Optics & Laser Technology 49 (2013) 125136

Contents lists available at SciVerse ScienceDirect

Optics & Laser Technology

journal homepage: www.elsevier.com/locate/optlastec

Laser beam welding of dissimilar ferritic/martensitic stainless steels

in a butt joint conguration
M.M.A. Khan n, L. Romoli, G. Dini
Department of Mechanical, Nuclear and Production Engineering, University of Pisa, Pisa 56126, Italy

a r t i c l e i n f o

abstract

Article history:
Received 13 August 2012
Received in revised form
10 November 2012
Accepted 17 December 2012
Available online 24 January 2013

This paper investigates laser beam welding of dissimilar AISI430F and AISI440C stainless steels.
A combined welding and pre-and-postweld treatment technique was developed and used successfully
to avoid micro-crack formation. This paper also examined the effects of laser welding parameters and
line energy on weld bead geometry and tried to obtain an optimized laser-welded joint using a full
factorial design of experiment technique. The models developed were used to nd optimal parameters
for the desired geometric criteria. All the bead characteristics varied positively as laser power increased
or welding speed decreased. Penetration size factor decreased rapidly due to keyhole formation for line
energy input in the range of 1520 kJ/m. Laser power of 790810 W and welding speed of 3.64.0
m/min were the optimal parameters providing an excellent welded component. Whatever the
optimization criteria, beam incident angle was around its limiting value of 151 to achieve optimal
geometrical features of the weld.
& 2012 Elsevier Ltd. All rights reserved.

Keywords:
Laser-welding
Steels
Butt joint

1. Introduction
Laser welding with high power density, high degree of automation and high production rate is extremely advantageous in
industrial applications as stated in [1]. Joints between dissimilar
metals are particularly common in components used in the
automotive, power generation, chemical, petrochemical, nuclear
and electronics industries as described in [2]. The ability to use
different metals and alloys in a product provides the designer and
production engineer with greater exibility, and often results in
technical and economic advantages over components manufactured from a single material.
Welding of metals and alloys is an experienced subject,
dissimilar welding represents a major scientic and technical
challenge. Emerging new technologies increasingly require dissimilar metals and alloys to be joined. Most metals and alloys are
very weldable either in conduction or keyhole modes. Two factors
most important to weldability are hardenability and susceptibility of the hardened structure to cracking. Hardenability is related
to the cooling rate of metals. The faster cooling rate tends to
produce higher hardness and hence, the hardened structure
becomes more susceptible to cracking. According to Kou [3], for
a constant mass of the metal and the particular welding process

Corresponding author. Tel.: 39 502218137.

E-mail addresses: muhshin92@gmail.com,
muhshin92@yahoo.com (M.M.A. Khan).
0030-3992/$ - see front matter & 2012 Elsevier Ltd. All rights reserved.
http://dx.doi.org/10.1016/j.optlastec.2012.12.025

and procedure, cooling rate depends only on weld preheat

temperature as it can modify the energy input used to make the
weld. It can also be changed to reduce the cooling rate. Fig. 1.1
shows effects of combined welding and pre-and-post-weld treatment technique on mitigation of crack formation during welding.
Also, cracking may result from any of the following factor:
restraint, weld shape, and material composition. Restraint
is always present in any weld because as the weld solidies it
acquires strength but continues to cool and shrink. It is the degree
of restraint that becomes critical. Restraint relates to the weld
design, the weldment design, and welding procedure. Khan et al.
[4,5] show that weld shape is also a function of the weld design,
weldment design and welding procedure. Welding procedure
relates to the placement of welds or beads in the weld, the shape
of the beads and the shape of the nished surface as stated in [6].
The third factor is the metal composition. Kou [7] describes that
segregation is important, however, since impurities such as sulfur
and phosphorus tend to form low-melting-point lms between
solidifying grains of the metal. These impurities relate to weld
joint detail and the welding process, since they affect the amount
of dilution.
Besides, the predominant phase transformation in the martensitic AISI440C stainless steels is austenite-to-martensite that
occurs in the fusion zone and the as-welded hardness is usually
in the range of 3555 HRC with greater cracking risk. Again,
the combined inuence of grain size and precipitation behaviour
on weld metal and HAZ toughness and ductility is analogous
to the high-temperature embrittlement phenomenon, which is

126

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

Fig. 1.1. butt welded surface (a) before and (b) after applying the combined preheat-and-postheat welding technique for the line energy, LE 15 kJ/m.

the characteristic in medium chromium alloys like ferritic

AISI430F stainless steels. Postweld treatment is, therefore, almost
always required for welding dissimilar martensitic and ferritic
stainless steels and used primarily to temper the martensite that,
metallurgically, promotes transformation of martensite to ferrite
and very ne carbides as stated by Lippold and Kotecki [8]. This
transformation reduces strength but improves ductility and
toughness. Besides, as described in [9], postweld thermal treatments acts to mitigate the high-temperature embrittlement effect
by promoting coarsening of the precipitates, thereby reduce their
negative inuence.
The severe heating rate induced by a moving laser source on a
metal surface generates a thermal shock having a deep impact on
the change in weld microstructure as well as the formation of
defects particularly when dissimilar steels are welded. This is
because the material is brought to the melting and vaporization
state almost instantaneously along isotherms ruled by the so
called modied Bessel function of the second type and zero order
as reported in [10]. Besides, Huang and Sun [11] state that, as
compared to heating rate, cooling rates can be less dangerous for
the material integrity since temperature decreases with exponential laws offering lower gradients in the tail of the melt pool.
Therefore, the idea of the present paper is to generate combined
pre-and-post-heating areas in which the weld material experiences less severe heating and cooling rates. This is obtained by
using a non-vertical conguration (A: 151301451) of the laser
beam with respect to the material surface as schematically shown
in Fig. 1.2.
Laser welding is usually done with the aim of producing an
excellent joint at low cost. However, achieving such a weld
junction without optimization is impossible as stated in [12].
Benyounis and Olabi [13] report that various optimization methods are applied to dene the desired output variables through
developing mathematical models. In the last two decades, the use
of factorial design of experiment (DOE) has grown rapidly and a
variety of industries have employed this technique to improve
products or manufacturing processes as described in [14]. It is a
powerful and effective technique to solve challenging quality
problems. In practice, as stated by Yang et al. [15], the factorial
design of experiment technique has been used quite successfully
in several industrial applications as in designing electrical/
mechanical components or optimizing various manufacturing
processes.
As seen in the above literatures, weldability is very complex
and all the welding factors are interrelated. To better understand
the weldability, it is necessary to study the weld joint and
geometrical features determining the weld shape. The alloy
composition and pre-and-postweld processing must be considered when selecting the welding parameters due to the high
cooling and heating rates associated with laser welding. Besides,

Fig. 1.2. Schematic diagram to show the variation in incidence angle, A in a plane
tangential to the sample trajectory.

no result has been reported yet in the literature on the laser butt
welding of these particular dissimilar ferritic AISI430F and martensitic AISI440C stainless steels or for similar industrial applications. Moreover, in this study, a novel combined laser welding and
pre-and-postweld treatment technique is developed for overcoming the problems associated with laser welding of dissimilar
stainless steels and applied successfully to the fabrication of fuel
injector. This paper, therefore, examines and optimizes the laser
welding of dissimilar martensitic/ferritic stainless steels in a
constrained butt joint conguration. This study is focused on:
Combined welding and pre-and-postweld treatment concept
and its effects,
Effects of laser welding parameters such as laser power,
welding speed and incidence angle of the beam, impinged
tangentially in the direction of rotation as shown in Fig. 1.2, on
the geometrical features of the weld i.e., on weld width,
resistance length, and penetration depth,
Effects of line energy or energy per unit length on the same
weld geometrical features to understand the energy dependent welding phenomena, and

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

Finally, determination of optimal range of laser welding

parameters, using the developed models with numerical
optimization, to minimize the weld width and maximize the
weld penetration depth.

2. Materials and experimental procedures

2.1. Materials
Two cylindrical shells made of ferritic AISI430F (cold drawn,
annealed and centerless ground) and martensitic AISI440C (prehardened and tempered) stainless steels are welded circularly to
make a butt joint. This dissimilar joint is selected based on both
technical and economical aspects, because they can provide
satisfactory service performance and considerable savings. Moreover, in automotive industries, these materials are frequently
used in welded form for making different types of fuel injectors.
The chemical compositions of base metals available in as-received
condition and the weld seam characteristics are shown in Fig. 2.1
and Table 2.1, respectively. The inside diameter of the outer shell
and the outside diameter of the inner shell are machined to
7.570.025 mm and 7.458 70.015 mm, respectively, to have a
clearance t between them when the shells are assembled.
2.2. Experimental procedures
Specimens are welded circularly in a llet joint conguration
using a 1.1 kW continuous wave Nd:YAG laser (Ron DY011).
The optical system consisted of a 300 mm ber and two lenses
of 200 mm focal and collimate lengths are used to deliver the
laser with a minimum focal spot diameter of 300 mm. A three-step
procedure is followed to locate the focal point. First, an exceptionally sharp-nosed tool of 200 mm in height is attached to laser

127

head mounted on Z motion stage. The laser head is then set

tangentially in the direction of rotation to an intended beam
incident angle. Finally, the positions of the XYZ motion stages
are adjusted in such a way that pointed tool tip touches the
planned point of focus. Laser beam is focused on this located point
through the laser head at the specied angle, and the necessary
rotary motion is provided to the specimen through specimen
holder mounted on an XY motion stage. Computer control panel
is interfaced with the linear XYZ as well as rotary motion
systems to regulate the aforesaid movements.
The experiment is initially planned based on statistical factorial
experimental design with full replication. During experimentation, laser power, (P), welding speed (S), and beam incident angle
(A) are selected as process input variables for laser welding.
Table 2.2 shows the experimental condition, laser welding input
variables, and design levels used at a glance. Each of the input
variables and its working range is selected based on industrially
recommended laser-welding parameters used in automotive
industries.
General Full Factorial Design is used as a statistical design of
experiment technique to develop statistical models relating the
welding input parameters to each of the two output responses
of the weld (weld width and penetration depth). The adequacies
of the models developed and their signicant linear and interaction model terms are measured by analyzing variance and other
adequacy measures. Finally, these mathematical models are used
to determine the optimal settings of welding parameters to
ensure the desired weld quality. In this study, the quality criteria
dened for the weld to determine the optimal settings of welding
parameters are the minimization of weld width and the
maximization of weld penetration depth.
Besides, the energy delivered per unit length of weld line is
referred to as line energy (LE), which is frequently used in various
laser-processing techniques and termed as a key-parameter when
continuous-wave laser is used. This term is calculated as the ratio

9.560.03

AISI
440C
440C

AISI
430F
430F
Rm

Inner Shell
Inner
Shell

Outer
Shell

Dp

Inner
Shell

Outer Shell
Outer
Shell

7.4580.015
Fig. 2.1. Characterization of welding cross-section (W: Weld width, Dp: Weld penetration depth, Rm: Minimum crack-path).

Table 2.1
Chemical compositions of base metals of the weld.
Base metal

AISI430F
AISI440C

Composition (in weight percent)

C

Cr

Ni

Mn

Si

Mo

Se

Fe

0.12
0.951.2

16.018.0
17.2

0.75

1.0
1.00

0.04

0.03
0.015

1.00
1.00

0.75

0.20

Remainder
Remainder

128

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

2.3. Weld bead characterization

of laser power over the welding speed as shown in Eq. (1):

LE 0:06 

P
S

kJ=m

where, LE is line energy; P is laser power in watt (W) describing

the thermal source; and S is welding speed in m/min determining
the irradiation time. According to the Eq. (1), the combinations of
laser power of 6001000 W and welding speed of 2.04.0 m/min
resulted in nominal line energy input in the range of 9.030.0 kJ/m.
During experimentation, laser beam is focused on the point of
interface at a certain angle with the direction of rotation. Argon is
used as shielding gas with a constant ow rate of 29 l/min to
protect weld surface from oxidation and suppress the generation
of plasma during welding. A standard washing procedure, which
is practised in the automotive industries, is followed to clean, cool
and dry the specimens. The experimental set-up for the laserwelding system is illustrated in Fig. 2.2.

Welding tests are carried out in a random order to avoid any

systematic error in the experiment. After welding, transverse
sections are prepared by cutting the samples axially using
SampleMet II (Beuhler, IL) model abrasive cutter. The sectioned
samples are mounted, polished, and etched for mechanical
characterization. Software, Leica IM500, incorporated with an
optical microscope (Leica MZ125) is used to measure weld width,
penetration depth, and minimum crack path as shown in Fig. 1.
Each set of experiments is replicated three times to ensure
statistical accuracy. The mean value of each measured response
parameter is determined and recorded for further analysis.
Table 2.3 shows the average measured responses for various
laser-welding conditions.
The guidance on quality levels for imperfections given in ISO
13919-1:1996 is followed to assure the desired weld quality.
At this point, each welded specimen is visually inspected before

Table 2.2
Experimental conditions and response factors.
Process factors
Laser power (W)
Welding speed (m/min)
Angle of incidence (1)
Constant factors
Base material
Laser source
Shielding gas
Response factors
Weld bead characteristics

Symbols
P
S
A

Actual values
600
2.0
15

800
3.0
30

Outer shell
Inner shell
Continuous wave Nd:YAG laser
Type
Flow rate

1000
4.0
45

AISI 430F
AISI 440C
Argon
29 l/min

Weld width (W), weld penetration depth (Dp), and minimum crack-path (Rm)

Table 2.3
Design matrix with actual factors and measured mean responses.
Standard order

Laser
Head

Specimen
Shielding
Gas Nozzle

Specimen
Holder

Fig. 2.2. Photographic view of Nd:YAG laser-welding system.

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27

Process factors

Response factors

P (W)

S (m/min)

A (deg)

W (mm)

Rm (mm)

Dp (mm)

600
800
1000
600
800
1000
600
800
1000
600
800
1000
600
800
1000
600
800
1000
600
800
1000
600
800
1000
600
800
1000

2.0
2-0
2.0
3.0
3.0
3.0
4.0
4.0
4.0
2.0
2.0
2.0
3.0
3.0
3.0
4.0
4.0
4.0
2.0
2.0
2.0
3.0
3.0
3.0
4.0
4.0
4.0

15
15
15
15
15
15
15
15
15
30
30
30
30
30
30
30
30
30
45
45
45
45
45
45
45
45
45

930.7
1050.3
1021.9
639.6
712.5
851.5
584.6
712.6
881.5
1002.1
1146.8
1217.7
742.3
897.6
899.5
733.6
820.8
848
912.8
1024.6
1073.5
715.8
843.4
933.5
643.3
707.9
901.4

532.5
832.8
1180.6
397.9
822.1
1129.1
292.3
563.4
618.2
375.8
524.9
799.4
195.7
601
669.2
249.3
456.1
510.7
166.8
448.1
539
174.5
376.6
486.1
146.2
298.2
413.6

599.9
956.6
1234.5
438.5
833.8
1148
301.7
594
1043.9
406.2
593.5
959.7
282.8
603.9
1074.4
287.4
464.2
819.2
256.3
631.3
801.4
261.9
458
643.6
205
306.4
505.3

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

and after the cut using the optical microscope. Hermetic weld is
ensured by performing leak test in vacuum for each of welded
specimens. During leak test, nitrogen is inated into the
assembled part at a pulsed pressure in the range 10150 bar for
the expected life cycles. This method also guarantees that the
weld will not fail during its service life. In case of failure, the
internal cracks generated during the welding process propagate
up to the free surface and N2 leakage is detected by a loss of
vacuum into the chamber.
3. Results and discussion
Various weld prole characteristics are measured with axially
cut specimens using an optical microscope and are recorded for
further analyses described in the succeeding sections.
Perturbation plots are used to illustrate the effects of individual process parameter such as laser power (P), welding speed (S),
and beam incident angle (A) on geometrical features of the weld
e.g., weld width, weld penetration depth, and minimum crackpath. Contour plots are used to show the two-factor interaction
effects on the same weld bead geometry.
The line energy is plotted against weld width, weld penetration depth, and minimum crack-path with a view to demonstrate

GL Ta
Tm
GF

Ta

Tv

GB Ta

Melt pool

Spot

GL

Ta
GL

GF

Tv

Melt pool

GB

Spot

GL

the effects of energy input on weld prole characteristics, and to

explain different laser welding phenomena as well.
3.1. Combined welding and pre-and-postweld treatment concept
and its effects
The main concept adopted to modify the experimental set up
has been described schematically in Fig. 3.1. When a moving laser
beam hits the surface of a metal in perpendicular direction, an
asymmetric melt pool is generated (Fig. 3.1(a)). Heat conduction
in the surroundings is not isotropic and generates different
thermal gradients in frontal (GF), lateral (GL) and back (GB)
directions. The entity of such difference can be calculated following [10]. The lengths of the arrows represent a realistic proportion
among the modules of the thermal gradients considering
the jump between vaporization temperature TV and ambient
temperature TA, and the distance in which it is achieved [3].
The frontal gradient is extremely higher with respect to the back
due to the deformation of isotherms caused by the welding speed.
The basic idea is to reduce this shock by decreasing GF: this effect
is obtained displacing and widening the laser spot (dot line),
as reported in Fig. 3.2(b).
From the experimental point of view, this is obtained by
inclining the beam on the tangential plane while the specimen
moves against the inclined beam as shown in Fig. 3.2(b). For any
generic position of the butt specimen welded with an inclined
beam shown in Fig. 1.2, the cross sectional view of the melt pool
is found to be also asymmetric with respect to the beam axis as is
reported in Fig. 3.2(a) [16]. This is because the portion of melt
pool surface exposed directly to the laser beam is depleted more
as compared to other portion of the keyhole surface.
The top view shows an elliptical print in which the left side is
wider than the right one thus generating a thermal gradient of
lower entity in the cross sectional plane. This phenomenon
causing the enlargement of the melt pool and the unsymmetrical
distribution of energy inside the print is adopted in the present
research to generate a preheating effect before fresh material
reaches the beam axis
During welding, the fresh material is rst exposed to the laser
radiation in a region with positive defocus which determines the
reduction in the thermal gradient. It then reaches the highest
intensity zone of the laser beam. At the exit (right) side, the beam
acting on welded material results in slight postheat treatment
of the weld. No relevant changes occur in the lateral gradient.
The ultimate result is a welded seam due to continuous rotation
of the unsymmetrical weld pool over 3601 that reduces the

io
n

Fig. 3.1. Top views of melt pools and basic draft of thermal gradients (G) in
different directions with: (a) laser beam perpendicular to the surface (b) laser
beam inclined on the tangential plane.

ct

La

di
re

se

rb

ea

nt
id
e
ri
nc
se
La

Melt pool

129

Fo
pl cal
an
e

+ defocus
sample rot. speed

-defocus

Solidified seam
AISI 430F
AISI 440C
Fig. 3.2. (a) Top and cross sectional views of an instant during the welding process, (b) draft of the basic concept to obtain a less severe heating rate on the fresh material.

130

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

Perturbation

Perturbation

1220

Weld Width, m

1060
P
900

740

S
A

P
A

Weld Penetration Depth, m

1300

1025
P
750

A
S

S
A

475
P
200

580
-1.00

-0.50

0.00

0.50

- 1.00

1.00

-0.50

0.00

0.50

1.00

Deviation from Reference Point (Coded Units)

Deviation from Reference Point (Coded Units)

Perturbation

Minimum Crack Path, m

1200

925
A
650

S
A

375
P
100
-1.00

-0.50

0.00

0.50

1.00

Deviation from Reference Point (Coded Units)

Fig. 3.3. Perturbation plot showing effect of all factors on (a) weld width, (b) weld penetration depth, (c) minimum crack path.

occurrence of crack formation as can be seen in Fig. 1.1. This

combined process also decreases the penetration depth of the
weld since welding with inclined laser beam causes reection
losses and a more complex heat transmission through refraction
described by Snells law.

3.2. Effects of process parameters

The Fig 3.3(a)(c) show the perturbation plots to compare the
effects of all the process parameters at the center point in the
design space. The results suggest that laser power has the most
signicant positive impact on the weld width; weld penetration
depth; and minimum crack-path. The opposite phenomena are
observed for the welding speed. This is because higher laser
power and slower welding speed result in higher energy deposition on the weld area, and longer irradiation time for the
deposited energy to diffuse into material.
These gures also illustrate that increase in beam incident
angle results in shallower weld penetration, and shorter minimum crack-path, whereas the larger beam incident angle causes
the wider weld width. These are due to following consequences:
(i) the larger the angle of incidence, the higher the energy loss to
the surrounding through reection; and (ii) dominance of uniform heat conduction in all directions over the z-preferential or
axial heat conduction for the lower energy input to the materials
to be welded.
The contour plots shown in Fig. 3.4(a)(c) demonstrate the
facts that interactions of higher laser power and slower welding
speed cause wider weld width; deeper weld penetration; and
longer minimum crack-path.

3.3. Effects of line energy

Fig 3.5(a)(c) show the effects of line energy input on the weld
penetration depth (DP), minimum crack path (Rm), and weld
width (W), respectively, whereas variation in penetration size
factor with line energy is illustrated in Fig 3.5(d).
For line energy in the range of 9.015 kJ/m, as illustrated in
Fig 3.5(c)(d), there is a rapid growth in weld width (W) with
energy input, whereas change in penetration size factor (W/Dp) is
insignicant. Slight negative variations in penetration size factor
prove that the laser welding is mainly conduction limited. Since
the melt pool geometry depends on energy intensity, uniform
conduction occurring in all directions usually results in semicircular weld prole. However, the heat conduction along the
beam axis becomes dominant with the increase in energy input
and weld shape changes from semi-circular to parabolic. Similar
parametric effects on the welding pool width at surface [17] and
the upper width [18] are also observed for welding dissimilar low
carbon and austenitic stainless steels in a butt joint conguration
using the CO2 laser.
Again, a small variation in weld width is observed for the line
energy input in the range of 1520 kJ/m. Nonetheless, as shown
in Fig 3.5(a) and (d), a sharp decrease (starting from 15 kJ/m) in
W/Dp demonstrates the fact that the weld penetration depth
increases at a faster rate than the weld width in this range and
establishes a keyhole formation regime. As a result, the weld bead
becomes almost cylindrical. Penetration size factor increases with
further increase in line energy. This is due to the creation of upper
keyhole plasma plume that acts as a point heat source above weld
plane. This generated plasma plume acts in the keyhole and forms
a chalice shaped weld bead prole when energy input is more

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

Weld Width, m

737

3.5
791
871

3.0

950

2.5

Weld Penetration Depth, m

4.0

Welding Speed, m/min

Welding Speed, m/min

4.0

131

3.5

355

487

3.0

618

750

882

2.5

1030
1110

2.0

2.0
600

700

800

900

1000

600

700

Laser Power, W

900

1000

Minimum Crack Path, m

4.0

Welding Speed, m/min

800

Laser Power, W

3.5

307

3.0

406

504
603

2.5

701

2.0
600

700

800

900

1000

Laser Power, W
Fig. 3.4. Contour graphs to show the interaction effects of P and S on (a) weld width, (b) weld penetration depth, and (c) minimum crack path at A 301.

than 20 kJ/m, which is quite similar to the result obtained from an

experimental study conducted by Khan et al. [5].
Besides, as illustrated in Fig 3.5(b), variation in minimum
crack-path with the line energy input shows the same trend as
the weld penetration depth. This is because of the existing linear,
positive relationship between them as can be seen in Fig. 3.6.
3.4. Process parameter optimization
The optimization part in Design-Expert software V7 searches
for a combination of factor levels that simultaneously satisfy the
requirements placed (i.e., optimization criteria) on each of the
responses and process input factors (i.e., multiple-response optimization). Numerical and graphical optimization methods are
used in this work by selecting the desired goals for each factor
and response. As mentioned before, the numerical optimization
process involves combining the goals into an overall desirability
function (D). The numerical optimization feature in the designexpert package nds one point or more in the factors domain that
maximizes this objective function. In this study, the objective is to
optimize the autogenous laser-welded joints subjected to minimize the weld width and maximize weld penetration depth a
characterizing factor that determines the minimum crack-path as
shown in Fig. 3.6. In order to achieve these objectives, mathematical models are developed to relate the aforesaid geometrical
features and the selected laser welding input variables.
3.4.1. Development of mathematical models
At this stage, the t summary in the design-expert software is
used to select the models that best describe the response factors.

The t summary includes sequential model sum squares to select

the highest order polynomial where additional terms are signicant and the model is not aliased. In addition, model summary
statistics of the t summary focuses on the model that maximizes
adjusted R-squared and predicted R-squared values. The sequential F-test is carried out using the same statistical software
package to check if the regression model is signicant and nd
out the signicant model terms of the developed models as well.
The step-wise regression method is also applied to eliminate the
insignicant model terms automatically.
Suitable response models for the response factors are selected
based on the t summaries. From t summary output of the
measured responses shown in Tables 3.13.4, it is evident that
quadratic model is statistically signicant for the weld width,
whereas for weld penetration depth; two-factor interaction (2FI)
models are statistically recommended for further analyses.
The test for signicance of the regression models and the test
for signicance on individual model coefcients are performed
using the same statistical package. By selecting the step-wise
regression method that eliminates the insignicant model terms
automatically, the resulting ANOVA Tables 3.5 and 3.6 for the
selected models summarize the analysis of variance of each
response and illustrate its signicant model terms as well. The
aforestated tables demonstrate that calculated Fishers Model-F
and Model-P values are, respectively, 57.17 and o0.0001 for
weld width quadratic model; and 107.74 and o0.0001 for weld
penetration depth 2FI model. These Model-F and Model-P
values imply that the selected models are highly signicant and
there is only a less than 0.01% chance that these large Model-F
values could occur due to noise. The associated P value is also
used to estimate whether F is large enough to indicate statistical

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

1400

1400
1200
1000
800
600
400
15
15
30
30
45
45

200

Minimum Crack Path, Rm (m)

Weld Penetration Depth, Dp (m)

132

1200
1000
800
600
400
15
30
45

200
0

10

15

20

25

30

35

10

1300

Weld Width, W (m)

1200
1100
1000
900
800
700
15
30
45

600
500
5

10

15

20

25

15

20

25

30

35

Line Energy, LE (kJ/m)

30

Penetration Size Factor, W/D (a.u.)

Line Energy, LE (kJ/m)

3.5

15
30

3.0
2.5
2.0
1.5
1.0
0.5
0.0
5

35

Line Energy, LE (kJ/m)

10

15

20

25

30

35

Line Energy, LE (kJ/m)

Fig. 3.5. Effect of line energy on (a) weld penetration depth, (b) minimum crack-path, (c) weld width, and (d) penetration size factor for different beam incident angle.

Table 3.1
Sequential model sum of squares for weld width model.

900

Minimum Crack Path, Rm (m)

800
700
600
500
400

Source

Sum of
squares

df

Mean
Linear
2FI
Quadratic
Cubic
Residual

4.372E 007
1.325E 006
3042.62
2.340E 005
53388.46
41493.34

Total

4.538E 007 26 1.745E 006

1
3
3
3
7
9

Mean
square

F
value

p-value
prob4 F

4.372E 007
4.417E 005 29.28 o0.0001
1014.21
0.059
0.9808
77994.39
13.15
0.0001
7626.92
1.65
0.2363
4610.37

Suggested
Aliased

300
200

30

100
0

200

400

600

800

1000

Table 3.2
Model summary statistics for weld width model.
Source

Std. dev.

R2

Adj R2

Pred R2

PRESS

Linear
2FI
Quadratic
Cubic

122.83
131.56
77.01
67.90

0.7997
0.8015
0.9427
0.9750

0.7724
0.7389
0.9105
0.9304

0.7298
0.6687
0.8369
0.7707

4.477E 005
5.490E 005
2.702E 005
3.799E 005

1200

Weld Penetration Depth, Dp (m)

Fig. 3.6. Relationship between the minimum crack-path and weld penetration
depth.

signicance. If P value is lower than 0.05, it indicates that the

model is statistically signicant as stated by Zulkali et al. [19].
The same tables also show other adequacy measures e.g.,
R-squared, adjusted R-squared, and predicted R-squared values. All
the adequacy measures are in logical agreement and indicate
signicant relationships. Moreover, adequate precision compares
range of predicted value at the design points to average prediction
error. The adequate precision ratios in all cases are dramatically
greater than 4 indicating adequate models discrimination.

Suggested
Aliased

Again, the ANOVA table for the weld width model shows that
there is a quadratic relationship between weld width and welding
parameters. The linear terms of laser power (P) and welding speed
(S); and the quadratic terms of welding speed and incident angle are
the signicant model terms associated with the weld width. However, linear term of beam incident angle is added to support hierarchy
of weld width model. For the weld penetration depth model, ANOVA
table demonstrates that all three linear terms i.e., laser power (P),
welding speed (S) and beam incident angle; and two-factor

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

interactions (2FI) of laser power-welding speed (PS), are the signicant model terms.
From the results shown in Tables 3.13.6, it is, therefore,
apparent that the developed statistical models for predicting weld
width and penetration depth are fairly accurate and can be of
following forms:
(i) Weld width
W1:06 2163:33 0:76P1154:6S 36:97A
154:87S2 0:58006A2
(ii) Weld penetration depth
Dp 446:02 2:16 P106:24 S 7:674 A0:024 P x A

Table 3.3
Sequential model sum of squares for weld penetration depth model.
Source

Sum of
squares

df Mean
square

Mean
Linear
2FI
Quadratic
Cubic
Residual

9.442E 006
2.046E 006
61301.75
8834.15
50208.79
37119.61

Total

1.165E 007

1
3
3
3
7
9

F
value

9.442E 006
6.821E 005 95.30
20433.92
4.04
2944.72
0.54
7172.68
1.74
4124.40

p-value
prob4 F

o0.0001
0.0223
0.6620
0.2158

Suggested
Aliased

26 4.479E 005

Table 3.4
Model summary statistics for weld penetration depth model.
Source

Std. dev.

R2

Adj R2

Pred R2

PRESS

Linear
2FI
Quadratic
Cubic

84.60
71.14
73.88
64.22

0.9285
0.9564
0.9604
0.9832

0.9188
0.9426
0.9381
0.9532

0.8994
0.9240
0.9006
0.8310

2.216E 005
1.674E 005
2.191E 005
3.725E 005

Suggested
Aliased

133

Normality of residual data and amount of residuals in prediction are then checked to ensure statistical validation of the
developed models. The normality of data is veried by plotting
the normal probability plot (NPP) of residuals. The residual is the
difference between observed and predicted values (or tted
value) obtained from the regression model. The data set is
normally distributed if the points on the plot fall fairly close to
the straight line. The normal probability plots of residual values
for weld width, and penetration depth are illustrated in Fig
3.7(a)(b), respectively. The experimental points are reasonably
aligned with predicted or tted points suggesting the normality of
data. This is an implication that empirical distribution of residual
data is well-compared with a normal distribution having the
same mean and variance
Fig 3.8(a)(b) are showing the relationships between the
actual and predicted values of weld width and penetration depth.
Since the points plotted are close to and around the diagonal line,
the difference between the predicted and actual value for each
point can be considered to be minimal. It is also an indication that
the statistical models for prediction are adequate and predicted
results are in good agreement with the measured data.
3.4.2. Numerical optimization
Two criteria are introduced in this numerical optimization.
The rst set of criteria is to maximize weld penetration depth
with no limitation on either process parameters or weld width. In
this case, all the process parameters and weld width (rst
response) are set within a specied range. Furthermore, lowering
the laser power and increasing the welding speed are the most
common techniques used in automotive industries to produce
relatively low-cost and excellent weld joints. Taking these cost
and quality aspects into account, second set of criteria for process
parameter optimization is xed to maximize weld penetration
depth and welding speed, and minimize the laser power and weld
width. Table 3.7 summarizes these two criteria, lower and upper
limits as well as importance for each input and response factor.
Tables 3.8 and 3.9 show the optimal solution based on the two
optimization criteria as determined by design-expert software.
The optimization results clearly demonstrate that, whatever the

Table 3.5
ANOVA table for weld width quadratic model.
Source

Sum of squares

df

Mean square

F value

p-value prob 4F

Model
P
S
A
S2
A2
Residual

1.549E 006
3.876E 005
8.518E 005
19093.30
1.405E 005
93119.36
1.083E 005

5
1
1
1
1
1
20

3.097E 005
3.876E 005
8.518E 005
19093.30
1.405E 005
93119.36
5417.49

57.17
71.54
157.24
3.52
25.93
17.19

o 0.0001
o 0.0001
o 0.0001
0.0751
o 0.0001
0.0005

Cor total
R2 0.9346

1.657E 006
Adj R2 0.9183

25
Pred R2 0.8889

Signicant

Adeq Precision 25.958

Table 3.6
ANOVA table for weld penetration depth 2FI model.
Source

Sum of squares

df

Mean square

F value

p-value prob 4F

Model
P
S
A
PA
Residual

2.101E 006
1.388E 006
1.885E 005
4.895E 005
55065.56
1.024E 005

4
1
1
1
1
21

5.254E 005
1.388E 006
1.885E 005
4.895E 005
55065.56
4876.13

107.74
284.74
38.65
100.38
11.29

o 0.0001
o 0.0001
o 0.0001
o 0.0001
0.0030

Cor total
R2 0.9535

2.204E 006
Adj R2 0.9447

25
PredR2 0.9324

Signicant

Adeq Precision 36.953

134

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

Normal Plot of Residuals

99

99

95
90
80
70

95
90
80
70

Normal % Probability

Normal % Probability

Normal Plot of Residuals

50
30
20
10
5

50
30
20
10
5
1

-1.70

-0.82

0.07

0.95

1.84

-1.98

Studentized Residuals

-0.85

0.27

1.40

2.53

Studentized Residuals

Fig. 3.7. Normal probability plot for weld (a) width, and (b) penetration depth.

Predicted vs. Actual

1900

1300

1625

1000

Predicted

Predicted

Predicted vs. Actual

1350

700

1075

400

800

100
856

1108

1360

1613

1865

124.7

407.6

Actual

690.5

973.4

1256.3

Actual

Fig. 3.8. Scatter diagrams of weld (a) width, and (b) penetration depth.

Table 3.7
Optimization criteria used in this study.
Parameters or
responses

Limits

Table 3.8
Optimal solutions as obtained based on rst criterion.

Lower Upper
P (W)
S (m/min)
A (1)
W (mm)
DP (mm)

600
2.0
15
584
206

Soln no.

Importance Criterion

1000
4.0
45
1218
1235

3
3
3
5
5

First

Second

Is in range
Is in range
Is in range
Is in range
Maximize

Minimize
Maximize
Is in range
Minimize
Maximize

optimization criteria, the beam incident angle has to be its

limiting value of 151 to achieve maximum weld penetration
depth and minimum weld width. This result also supports the
discussion made earlier on the effect of beam incident angle on
the geometrical features of the weld.
Again, Table 3.8 demonstrating the optimal welding conditions
according to the rst criterion, it is found that maximum weld
penetration depth of 1250 mm is obtained when laser power and
welding speed are set to their respective highest and lowest limits.
However, with acceptable weld penetration depth, the laser power
can be minimized to around 790 W and the highest limit of the
welding speed can be used instead of its lowest limit of 2 m/min.

1
2
3
4
5
6
7
8
9
10

P (W)

S (m/min)

F (lm)

DP (lm)

W (lm)

998.6
999.3
998.9
997.9
999
999.5
998.9
998.5
999.6
999.7

2.02
2.16
2.03
2.00
2.02
2.04
2.03
2.14
2.18
2.01

15.2
15.1
15
15
15.5
15.6
15.2
15.2
15
15.8

1248.1
1235.9
1250
1251.7
1243.8
1240.4
1247.4
1235.3
1236
1240.8

1084.6
1039.7
1078.5
1088.6
1088
1082.3
1081.1
1048.7
1033.9
1095

In this case, as shown in Table 3.9, the weld width and penetration
depth would be of 665 mm and 706 mm, respectively, which are
much higher than the prerequisite values for the weld. These results
also indicate the fact that laser welding ought to be conductionlimited for this particular joint type and laser-material combinations
in order to obtain the optimal geometrical features of the weld.
Since optimal range of laser power and welding speed selected
based on second criterion is, respectively, much lower and higher
than that obtained for rst set of criterion, any combination of process

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

parameters of the second optimal set would cause less energy input
to constrained butt joints to be made. This reduced energy input to
weld materials would, ultimately, result in less distortion, and

135

formation of cracks, blow holes, and spatter with a consequent

improve in the weld quality. The results given in Table 3.10 also
support the improvement of weld quality at lower laser power and
higher welding speed.

Table 3.9
Optimal solutions as obtained based on second criterion.
Soln no.
1
2
3
4
5
6
7
8
9
10

4. Weld microstructure and microhardness

P (W)

S (m/min)

F (lm)

DP (lm)

W (lm)

792.7
794.9
790.1
795.9
789.2
797.2
826.3
816.9
789.5
810.9

3.95
3.94
3.98
3.99
4.00
3.73
3.99
3.78
3.66
3.63

15.00
15.00
15.00
15.00
15.00
15.00
15.00
15.00
15.00
15.00

676.3
681.3
668.5
678.2
664.6
707.6
731.9
737.6
700.9
742.8

705.4
706.1
705.5
708.7
706.2
702.8
723.9
712.6
699.5
710.3

Table 3.10
Visual inspection of weld quality.
Process parameters
P (W)
600
800
1000
600
800
1000
600
800
1000

Visual check

S (m/min)

A (1)

Cracks

Blow holes

Spatter

2.0
2.0
2.0
3.0
3.0
3.0
4.0
4.0
4.0

15
15
15
15
15
15
15
15
15

0
1
0
1
1
1
0
0
0

0
0
0
0
0
0
0
0
0

1
2
2
1
1
2
0
0
1

0 no defect, 1 exist but acceptable, 23 not acceptable.

This study is carried out to investigate the various microstructures typically formed at the weld zone and various locations
close to both base metals for a line input of 12.0 kJ/m.
The particular sample selected for the metallurgical study is one
of those samples that provide the weld with designed geometry.
Fig 4.1(a) shows the micrograph of the weld made of AISI440C
and AISI430F stainless steels.
As shown in Fig 4.1(b) and (c), the microstructures at the inner
shell are cellular dendritic, whereas columnar dendritic structures
form at the outer shell. These microstructures are a result of
solidication behavior and subsequent solid-phase transformation controlled by melting ration of two materials to be joined
and weld cooling rate [17]. Since base metals consist of martensitic AISI440C and ferritic AISI430F stainless steels, the microstructures that form in the fusion zone must contain a variety of
complex martensite-ferrite structures. Moreover, as shown in
Fig (c), the intra-granular Cr23C6 carbide formation is evident in
the microstructures forming near the fusion zone boundary. As a
result, at the outer shell, average microhardness of the fusion
zone is found to be much higher than that of both the heat
affected zone and the base metal as can be seen in Fig. 4.2.
Fig. 4.1(b) shows the microstructures of base metal and HAZ of
the inner shell, which is pre-hardened and-tempered. Both
microstructures contain partly spherodized primary carbide particles in a tempered martensitic matrix. However, microstructures

Fig. 4.1. (a) typical micrograph of laser welding of AISI440C and AISI430F stainless steels, and base metal (BM), heat affected zone (HAZ) and Fusion zone (FZ) of (b) inner
shell and (c) outer shell.

136

M.M.A. Khan et al. / Optics & Laser Technology 49 (2013) 125136

600

Vicker's Hardness Number

Local Microhardness

Intersection
Line

500

400

300

AISI
440C

AISI
430F

For these particular joint type and laser-material combinations:

| A laser power of 790810 W and a welding speed of 3.64.0
m/min are the optimal input parameters to obtain an excellent
weld.
| Whatever the optimization criteria, the beam incident angle
has to be around its limiting value of 151.
Efcient and low-cost weld joints could be achieved using the
welding conditions drawn from the numerical optimization
Acknowledgements

HAZ

200

BM

HAZ

Welding Pool
FZ
FZ
Boundary
Boundary

BM

Fig. 4.2. Vickers local microhardness prole along the line shown in Fig. 4.1(a).

formed in HAZ are ner and contain secondary carbide particles

too. Formation of these microstructures can be attributed to
combined effects of dissolution of base metal carbide into the
solution in the austenite due to sufciently high temperature that
prevails in HAZ and repeated hardening due to self-quenching,
which occurs inherently during laser welding. As a result, inner
shell HAZ becomes fully martensitic on cooling, and peak weld
hardness occurs in this region as shown in Fig. 4.2. The carbides
formed in this HAZ are normally of M23C6 types, where the M is
predominantly of Cr and Fe.

5. Conclusions
Using the laser machine and within the limits of the laser
parameters considered in this study the following points can be
concluded:
The developed combined welding and pre-and-postweld
treatment technique is able to overcome the crack formation
problem associated with laser welding of dissimilar stainless
steels.
Laser power and welding speed are the most signicant laser
welding input factors and have opposite effects.
For welding dissimilar stainless steels in a butt joint conguration, weld penetration depth determines the minimum
crack-path of the weld.
Formation of keyhole results in rapid change in the weld
geometrical features within a certain range of energy input. After
the upper limiting value, formation of upper keyhole plasma
plume only contributes to the change in shape of the weld bead.
Various, complex martensiteferrite microstructures develop
in the fusion zone. As a consequence, average microhardness
of the fusion zone becomes much higher than that of base
metal of the outer shell and lower than that of base metal of
the inner shell.
By means of design of experiment inspired by full factorial
design, it is possible to achieve the best operating parameter
window.

Authors would like to acknowledge Continental Cleanliness

and Material Lab for its support while cutting and preparing the
samples for analysis. Thanks are due to M. Fiaschi and F. Sarri,
Industrial Engineering Department for their kind assistance at
various stages of this research. The help extended by V. Colombini
and C. Fierro during experimental investigation is also sincerely
acknowledged.
References
[1] Dawes C. Laser welding. New York, USA: Abington Publishing; 1992.
[2] Sun Z, Ion JC. Review: laser welding of dissimilar metal combinations. Journal
of Materials Science 1995;30:420514.
[3] Kou S. Welding metallurgy. 2nd ed. NJ, USA: John Willey & Sons Inc.; 2002 pp.
199209.
[4] Khan MMA, Romoli L, Fiaschi M, Dini G, Sarri F. Experimental investigation on
laser beam welding of martensitic stainless steels in a constrained overlap
joint conguration. Journal of Material Processing Technology 2010;210:
134053.
[5] Khan MMA, Romoli L, Fiaschi M, Dini G, Sarri F. Laser beam welding of
dissimilar stainless steels in a llet joint conguration. Journal of Materials
Processing Technology 2012;212:85667.
[6] EN ISO 15609-4:2009, Specication and qualication of welding procedures
for metallic materials-welding procedure specication-Part 4: laser beam
welding, pp 113.
[7] Kou S. Solidication and liquation cracking issues in welding. Journal of the
Minerals, metals, and Materials Society 2003:3742.
[8] Lippold JC, Kotecki DJ. Welding metallurgy and weldability of stainless steel.
1st ed. NJ, USA: John Willey & Sons; 2005 pp. 7180.
[9] Lippold JC, Kotecki DJ. Welding metallurgy and weldability of stainless steel.
1st ed. NJ, USA: John Willey & Sons; 2005 pp. 116119.
[10] Lankalapally KN, Tuy JF, Gartner M. A model for estimating penetration depth
of laser welding processes. Journal of Physics. D: Applied Physics
1996;29:183141.
[11] Huang H, Su F, Sun Y. Thermal shock of semi-innite body with multi-pulsed
intense laser radiation. Acta Mechanica Solida Sinica 2010;23(2):17580.
[12] Benyounis KY, Olabi AG, Hashmi MSJ. Optimizing the laser-welded butt joints
of medium carbon steel using RSM. Journal of Materials Processing Technology 2005;164165:9869.
[13] Benyounis KY, Olabi. AG. Optimization of different welding processes using
statistical and numerical approachesa reference guide. Advances in Engineering Software 2008;39:48396.
[14] Montgomery DC. Design and analysis of experiments. 6th ed. New York: John
Wiley and Sons, Inc.; 2004.
[15] Yang YK, Chuang MT, Lin SS. Optimization of dry machining parameters for
high-purity graphite in end milling process via design of experiments
methods. Journal of Materials Processing Technology 2009;209:4395400.
[16] Liao Y-C, Yu M-H. Effects of laser beam energy and incident angle on the
pulse laser welding of stainless steel thin sheet. Journal of Materials
Processing Technology 2007;190:1028.
[17] Anawa EM, Olabi AG. Optimization of tensile strength of ferritic/austenitic
laser-welded components. Optics and Lasers in Engineering 2008;46:5717.
[18] Ruggiero A, Tricarico L, Olabi AG, Benyounis KY. Weld-bead prole and costs
optimization of the CO2 dissimilar laser welding process of low carbon steel
and austenitic steel AISI316. Optical Laser Technology 2011;43:8290.
[19] Zulkali MMD, Ahmad AL, Norulakmal NH. Oryza sativa L. husk as heavy metal
adsorbent: optimization with lead as model solution. Bioresour. Technol
2006;97:215.