k G
1016 CE
104 ss
1012
1=m J =m3 K W =m K W =m3 K J =m3 K s
Silver 7.1 1.5 35 14.4 1.1664 0.081
Chromium 6.7 3.3 94 42.4 5.8088 0.137
Copper 6.7 3.43 386 26 4.0612 0.1562
Table 4.3 Properties used in Property Numerical value

the simulations
d 5
1012 s1
sp 6
1012 s
T 300 K
I 1:5
1015 W=m2
f 1
rf 0:99
L 25
109 m

temperature variation along
the x-axis for three materials
and three heating periods
Figure 4.14 shows dimensionless lattice temperature inside the nano-wire for
different dimensionless time. The x-axis location is at the center of the wire.
Temperature decay at the surface vicinity of the wire is gradual and it becomes
sharp in the region next to the wurface vicinity. Since the volumetric heat source is
considered in the analysis, energy absorbed by the substrate material within the
absorption depth is responsible for the gradual decay of temperature in the surface
vicinity. In this case, energy absorbed from the irradiated field increases internal
energy gain of the material while increasing temperature in the irradiated region.
Since, the absorption takes place according to the Lamberts Beer law, absorbed

temperature variation along
the y-axis for three materials
and three heating periods
power reduces with increasing depth below the surface. This, in turn, gives rise to
sharp decay of temperature in the region next to the surface vicinity. Since the
laser pulse intensity reduces with time (exponential pulse), temperature inside the
nano-wire reduces due to low energy being absorbed in the nano-wire with pro-
gressing time. As the depth below the surface increases, temperature decay
becomes gradual. It should be noted that increasing temperature gradient enhances
the diffusional energy transport from the surface region to the solid bulk. However,
this may not occur substantially during the short time period. The maximum
temperature is higher for silver, then, follows chromium, and copper.
Fig. 4.15 Temporal

surface temperature variation
for three materials. The x, y-
axes locations are at x = 0
and y = 0
Figure 4.15 shows temporal variation of dimensionless surface temperature

distribution at the nano-wire center for three different substrate materials. The rate
of temperature decay follows almost the rate of laser pulse intensity decay, par-
ticularly for silver. This is associated with the short-heating period. Electrons gain
energy from the irradiated field and transfer their excess energy to the lattice site
through the collisional process. Consequently, depending on the rate of excess
energy gain by electrons from the irradiated field and number of collisions during
the short-period of heating, lattice site temperature increases accordingly.
Although, the number of collisions between electrons and lattice site increases
with progressing time, energy absorbed by electrons from the irradiated field
reduces-because of time decay of the laser pulse, which is in the exponential decay
form. This, in turn, does not affect notably the rate of lattice site temperature decay
with progressing time. Temperature attains high values for silver, then, follows
chromium, and copper.
References
1. B.S. Yilbas, Improved formulation of electron kinetic theory approach for laser short-pulse
heating. Int. J. Heat Mass Transf. 49(1314), 22272238 (2006)
2. B.S. Yilbas, A.Y. Al-Dweik, Non-equilibrium heating and thermal stress development.
J. Thermophys. Heat Transf. (2012) (in print)
3. B.S. Yilbas, A.Y. Al-Dweik, Closed form solutions for thermal stress field due to non-
equilibrium heating during laser short-pulse irradiation. Physica B 407(12), 21692175
(2012)
4. B.S. Yilbas, A.Y. Al-Dweik, Laser short pulse heating of metal nano-wires. Physica B (2012)
(in print)
5. B.S. Yilbas, S.Z. Shuja, Laser short-pulse heating of surfaces. J. Phys. D Appl. Phys. 32,
19471954 (1999)
6. C.L. Tien, J.H. Lienhard, Statistical Thermodynamics (Hemisphere, Washington DC, 1979)
References 165
7. B.S. Yilbas, A.F.M. Arif, Material response to thermal loading due to short pulse laser
heating. Int. J. Heat Mass Transf. 44, 37873798 (2001)
8. H.E. Elsayed-Ali, M.A. Norris, M.A. Pessot, G.A. Mourou, Time-resolved observation of
electron-phonon relaxation in copper. Phys. Rev. Lett. 58, 12121215 (1987)
9. G. Chen, Ballistic-diffusive heat-conduction equations. Phys. Rev. Lett. 86, 22972300
(2001)
10. G.L. Eesley, Generation of nonequilibrium electron and lattice temperatures in copper by
picosecond laser pulses. Phys. Rev. B 33, 21442151 (1986)
11. L.B. Loeb, The Kinetic Theory of Gases (Dover, New York, 1961)
12. M.I. Kaganov, I.M. Lifshiftz, L.V. Tanatarov, Relaxation between electrons and crystalline
lattice. Soviet Phys. JETP 4, 173178 (1957)
13. S.D. Brorson, A. Kazeroonian, J.S. Moodera, D.W. Face, T.K. Cheng, E.P. Ippen, M.S.
Dresselhaus, G. Dresselhaus G., Femtosecond room-temperature measurement of the
electron-phonon coupling constant k in metallic superconductors. Phys. Rev. Lett. 64,
21722175 (1990)
14. M. Honner, J. Kunes, On the wave diffusion and parallel nonequilibrium heat conduction.
ASME J. Heat Transf. 121, 702707 (1999)
15. T.Q. Qiu, L. Tien, Femtosecond laser heating of multi-layer metals-I analysis. Int. J. Heat
Mass Transf. 37, 27892797 (1994)
16. S.D. Brorson, J.G. Fujimoto, E.P. Ippen, Femtosecond electron heat-transport dynamics in
thin gold film. Phys. Rev. Lett. 59, 19621965 (1987)
17. B.S. Yilbas, A.Y. Al-Dweik, Exact solution for temperature field due to non-equilibrium
heating of solid substrate. Physica B 406(23), 45234528 (2011)
18. A. Kovalenko, Thermoelasticity (Basic Theory and Applications) (Wolters-Noordhoff,
Publishing, Groningen, 1969)
19. L. Debnath, D. Bhatta, Integral Transforms and Their Applications, 2nd edn. (Chapman &
Hall/CRC, New York, 2007)
20. H. Bateman, Tables of Integral Transforms, vol. 1 (McGraw-Hill, New York, 1954)
21. L.F. Shampine, Vectorized adaptive quadrature in MATLAB. J. Comput. Appl. Math. 211,
131140 (2008)
Chapter 5
Concluding Remarks
Abstract Thermal stress developed in the laser irradiated material is governed by

laser and workpiece materials. The main laser parameters include the pulse length
and the laser power intensity while the important material properties are the
thermal conductivity, absorption depth, elastic modules, thermal expansion coef-
ficient, and Poissons ratio. Depending upon the duration of the laser pulse, the
heating process can be classified into two categories, which are equilibrium and
non-equilibrium heating situations. In the case of equilibrium heating, duration is
longer than the thermalization time of the substrate material and Fourier heating
law governs the heating process. However, laser pulse durations comparable and
shorter than the thermalization time of the substrate material, wave behavior of
stress field takes place and the hyperbolic nature of the governing equations are
used to account for the wave behavior. Consequently, care must be taken to
formulate the heating and thermal stress problems in line with the physical aspects
and scale of the problem. The conclusions derived from the body of this book are
presented according to the following sub-headings and in line with the previous
studies [118].
5.1 Equilibrium Heating
The maximum laser pulse duration is limited with the thermalization time of the
substrate material for the equilibrium heating situation. In this case, volumetric
and surface heat sources can be incorporated in the analysis to resemble the laser
heating process. Since the laser pulse profile has significant effect on the thermal
stress field, step input and exponential lasers pulses should be considered to
identify this effect. In addition, insulated and convection boundary conditions at
the surface together with stress free surface and stress continuity at the surface are
important concerns influencing the thermal stress behavior in the irradiated region.
Therefore, in line with the previous studies [18], the followings are concluded.

168 5 Concluding Remarks
5.1.1 Step Input Laser Pulse Intensity
In the analysis, stress free surface and zero stress gradient at the surface conditions
are considered. Temperature and stress fields are computed for two cases sepa-
rately. In general, it is observed that stress level increases considerably at some
depth below the surface. This is more pronounced as the heating period progresses.
The behavior of thermal stress inside the substrate material differs considerably as
the boundary condition at the surface changes, i.e. stress free surface and zero
stress gradient at the surface conditions result in different stress behavior inside the
substrate material. In the case of stress free surface condition, the stress wave
propagates towards the solid bulk of the substrate material, since at the surface
stress is kept at zero. The stress level increases as it propagates inside the substrate
material and the stress level enhances as the heating period progresses. In the case
of zero stress gradient at the surface condition, stress is compressive in the surface
vicinity of the substrate material and it becomes tensile as the distance from the
surface increases. The stress level reaches maximum at some depth below the
surface. The magnitude of stress reduces sharply to almost zero as the depth inside
the substrate material increases further. The point of sharp decay of stress level
varies with heating time while the location of maximum stress inside the substrate
material remains almost constant. Temporal behavior of the thermal stress inside
the substrate material differs considerably as the boundary condition at the surface
changes. In the case of stress free surface, stress peaks occur inside the substrate
material as the stress wave propagates. The stress level attains low values in the
early heating period for the zero stress gradient at the surface. However, as the
heating period progresses, stress level increases rapidly, provided that the mag-
nitude of stress level exceeds its counterpart corresponding to stress free condition.
Consequently, the boundary condition at the surface modifies the stress wave
propagation and its level inside the substrate material. Temporal variation of
maximum stress inside the substrate material differs for two boundary conditions
employed in the present study. In the case of zero stress gradient at the surface
condition, maximum stress attains low values in the early heating period and it
increases rapidly as the heating period progresses. The maximum stress follows
almost an exponential increase with time. However, maximum stress behavior is
almost parabolic with time for stress free surface condition. In this case, maximum
stress increases rapidly in the early heating period and the rate of rise of maximum
stress becomes less as the heating period progresses.
5.1.2 Exponential Laser Pulse Intensity
Volumetric heat source is considered to resemble the absorption of the laser pulse
during the heating process. The conclusions for various boundary conditions are
given below.
5.1 Equilibrium Heating 169
5.1.2.1 Stress Free Surface Consideration
It is found that the stress wave propagates into the substrate material, and the stress
level behind the wave front is tensile and after the wave front it is compressive.
The magnitude of maximum stress level varies inside the substrate material.
Temperature rises rapidly in the early heating period and the rate of temperature
rise reduces as the heating period progresses. In this case, internal energy gain
dominates over the conduction losses due to high temperature gradient. Thermal
stress wave propagates into the substrate material at a speed of c1. The magnitude
of stress wave reduces at different locations inside the substrate material as the
heating period progresses. This is because of the magnitude of temperature gra-
dient corresponding to different heating periods, since pulse intensity varies
exponentially with time.
5.1.2.2 Exponential Stress Distribution at the Surface
In order to accommodate the temporal variation of recoil pressure generated

during the surface ablation, time exponential stress distribution is considered at the
free surface of the solid substrate. The stress free boundary condition results in low
magnitude of stress waves and the stress boundary at the surface modifies the
shape of the stress wave developed inside the substrate material. In this case, the
stress curve becomes similar to that introduced at the surface. The magnitude of
stress wave reduces as the depth below the surface increases towards the solid
bulk. The time corresponding to the stress peak inside the substrate material differs
for stress free boundary and stress boundary at the surface cases. This is because of
the temporal distributions of the laser pulse intensity and stress at the surface,
which differ considerably.
5.1.2.3 Stress Continuity at the Surface
The zero stress gradient at the surface is considered to incorporate the stress
developed during the laser heating of surfaces initially coated or interfaced with
other substrate surface. Temperature rise due to internal energy gain is important
in the early heating period and as the heating period progresses, conduction energy
transfer from the surface vicinity results in gradual temperature rise inside the
substrate material. This is due to the temporal behavior of the pulse power, which
reduces exponentially with progressing heating period. Stress level in the surface
region attains considerably high values. The stress wave is compressive in nature
and propagates with a wave speed c1. Moreover, the rise of the stress wave is
modified by the stress boundary condition at the surface.
5.1.2.4 Convection Boundary Condition at the Surface
A convection boundary condition at the surface is considered to incorporate the

cooling effect of the assisting gas during the heating process. The influence of the
heat transfer coefficient on temperature profiles is significant as the dimensionless
heat transfer coefficient at the surface increases to 0.0202. The temperature gra-
dient 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 coef-
ficient of h 0:0202. Moreover, the heat transfer coefficient influences the
temperature gradient in the surface region, which is significant for h 0:0202.
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 for a heat transfer
coefficient of 0.0202. The magnitude of stress wave is reduced as the depth
increases from the surface towards the bulk solid.
5.1.2.5 Entropy Analysis
Entropy generation due to temperature field decays sharply in the surface region,
which is due to the behavior of temperature gradient and energy storage in this
region. The location of minimum entropy generation due to temperature field
inside the substrate material is in the same region of the temperature equilibrium as
described earlier. The location of minimum entropy generation moves away from
the surface as the heating period progresses. Entropy generation due to stress field
shows a cyclic behavior provided that the maximum entropy occurs at the location
where the stress is the maximum inside the substrate material. The cyclic behavior
of the entropy is because of the propagation of the stress waves. Entropy gener-
ation due to temperature field dominates over its counterpart corresponding to the
stress field in the surface region. As the depth from the surface increases towards
the solid bulk, entropy generation due to stress field becomes important; in which
case, entropy peaks appear on the entropy curve. In the case of stress field, the
negative sign of the entropy generation is because of the compression component
of the stress waves and, in all cases, the entropy generation is positive.
5.2 Cattaneo Heating Model and Thermal Stresses
The conclusions derived from the solution of Cattaneo heat equation and the
thermal stress are given below in line with the previous studies [913].
5.2 Cattaneo Heating Model and Thermal Stresses 171
5.2.1 Exponential Laser Pulse Intensity
5.2.1.1 Surface Heat Source Consideration
It is observed that temperature rise in the early heating period is gradual. As the
heating period progresses, temperature rises rapidly reaching its maximum, which
is more pronounced at the surface. The time occurrence of the peak temperature
changes as the distance below the surface increases. Temperature decay rate
becomes small when the pulse intensity reduces 5 % of its peak value. In this case,
temporal gradient of temperature becomes small and the wave nature of the
heating replaces with the diffusional heating. Temperature decay is sharp inside
the substrate material during the heating cycle. This results in large temperature
gradients and stress field below the surface. Temporal behavior of thermal stress
reveals that the compressive stress waves are formed due to initial contraction of
the surface during the early heating period. The compression wave reaches its peak
value rapidly and decays gradually similar to the pulse intensity distribution. The
stress wave generated propagates at a constant speed inside the substrate material.
The thermal expansion of the surface during the late heating period results in the
tensile wave formation in the surface region. This appears as a tensile tail in the
compressive wave generated earlier. The magnitude of the tensile wave is sig-
nificantly lower than that of the compressive wave.
5.2.1.2 Volumetric Heat Source Consideration
Temperature decay is sharp inside the substrate material during the heating cycle.
This results in large temperature gradients and stress field below the surface.
Temporal behavior of thermal stress reveals that the compressive stress waves are
formed due to initial contraction of the surface during the early heating period. The
compression wave reaches its peak value rapidly and decays gradually similar to the
pulse intensity distribution. The stress wave generated propagates at a constant
speed inside the substrate material. The thermal expansion of the surface during the
late heating period results in the tensile wave formation in the surface region. This
appears as a tensile tail in the compressive wave generated earlier. The magnitude
of the tensile wave is significantly lower than that of the compressive wave.
5.2.2 Step Input Laser Pulse Intensity
5.2.2.1 Surface Heat Source Consideration
The high rate of increase in temperature is attributed to the internal energy gain of
the substrate material from the heat source and small amount of energy transfer
by diffusion from the surface vicinity to the solid bulk in the early heating period.
In this case, qT/qt, and q2T/qt2 attain high values and varies sharply with progressing
time. Consequently, the heat diffusion is not governed by the classical Fourier law in
the early heating period. This situation is also observed onset of the initiation of the
cooling period for which temperature decays rapidly with progressing time. How-
ever, as the heating progresses, the rate of temperature rise becomes gradual and the
term q2T/qt2 becomes small in Cattaneo equation. The heat diffusion is governed
mainly by the Fourier law. This is also true during the long cooling periods. Thermal
stress developed in the surface vicinity is in the form of a stress wave which
propagates into the substrate material with a constant speed. Moreover, the stress
wave is tensile in the heating cycle due to the thermal expansion of the surface while
it is compressive in the cooling cycle because of the thermal contraction of the
surface region during the cooling period. The thermal wave generated has a tail with
decaying amplitude and it extends over the heated region.
5.2.2.2 Volumetric Heat Source Consideration
Temperature and stress distribution are demonstrated for a step input laser short-
pulse heating of metallic surfaces. It is found that the temperature rise towards the
laser mid-pulse-length is rapid, which results in large temporal gradient of tem-
o2 T
perature (oT
ot and ot2 ). In this case, wave nature of heat transfer dominates and
governs the heat transfer in the substrate material. This is true during the heating
and in the early cooling periods, except at some depth below the surface; in which
case, temporal variation of temperature is not significantly high and resulting in
diffusional heat conduction. This situation is also observed during the late cooling
periods, i.e. wave nature of the heat transfer is replaced by the diffusional con-
duction governed by Fourier law. Thermal stress generated during the heating
period is compressive and the stress wave amplitude is influenced by the volu-
metric heat source in the early heating period. However, stress wave becomes
tensile in the cooling period. This is attributed to the expansion of the irradiated
surface during the cooling period.
5.2.2.3 Entropy Analysis
Entropy generation rate due to heat transfer and thermal stress field is computed
for the time exponentially varying laser short-pulse. The volumetric heat source
resembling the laser irradiated energy is incorporated in the analysis. Stress free
boundary condition is considered to account for the free expansion of the surface
during the laser heating pulse in the analytical solution. It is observed that tem-
perature gradient remains low in the surface vicinity because of the internal energy
gain of the substrate material during the heating pulse. The maximum value of
thermal stress increases with time, which is attributed to the formation of high
temperature gradient with progressing time. Thermal stress is tensile in the surface
5.2 Cattaneo Heating Model and Thermal Stresses 173
region and becomes compressive with increasing depth below the surface. This is
associated with free expansion of the surface. As the heating progresses, the depth
of tensile stress extends inside the substrate material. Entropy generation rate,
during the heating pulse attains low values in the surface region and increases with
increasing depth below the surface, which is true for all heating periods. The
attainment of low entropy generation rate is associated with the term 1/T in the
entropy equation, which is low in the surface region. Entropy generation rate due
to heat transfer is significantly higher than that of due to thermal stress developed
for all heating periods.
5.3 Non-Equilibrium Heating
Non-equilibrium energy transfer takes place during the laser-short pulse heating
process. Thermal separation results in thermally communicating two sub-systems
co-existed in the irradiated region. Although the temperature gradient is low in the
lattice sub-system, high thermal stress field is developed due to short-heating
duration. The findings due to non-equilibrium heating situation are given below in
line with the previous findings [1418] under the appropriate sub-headings.
Two boundary conditions are introduced for thermal stress analysis. These include
stress free boundary and stress continuity at the surface conditions. Stress free
boundary condition assumes free expansions of the surface during the heating
process while stress continuity boundary condition incorporates the presence of
optically transparent coating at the surface. It is found that lattice temperature
increases along the x-axis because of electron excess energy transfer through the
collisional process. The temperature gradient remains high in the surface region
during the initial heating period, which in turn results in high magnitude of thermal
stress wave generation in the surface region. The thermal stress wave behavior
differs significantly for different stress boundary conditions at the surface. Stress
free boundary condition causes free expansion of the surface resulting in tensile
thermal stress in the surface region. As the distance from the surface increases,
tensile behavior replaces with compressive behavior. The maximum magnitude of
compressive stress increases with increasing depth below the surface. In the case
of the stress continuity boundary condition, the amplitude of thermal stress
remains high in the surface region and thermal stress becomes compressive inside
substrate material for all heating periods. Stress wave decays gradually with
progressing distance along the x-axis towards the solid bulk.
Two boundary conditions for thermal stress were introduced, namely stress free
surface and stress continuity at the surface. It is found that lattice site temperature
rise is sharp in the early heating period and it becomes gradual as the heating
period progresses. This is attributed to the collisional energy transfer from elec-
trons to lattice site during the heating period. Thermal stress developed in the
surface region is tensile while it becomes compressive at same depth below the
surface for stress free boundary condition. This is attributed to free expansion of
the surface. Thermal stress wave propagates into the substrate material at a con-
stant speed and thermal stress behavior in the surface region is modified due to
stress continuity condition at the surface; in which case, the peak stress reduces
with progressing time. In the case of stress continuity at the surface, thermal stress
becomes compressive inside the substrate material for all heating periods.
5.3.3 Two-Dimensional Heating
Laser short-pulse heating of metal nano-wire is considered and the analytical

solution of temperature field in the irradiated region is formulated after considering
two-dimensional axisymmetric heating situation. It is found that temperature decay
inside the substrate material is gradual in the surface vicinity, which is associated
with the internal energy gain from the irradiated field in this region. Temperature
decay is sharp in the region next to the surface vicinity. Although high temperature
gradient enhances heat diffusion from surface region to the solid bulk, this is not
substantiated due to the short heating duration. Temperature decays sharply towards
the edge of the nano-wire because of the low temperature boundary condition at the
wire edges. Temporal variation of temperature at the irradiated spot center follows
almost the laser pulse intensity variation with time. This is attributed to energy gain
by electrons from the irradiated field; in which case, electrons transfer their excess
energy to lattice site through the collisional process.
References
1. M. Kalyon, B.S. Yilbas, Analytical solution for thermal stresses during laser pulse heating
process. Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 215, 14291445 (2001)
2. B.S. Yilbas, M. Kalyon, Repetitive laser pulse heating with a convective boundary condition
at the surface. J. Phys. D Appl. Phys. 34, 222231 (2001)
3. B.S. Yilbas, N. Ageeli, Thermal stress development due to laser step input pulse intensity
heating. J. Therm. Stresses 29(8), 721751 (2006)
4. B.S. Yilbas, N. Ageeli, M. Kalyon, Laser induced thermal stresses in solids: Exponentially
time decaying pulse case. Lasers Eng. 14(1), 81101 (2004)
References 175
5. B.S. Yilbas, N. Al-Ageeli, Formulation of laser induced thermal stresses: Stress boundary at
the surface. Proc. Inst. Mech. Eng. Part C J. Mech. Eng. Sci. 217, 423434 (2003)
6. B.S. Yilbas, N. Ageeli, Thermal stresses due to exponentially decaying laser pulse and a
convection boundary at the surface. Lasers Eng. 16, 235265 (2006)
7. B.S. Yilbas, M. Kalyon, Analytical approach for entropy generation during a laser pulse
heating process. AIChE J. 52, 19411950 (2006)
8. B.S. Yilbas, Entropy analysis due to temperature and stress fields in the solid irradiated by a
time exponentially varying laser pulse. Heat Transf. Eng. J. 26(8), 8089 (2005)
9. H. Al-Qahtani, B.S. Yilbas, The closed form solutions for Cattaneo and stress equations due
to step input pulse heating. Phys. B 405(18), 38693874 (2010)
10. H. Al-Qahtani, B.S. Yilbas, Temperature and stress fields for short pulse heating of solids.
J. Thermophys. Heat Transf. 25(1), 173176 (2011)
11. H. Al-Qahtani, B.S. Yilbas, Closed form solution of Cattaneo equation including volumetric
source in relation to laser short-pulse heating. Can. J. Phys. 89(7), 761767 (2011)
12. H. Al-Qahtani, A closed form solution of temperature and stress fields for laser short-pulse
heating of a solid: Exponentially decaying volumetric source. Lasers Eng. 22(12), 109124
(2011)
13. H. Al-Qahtani, B.S. Yilbas, Entropy generation rate during laser short-pulse heating:
Contribution of heat transfer and thermal stress. Lasers Eng., in print (2012)
14. B.S. Yilbas, Improved formulation of electron kinetic theory approach for laser short-pulse
heating. Int. J. Heat Mass Transf. 49(1314), 22272238 (2006)
15. B.S. Yilbas, A.Y. Al-Dweik, Non-equilibrium heating and thermal stress development.
J. Thermophys. Heat Transf., in print (2012)
16. B.S. Yilbas, A.Y. Al-Dweik, Closed form solutions for thermal stress field due to non-
equilibrium heating during laser short-pulse irradiation. Phys. B 407(12), 21692175 (2012)
17. B.S. Yilbas, A.Y. Al-Dweik, Laser short pulse heating of metal nano-wires. Phys. B, in print
(2012)
18. B.S. Yilbas, S.Z. Shuja, Laser short-pulse heating of surfaces. J. Phys. D Appl. Phys. 32,
19471954 (1999)
Index
A E
Adaptive Gauss-Kronrod quadrature, 151 Elastic module, 66
Analytical solution, Cattaneo equation. See Electron-phonon collision, 122, 126, 132, 134
Cattaneos equation Energy transport, microscopic level, 122
Analytical treatment, hyperbolic equations. conservation of energy, 127, 128
See Energy transport, microscopic level electron distribution, 123, 124
electron excess energy transfer, 126
electron kinetic theory approach, 122
B electron mean free path, 134
Beer Lamberts law, 68, 74, 95, 134, 153 electron movement in surface
Boltzmanns constant, 134 region, 123f
electron phonon coupling
factor, 133, 134
C electrons, collision, 125
Cattaneo heat equation, 82, 96 electrons, without collision, 125
See also Cattaneos equation, 000 energy balance, 126
Cattaneo heating model and thermal Fourier integral transformation, 129
stresses, 81, 170 Fourier inversion, 130
exponential laser pulse intensity, 171 Fourier transformation, 130
step input laser pulse intensity, 171, 172 kinetic theory approach, 132
Cattaneos equation, 81, 82 rectangle function, 124
surface heat source consideration, 8695, simple kinetic theory, 128
110114 total electron energy, 127
volumetric source consideration, 95110, Entropy analysis, thermal stress field, 6467,
114119 108110
dimensionless entropy generation, 80f, 81f,
82f, 83f, 84f
D dimensionless stress distribution, 80f
Debye temperature, 134 dimensionless temperature gradient, 81f
Dimensionless form, 4850, 50, 51, 5153, 53, dimensionless thermal stress, 83f
54 elastic module, 66
Dimensionless stress field equation, 89, 94 entropy generation, 64, 65
Dimensionless temperature distribution, 8, 28, Gouy-Stodola theorem, 66
38, 56, 73, 75f, 78, 79f, 116f, 160, 162f non-dimensional form, 66
Dirac delta function, 48, 131 Poissons ratio, 66

DOI: 10.1007/978-3-319-00086-2, Springer International Publishing Switzerland 2014
178 Index
Entropy analysis, thermal stress field (cont.) L

step input pulse, findings, 7981 Lamberts law, 72, 79, 162
strain, in terms of displacement, 66 Laplace transformation, 7, 9, 19, 21, 27, 29,
thermal expansion coefficient, 66 36, 38, 39, 55, 60, 104
time exponentially decaying pulse, find- See also Inverse Laplace transform
ings, 8284 boundary conditions, 87, 97
Equilibrium heating, 167 respect to time, 9, 38, 39, 57
exponential laser pulse intensity, 168170 temperature, 9, 21, 57
step input laser pulse intensity, 168 of Term1, 41
Exponential laser pulse intensity, 168, 171 of Term2, 44
convection boundary condition at surface, of Term22, 46
170 thermal stress, 9, 21, 38, 39, 57
entropy analysis, 170 t-shifting rule of, 107
exponential stress distribution at surface, Laser heating process, 13, 86, 108
169 conduction heating, 2
stress continuity at surface, 169 Fourier heating law, 2
stress free surface consideration, 169 one-dimensional modeling, 6
surface heat source considerations, 171 short-pulse heating, 3
volumetric heat source considerations, 171 solid state heating, 1
Lasers, 1
F
Fourier heat transfer equation, 7, 26, 36, 54 M
Fourier heating law, 1, 5, 167, 170, 172 Metallic substrates at microscopic level, for-
Fourier heating model, 85, 117 mulation of energy transport. See
Fourier cosine transform, 137, 139, 140, 145, Energy transport, microscopic level
150, 152 Modified Bessel functions, 88, 93, 97, 109,
Fourier sine transform, 137, 138, 140 150, 151
Fourier transformation, 130
N
G Non-equilibrium energy transport, 2, 3, 121
Gouy-Stodola theorem, 66 Non-equilibrium heating, 3, 85, 173
surface heat source consideration, 173
two-dimensional heating, 174
H volumetric heat source consideration, 174
Hyperbolic equations. See Energy transport,
microscopic level
P
Parabolic heating model, 131134
I Partial differential equation (PDE)
Inverse Fourier cosine transform, 140, 146, second order, 149, 152
151, 153 third order, 132
Inverse Fourier sine transform Partial fraction, 11, 12, 32, 36, 60, 98, 100,
Inverse Laplace transform, 11, 20, 55, 56 105, 107
of Term1, 11, 12, 24, 33, 4044, 60, 61 Term1, 24
of Term2, 12, 13, 24, 33, 4447, 60, 61 Term2, 12
of Term3, 13, 14, 25, 33, 47, 60, 61 Term3, 14, 25
of Term4, 15, 33, 47, 48, 60, 62 Term4, 15
of Term5, 60, 62 Poissons ratio, 9, 66, 136, 141, 167
K Q
Kinetic theory, 2, 121, 122, 128, 132 Quasi-ballistic approach, 131
Index 179
S Stress free boundary and convection at surface,

Short-pulse laser heating, 81 time exponentially varying laser pulse
Step input laser pulse heating, 6 heating, 3654
stress continuity boundary at surface, closed form solution of stress distribution,
1926 48
stress free boundary at surface, 618 dimensionless form, 4850, 50, 51, 5153,
Step input laser pulse intensity, 168 53, 54
entropy analysis, 172, 173 dimensionless stress distribution, 77f, 78f
surface heat source consideration, 171, 172 dimensionless temperature distribution, 38,
volumetric heat source consideration, 172 75f
Step input pulse intensity, 88, 96, 114 dimensionless temperature gradients, 76f
Stress boundary at surface, time exponentially findings, 7378
varying laser pulse heating, 5464 Fourier heat transfer equation, 36
boundary conditions, 54 general solution for stress field, 40
closed form solution of stress inverse Laplace transform, 37, 40, 41
distribution, 62 Laplace inversions of composing terms, 43,
complementary error function, 55, 56 44, 4446, 4648
dimensionless stress distribution, 6264 Laplace transformation, 36, 37
dimensionless temperature distributions, Laplace transformations of terms, 4143
79f Laplace transformation to time, 38, 39
findings, 78, 79 laser pulse in simulations, 75f
general solution for stress field, 58, 59 solving momentum equation, 38
initial conditions, 55 solving stress distribution, 38
inverse Laplace transform, 55 temperature distribution in Laplace
Laplace inversions of terms, 61, 62 domain, 39
Laplace transformation of boundary Stress free boundary at surface, step input laser
condition, 59, 60 pulse heating, 618
Laplace transformation to time, 56, 57 complementary error function, 8
partial fraction, 60 dimensionless laser step input pulses, 69f
solving stress distribution, 56 dimensionless temperature profiles, 69f,
solving stress equation, 56 70f
stress distribution in dimensionless form, dimensionless thermal stress, 70f
62 findings, 6770
stress due to recoil pressure, 56 inverse Laplace transform, 8, 9
Stress continuity boundary at surface, step inverse of Laplace transform of terms,
input laser pulse heating, 1926 1015
dimensionless stress distribution, 71f laser heating pulse, 6
dimensionless stress, temporal variation, laser pulse properties in simulations, 69t
72f pulses in analysis and simulations, 68f
findings, 71 stress distribution, dimensionless form,
heat transfer equation, 19 1618
initial and boundary conditions, 19, 20 stress equation, closed form, 15
inverse Laplace transformation, 20, 21 Stress free boundary at surface, time expo-
inversion of Laplace nentially varying laser pulse heating,
transformation, 2426 2736
Laplace transformation, 19, 20 boundary conditions for temperature, 29
new boundary conditions, 22, 23 closed form solution of stress distribution,
new initial conditions, 21, 22 34
zero stress gradient, 21 closed form solution, Laplace transforma-
Stress distribution, 11, 32, 38 tion method, 27
closed form, 15, 16, 34, 48 complementary error function, 28
dimensionless form, 1618, 26, 28, 34, 48, dimensionless temperature, temporal vari-
54, 62, 71f, 74f, 77f, 78f, 80f, 111, 154, ation, 72f, 74f
156f, 157, 159f, 160f, 161f findings, 7173
180 Index
Stress free boundary at surface, time expo- Thermal stress. See Cattaneo heating model
nentially varying laser pulse heating (cont.) and thermal stresses; Energy transport,
initial conditions for temperature, 29, 30 microscopic level; Entropy analysis,
inverse Laplace transform, 27, 28, 3234 thermal stress field
normalized power intensity distribution, Thermoelasticity, 94
73f Time exponentially varying laser pulse heat-
partial fraction, 32 ing, 26
stress distribution in dimensionless form, stress boundary at surface, 5464
3436 stress free boundary and convection at
stress field, general solution for, 30, 31 surface, 3654
Stress free substrate material, 56, 57 stress free boundary at surface, 2736
Surface heat source consideration, Cattaneos
equation, 86
convolution theorem, 95 V
dimensionless stress distribution, 112f Volumetric source consideration, Cattaneos
dimensionless stress field equation, 94 equation, 95, 96
dimensionless stress versus time, 111f convolution theorem, 102
exponential pulse heating, 9295, 112114 dimensionless entropy generation, 118f
exponential pulse, temporal variation, 93f dimensionless temperature distribution
findings, 110114 versus time, 116f
inverse Laplace transform, 95 dimensionless temperature, temporal vari-
step input pulse heating, 8792, 111 ation, 114f
temperature distribution, 93 dimensionless thermal stress, temporal
variation, 115f
dimensionless thermal stresses distribution
T with time, 117f
Thermal expansion coefficient, 5, 65, 136, 167 entropy generation rate, 117119
Thermal strain, 2, 5, 9, 66, 67, 74, 112, 170 exponential pulse heating, 103108, 116,
Thermal stress field 117
entropy analysis, 6467 Laplace inversion, 97
stress continuity boundary condition at laser step input pulse, temporal variation,
surface, 145147, 156158 96f
stress gradient-free boundary condition, modified Bessel function, 97
139, 140 space-time domain, 97
stress-free boundary condition, 137139, step input pulse heating, 96103, 114, 115
142145, 154156 temperature distribution, 104
surface heat source considerations,
134140, 154158
surface source, 151153 Y
two-dimensional considerations, 147149, Youngs modulus, 65
160164
volumetric heat source considerations,
140147, 158160 Z
volumetric source, 148151 Zero stress gradient, 19, 21, 168, 169

Laeser Pulse Heating of Surfaces and Thermal Stress Analysis - BYilbas

Enviado por

Dados do documento

Título original

Direitos autorais

Formatos disponíveis

Compartilhar este documento

Compartilhar ou incorporar documento

Opções de compartilhamento

Você considera este documento útil?

Este conteúdo é inapropriado?

Direitos autorais:

Formatos disponíveis

Laeser Pulse Heating of Surfaces and Thermal Stress Analysis - BYilbas

Enviado por

Direitos autorais:

Formatos disponíveis

Materials Forming, Machining and Tribology

For further volumes:

Nasser Al-Aqeeli Hussain M. Al-Qahtani

Laser Pulse Heating

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

Library of Congress Control Number: 2013940306

Springer International Publishing Switzerland 2014

Printed on acid-free paper

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

surface and volumetric heat sources. In addition, equilibrium and non-equilibrium

2 Equilibrium Laser Pulse Heating and Thermal Stress Analysis . . . 5

3 Analytical Solution of Cattaneo and Thermal Stress Equations . . . 85

4 Analytical Treatment of Hyperbolic Equations

5 Concluding Remarks . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . . 167

Abstract Laser heating offers considerable advantages over the conventional

B. S. Yilbas et al., Laser Pulse Heating of Surfaces 1

categorized as conduction limited heating. The conduction heating is, in general,

operations. However, the evaporation of the surface results in recoil pressure

B. S. Yilbas et al., Laser Pulse Heating of Surfaces 5

2.2 Step Input Laser Pulse Heating

2.2.1 Stress Free Boundary at the Surface

The mathematical arrangement of temperature rise in the solid substrate due to

which gives the thermal solution in the Laplace transform domain.

where erfc is the complementary error function, which is:

Equation 2.15 is the closed form solution for temperature distribution.

where r s are the Laplace transforms of thermal stress and tem-

Considering the temperature distribution for a pulse with exponential temporal

The second part of the particular solution is:

Solving for the particular solutions yields:

Form the boundary condition (x = ? ) rx = 0), it yields D1 = 0:

Term1 G1 Term11 G1 Term21 G1 Term31 2:36

1 Term1 1 Term11 1 Term21 1 Term31 2:38

The inverse of Laplace transform of terms becomes:

The inverse of Laplace transform of Term 2 comes out to be:

1 Term2 1 G2 Term12 1 G2 Term22 1 G2 Term32 2:46

c21 aedx c21 aedx

1 Term3 1 G3 Term13 1 G3 Term23 1 G3 Term33 2:54

c21 aedx h c1 dtDt i

c21 a2 edx h ad2 t ad2 tDt

Term4 can be written as:

1 Term4 1 G4 Term14 1 G4 Term24 1 G4 Term34 2:62

rx x; t 1 Term1 1 Term2 1 Term3 1 Term4 2:66

Presenting the stress distribution in dimensionless form, the additional

c2 ex h c t i

Consequently, the dimensionless form of stress equation becomes:

2.2.2 Stress Continuity Boundary at the Surface

The boundary condition, T 0 at x 1, results A 0 in Eq. 2.91. Therefore,

The Inverse Laplace Transformation of Eq. 2.95 results:

At t 0 !; for thestress equation, rx 0

However, it is noted from Eq. 2.94 that T is:

Substitution of Eq. 2.104 into Eq. 2.103 yields:

If we define M1 and M2 as:

then Eq. 2.106 becomes:

The homogenous solution is:

Substituting Eq. 2.111 into Eq. 2.103 yields:

The second part of the particular solution is:

Substituting Eq. 2.113 into Eq. 2.108 yields:

Therefore, Eq. 2.109 becomes:

Let us define the constants C10 and C20 as:

Let us introduce the followings:

The inversion of Laplace transformation of Term1 yields:

Similarly, the inversion of Laplace transformation of Term2 yields:

The inversion of Laplace transformation of Term3 yields:

Term1 G1 Term11 G1 Term21 G1 Term31 2:36

1 Term1 1 Term11 1 Term21 1 Term31 2:38

1 Term2 1 G2 Term12 1 G2 Term22 1 G2 Term32 2:46

1 Term3 1 G3 Term13 1 G3 Term23 1 G3 Term33 2:54

1 Term4 1 G4 Term14 1 G4 Term24 1 G4 Term34 2:62

rx x; t 1 Term1 1 Term2 1 Term3 1 Term4 2:66

rx x; t 1 Term1 1 Term2 1 Term3 1 Term4 2:187

1 Term1 1 Term11 1 Term21 2:230

1 Term11 1 Term111 1 Term211 1 Term311

1 Term21 1 Term121 1 Term221

1 Term2 1 Term12 1 Term22 2:245

1 Term12 1 Term112 1 Term212 1 Term312

1 Term22 1 Term122 1 Term222

rx x; t 1 Term1 1 Term2 1 Term3 1 Term4 2:268

rx x; t 1 Term1 1 Term2 1 Term3 1 Term4