Characteristics optimization of different welding processes on duplex stainless

steels using statistical approach and taguchi technique - A review guide

A.Balaram Naik 1, Dr.A.Chennakesava Reddy 2 , Dr.B.Balakrishna 3

1. Sr.Asst.Prof of Mech Engg, JNTUHCE Sulthanpur, Medak Dist. [email protected]

2 .Professor of Mech Engg, JNTUHCE Hyderabad, Kukatpalli, [email protected]

3. Assoc. Professor of Mech Engg, JNTUKCE Kakinada, [email protected]

Abstract

Material characteristics and welding parameters play a very important role in determining the quality of a weld bead.
Duplex stainless steel (DSS) consists of approximately equal proportions of ferrite (δ) and austenite (γ) phases. The
DSS joint quality can be defined in terms of mechanical behavior, heat input, heat-cooling rate, weld oxide
formation, sigma phase formation in weld metal and base metal. Different welding process is used with the aim of
obtained a best welding joint with the desired input parameters, good mechanical properties with minimum defects.
Generally welding thermal cycle, dual torch, aging, heat treatment techniques are more influence on quality of weld
joint. In order to determine optimize process, design of experiment and computational network are widely using
nowadays. A comprehensive review of the application of these techniques in the area of welding on different grades
of duplex stainless steel has been introduced herein. This paper was classified according to the output features of the
weld i.e. to optimize the mechanical properties by changing the relative amount of ferrite, austenite and martensite
content in base metal and weld metal. This review shows many investigations and effects of phase transformation,
heat treatment, dual torch technique on duplex stainless steel weldments and examined microstructure analysis at
different experimental conditions in order to obtain superior welded joints.

1. Introduction understanding of the evolution of microstructure

during welding is need in order to predict the final
Duplex stainless steel(DSS) typically comprises
microstructure consisting approximately equal weld properties. The quality of DSS weld joint is
proportion of Body-centered cubic ferrite (δ) and directly influenced by heat input parameters,
face centered cubic austenite (γ),which offers type of welding process and type of technique,
greater mechanical strength and high corrosion, therefore welding can be considered as a multi-
pitting and fatigue resistance than most types of input multi-out put process. By changing
stainless steel. Duplex stainless steel is a common chemical composition of base metal and filler
structural material in the oil, gas, manufacturing metal of duplex stainless steel, it requires time-
industries and has special application in chemical, consuming trial and error development method
waste water, marine engineering field. An for choosing best weld parameter to produce a
important limitation of the DSS is its maximum welded joint that closely meets the required
temperature of service, below 300 0C,due to the specification of joint. The various optimization
problem of brittleness at 475 0 C.The performance methods of different welding process can be
of DSS can be significantly affected by welding applied to define the desired output variables on
technique-induced phase transformation and DSS through developing design of experiment
(DoE) technique, taguchi technique and

1
 statistical approaches have been used to such 2.1 Ferrite formation
optimization. Sigma phase formation on duplex W.Zhang and T Debray [1] have highlighted to
stainless steel weld metal is greatly influence by developed ferrite formation in duplex stain less
single pass and multi pass gas tungsten arc steel by one-dimensional numerical diffusion
welding (GTA) without filler material in argon model.DSS are two phase alloys consisting of
gas. Activated tungsten inert gas(ATIG) welding approximately equal proportions of ferrite (δ) and
technique, which control the ferrite/austenite austenite (γ) phase with high toughness, good
contents of the weld metal are very important for weldability, satisfactory resistance and high
DSS to avoid degrade the strength and corrosion tensile strength. Investigation on heat affected
resistance. The optimal mechanical properties zone (HAZ) has been done to understand the
were obtained when precipitation of sigma and phase formation, phase fraction, grain size during
the γ-to- δ ferrite transformation are suppressed, the weld heating and cooling cycles.
0
i.e., at 1050 C.Ni percentage on filler metal and By using the numerical diffusion model and
weld oxide formation on DSS also improve the computed heating rates, time-temperature-
quality of weld and develop corrosion fatigue transformation (TTT) and continuous-heating
resistance. Effect of solution heat-treatment on transformation (CHT) diagrams for DSS were
DSS weld metal will increases pitting resistance constructed and the kinetics of the γ-δ
and content of γ-phase in weld bead and in heat transportation at various locations in the HAZ
affected zone (HAZ) increases which promotes could be calculate and validate by comparing the
weldability, excellent mechanical properties. calculated result with data obtained using a
spatially resolved X-ray diffraction (SRXRD)
In this paper a comprehensive detail technique.2205 DSS consists of alternate layers
review of the application of different optimize of 54% δ ferrite and 46 % γ-austenite phases has
technique is presented. This literature review shows been solution annealed at temperature of 1338 K
the relations between the welding input parameters for a period of 2.5 hours following by water
and the output weld bead variables, the paper also quenching to ambient temp to achieve starting
present the characteristic optimization of the microstructure. The initial nitrogen
different welding process on duplex stainless steel concentration and total nitrogen concentration
through experimental microstructure analysis. The can be known using governing equations of
classification of the literature review will be nitrogen diffusion.
according to the weld joint features.
∂N ∂2 N
2. Phase formation in DSS weld bead =D δ in δ (0 < x <
∂t ∂ X2
The performance of duplex stainless steel can be M)
significantly affected by welding induced phase (1)
transformations; hence formation of ferrite, 2
∂N ∂ N
austenite phases in weld metal and base metal =D γ in γ ( M < x <
after performining welding process is very
∂t ∂ X2
important.Controlng of weld-bead phase is xγ ¿
essential as the mechanical properties of welds are Where N is the nitrogen concentration, t is time,
affected by the weld-bead phase. Therefore, it is D δ ∧D γ are the diffusion coefficients of
clear that precise selection of the process nitrogen in the δ and γ phases respectively.
parameters is necessary.
2.2 Thermodynamics of γ-δ transformation

2
 interface on sigma phase formation in a DSS
The equilibrium fraction of (γ) has been weld metal where the chemical element
experimentally measured in the literature [2] by distribution is relatively uniform because of
Hertzman for the number of duplex stainless rapid cooling during weld thermal cycle. The
steels.The relationship of f γ as a function of chemical composition(Wt %) isFe-25.05%,Cr-
temperature and composition is given as 4.7%,Ni-1.90%,Mo-0.4% of 50 ×150×90 mm
in size of base metal plate was welded by sigle
f γ =0.01[75-6.8 pass gas tungsten arc (GTA) welding without
filler material in argon gas. Relationship
×10−15 ×T 5 +190 ( CC −0.03 ) +6 ( 22−Cr C )+ 9 ( ¿C −5 ) +6.5 ( 3−Mo C ) +160( N C −0.15)¿
between the pair of deviation angle ∆ θ K−S of
(2)
intergranular-austenite and the misorientation
at prior ferrite grain boundary in the weld
Where C Cis the carbon concentration, Cr Cis the
metals and relationship between frequency of
chromium concentration ¿C is the nitrogen
α/γ interfaces and the deviation angle ∆θ K−S
concentration. The concentration of C, Cr, Ni, Mo
and N in Eq (2) are the average concentrations of are examined. Sigma phase forms preferentially
elements in the base metal determining by chemical at one side of ferrite/austenite interface along
analysis. It found that 1338 K is only anominal start intergranular austenite in a DSS weld metal
temperature and not the equilibrium temperature for annealed at 1100K.The minimum deviation
the start of the γ-δ transition. angle for sigma phase formation is strongly
influenced by the interfacial energy and the
The spatially resolved X-ray diffraction (SRXRD) coherency of ferrite/austenite interface. High
technique was used to obtain real-time kinetic data coherent and low energy ferrite/austenite
for the γ-δ transformation during welding gas interface can delay sigma phase formation.
tungsten arc (GTA) welds were made on 2205
duplex stainless steel cylindrical bar samples 2.3. Weld oxide formation on DSS
.During welding sample was irradiated with a high-
intensity synchrotron beam, and the diffracted E.M.Westin and O.A.Olsson [6] have
beam was collected to determine the crystal continued their previous study and successfully
structure at discrete location. Additional details of investigated the effect of weld oxide have a
the SRXRD experiments are available elsewhere strong influence on corrosion resistance of
[3,4].It is found that the overall transformation rate Duplex stainless steel. Chromium dominates in
is 30 % faster when the starting structure is the weld oxide formed at high temperature and
uniform. The transformation rate increase with reaches a maximum above 9000C related the
temperature and the non-uniform starting structure chromium reduction in heat tint formed at the
delays the completion of the phase transformation. highest temperatures to evaporation of
chromium oxide, CrO3.Best corrosion
2.3 Sigma phase formation on DSS. resistance accurse by mechanical cleaning
followed by pickling.
Yutaka S.Sato and Hiroyuki [5] have extended Higher alloyed grade steels does not
their study and examined the effect of chemical affect the corrosion resistance as compare with
driving force on sigma phase formation in lower alloyed grade steels. Gas tungsten arc
duplex stain less steel weld metals with welding (GTAW) was performed on different
different chemical compositions. The effort of welded duplex stainless steel grades (LDX
crystallographic orientation relationship at δ/γ 2101, 2304, 2205 and 2507) and followed by

3
 corrosion testing,ferroxyl test, electron probe Post weld heat treatment (PWHT) is very
micro analysis-ray photo electron spectroscopy important to control over the balance of a α, γ
test were carried out to examined the effect on phases in the weld metal (WM) and HAZ.
corrosion resistance and weld oxide formation. Nitrogen (N) is a strong γ-stabilizer and it
The influence of nitrogen, nickel, increase the temperature at which the
manganese on heat tint formation is more. transformation of the α-to the γ phase occurs.
After examined all experimental test on all
samples it shows that discoloration only for the PREN = wt % Cr + 33 (wt % Mo + 0.5 × wt %
welds in nitrogen-containing atmosphere, not W) + 30 × wt. %N.
those with argon. The reason seems to be the (3)
manganese or chromium oxynitrides formed in
the former case and appear to be responsible The formula in Eq (3) for DSS has been used by
for the staining. several researchers to investigate DSS’s
resistance to localized corrosion [8].
3. Heat treatment analysis on DSS.
Metallographic examination, polarization test,
critical pitting temperature test, scanning electron
Welding is usually done with the aim of
microscope and energy dispersive spectroscope
getting a welded joint with best mechanical
(SEM-EDS) analysis and transmission electron
properties.
microscope (TEM) analysis of Cr2N precipitate
On any type of material, the property of base
were carried out and examined the effect of heat
matel and quality of welded joint is greatly
treatment on the resistance to pitting corrosion of
influence by the heat treatment process. The
a HDSS welds. The phase diagram and
effect of heat treatment also have great change
equilibrium fraction of each phase were
in microstructure and it balance the ferrite,
calculated against the temperature for the HDSS
austenite content in the weld metal and in heat
alloy using a commercial thermo calculation
affected zone (HAZ) which promotes
software package. Based upon the PRENγ and
weldability and excellent mechanical
PRENα [9] values calculated using a N factor of
properties.
30,pitting corrosion in the weld metals and
HAZ’S in the solution heat-treatment HDSS tube
3.1 Effects of solution heat-treatment on DSS
after welding was selectively initiated at the α-
welds.
phase, irrespective of the chemical compositions
Soon-Tae Kim and Seok-Hwan Jang [7] have of the shielding gas. The pitting corrosion was
presented an experimental analysis of the effects finally propagated from the α-phase to the γ-
of solution heat-treatment on hyper duplex stain phase.
less steel (HDSS) weldments.HDSS such has
3.2 Effects of Aging on DSS weldments.
UNS 532707 is defined as a highly alloyed DSS
with a PREN Value is greater than 45.After
Omyma Hassan Ibrahim and Ibrahim Soliman
performing a solution heat-treatment on hyper
Ibrahim [10] have explain about the effect of
DSS weld metal with an Ar shielding gas
aging heat treatment at (650,750 and 850 0C) on
supplemented with N2,the pitting resistance-
2205 duplex stain less steel after performing TIG
phase in the weld metal and in heat affected
(Tungsten inert gas) welding technique.DSS
zone was increase which promotes
weldments is the degradation of corrosion and
weldability,excellent mechanical properties.
mechanical properties with in certain high

4
temperature due to the microstructural changes.In Material is divided into several blooms
comparision to austenite stainless with a dimension of 150mm×100mm ×42mm is
steel,precipitation of sigma phase in DSS accurs heated in vacuum furnaces from 900 to 1200 0C
within the ferrite phase at shorter time,at higher temperature range and then reheated at 1250 0C
temperature and with larger volume fractions. for 2 hours and hot-rolled in to 12 mm thickness
Material with dimension of 55 mm plates. Chemical composition of DSS.
×10 mm × 4.5 mm were subjected to solution
Chemical composition of two kinds of duplex stainless steel 2304 (Wt %)
heat treatment at 10500C and 11500C respectively
for 1 hour and then water quenched and then
isothermal aging treatment was applied to base Mo N N Cu C Mn Si P S Fe
and weld metals at 6500,7500 and 8500C for 3hour
followed by furnace cooling. Microstructure
analysis, impact properties, load time plots for 23.06 0.31 0.13 4.03 0.29 0.021 1.56 0.40 0.010 0.004 Bal
austenite stainless steel and duplex stainless steel
weldments were verified and found that impact
energy is higher for DSS in the un-aged 23.82 0.91 0.18 4.50 0.35 0.025 1.52 0.38 0.008 0.004 Bal

conditions. The degree of embrittlement is greater

with the increases of the aging temperature.SEM
The above two grade specimens were solution –
examination reveals that the reduction in impact
annealed for 12 mints at 1250 0C and 10500C
energy due to aging is associated with the
respectively quenched in water. Electrochemical
formation of shallow ductile dimples and
measurements-critical pitting temperature test is
cleavage facets on the fracture surface of 316 L
done at least three times for the same specimen
and duplex stainless steels, respectively.
and microstructure analysis also carried and it so
that all the welding thermal simulated specimens
3.3 Welding thermal cycles on DSS.
exhibited impaired microstructure and pitting
corrosion resistance compared with the
HuaTan and Zhiyu wang [11] have investigated
corresponding base metal. As welding pass
the welding thermal cycle on 2305 DSS. Duplex
increases CPT value increases the ferrite phase
stainless steel are widely used as alternatives to
fraction decrease and improvement of the
austenite stainless steels and nickel based alloys,
microstructure and pitting corrosion resistance.
which consists of approximately equivalent
PREN (pitting resistance equivalent number)
amount of γ and α-ferrite with an excellent
increases.2304-B DSS with higher alloyed
combination of mechanical strength and
specimens showed better microstructure and
corrosion. Different welding thermal cycles from
pitting corrosion resistance than the
single-pass to triple pass were performed on two
corresponding 2304-A specimen.
kinds of 2304 DSS. The most typical problems of
welded DSS are associated with heat-affected
3.4. Hydrogen-enhanced cracking of DSS Welds
zone (HAZ), not with weld metal zone (WMZ),
since the properties of WMZ could be modified
M.C.Young and S.L.I.Chan [12] have continued
by using high-alloyed filler metal and N2
their investigate the effects of hydrogen, nitrogen
containing shield gas. More sophisticated and
addition during laser welding to improved the
complicated microstructure evolution in HAZ
weldability of duplex stainless steel. The
during multi-pass welding.
susceptibility decreased with increasing austenite
content in the weld metal of DSS. Slow

5
displacement rate tensile tests were carried out to energy input in welding. After performing
investigate the effect of hydrogen embrittlement welding on four different size of DSS pipe,
on notched tensile strength (NTS) and fracture metallographic test done by scanning electron
characteristics of 2205 DSS weld. Plastic microscope (SEM) and energy dispersive X-ray
deformation occurs more easily in low alloy DSS, spectrometer (EDX) test were performed to
resulting growth of stress corrosion cracks in the reveal the chemical composition of the phases
austenite. In highly alloyed DSS the strength of present in the microstructure and polarization
austenite is higher and cracking in the ferrite in curves of various specimens can be analyzed.
chloride solution is promoted. By controlling the The results of various test shows that
heat input and interpass temperature during formation of sigma and Cr2N phases reduced
welding are required to correct the α/γ ratio in the corrosion potential. The lower austenite content
DSS weld. Laser welding offers many advantages due to the high heat input and sufficient time for
over the conventional arc welding process on ferrite-austenite transformation.
DSS, by varying welding parameters γ phase
content in the fusion zone of a 2205 DSS is 4. Mechanical properties.
increases.
Microstructure observations, charpy Welding is usually done with the aim of getting a
impact test, notched tensile test, fracture welded joint with excellent mechanical
morphology were carried out after laser welding properties. By changing welding technique, input
on DSS and found the α/γ ratio in the fusion zone parameters, heat treatments, chemical
was drastically increased. The effect of N composition of duplex stainless steel, the welded
addition during laser welding also improved the joint may be examined and investigate to improve
weld’s ability to resist impact fracture. Hydrogen their mechanical properties. To determine these
embrittlement susceptibility is more obvious for welding combinations that would lead to
the specimens containing a greater amount of α excellent mechanical properties for DSS
phase.SEM fractographs shows that all specimens weldments. Different technique, methods,
underwent a significant change in fracture mode approaches and composition have been used to
from ductile in air to mainly quasi-cleavage achieve this aim. The following is a review of
fracture in H2. some articles that utilized these techniques for the
3.5. Influence of heat input on DSS welds. purpose of optimizing the welding process for
DSS in order to achieve the desired mechanical
M.Yousefieh and M.Shamanian[13] have properties of the welded joint.
presented an experimentally the effect of heat 4.1. Mechanical behavior in annealed DSS welds.
input variations on the microstructure of a duplex
stainless steel UNS S32760 in artificial sea water Riad Badji and Mabrouk Bouabdallah [14] have
media. Approximately 0.95 k J / mm of heat examined the effect of annealing treatment of
input have the best corrosion characteristics 2205 DSS on mechanical properties. The tube
which are the result for the lack of deleterious type DSS material of 170 diameter and 7 mm in
phases such as sigma and Cr2N and balanced wall thickness is welded using a gas tungsten arc
ferrite –austenite proportion. The pulsed current welding process in multi pass with Ar+2%N 2
of gas tungsten arc welding (GTAW) process has shielding gas and ER 2209 filler metal. After
a numerous advantages over the conventional welding, series of heat treatment were performed
GTA process, the ferrite-austenite ratio in base from 80000C to12000C and then etched with
metal as well as in weld metal depends on the

6
solution like 10 ml HNo3, 20ml glycerol and 30 As a result arc voltage increases and the ferrite
ml Hcl. content in the welded bead is decreases.
By ZEISS optical microscope, optical
micrograph of welded specimen over heated parts 4.3. Corrosion fatigue of DSS welds.
of the HAZ, and partially annealed region of the
HAZ at different ranges was done. The EDX S.A Tavara and M.D.Chapetti have stated that the
analysis technique has been used to identify the duplex stainless steel of two phase microstructure
phases constituting each microstructure region. austenite and delta ferrite in approximately
The best combination of strength stress σ =775 similar percentages have greater yield strength
MPa and elongation δl=36.45 % is obtained after than austenitic stainless steels. Welded parts has
annealing at 10500C.Asignificant change in the less resistance than base metal by control of
toughness and tensile properties has been carbon content(low carbon) and welding
observed as a consequence of precipitation conditions to ensure austenite formation can
phenomena or an increase of ferrite content. The make corrosion resistance of DSS equivalent to
optimal mechanical properties were obtained base metal.
when precipitation of sigma and the γ-to-δ ferrite Austenite/ferrite contents of welded
transformation are suppressed i.e. at 10500C. DSS is dependent upon electrode composition, Ni
plays an important role in the balance as austenite
4.2. Activated TIG welding on DSS. promoter. Corrosion fatigue in welded joints is
considered a very dangerous type of propagation
Tsann-Shyi Chern and Kuang-Hung Tseng [15] of defects in materials that are subjected to
have continued their study and examined the dynamic loads.Welding has performed on four
tensile strength. The effect of the specific fluxes different composition of base metal with four
used in the Activated tungsten inert gas welding different filler matel (electrode A,B,C,D) and
process on 6 mm thick duplex stainless steel also tested mechanical properties after heat treatment
investigated. Penetration during TIG welding was made to 10500C for 1 hour.Three point
process depends upon weld current, weld speed, bending fatigue test were carried out in a load
thickness of matel, material, flux used and type of control walking –beam fatigue machine in air and
electrode. By using activated flux, the 200%- sea water of different frequency and found that
300% deep penetration takes place. The Ni content has a direct relation to the number of
technique that control the ferrite /austenite cycles necessary to reach a depth of 1.5 mm.The
contents of the weld metal are very important for environmental contribution to the crack growth
DSS to avoid degrade the strength and corrosion rate is:
resistances. Effect of activated TIG welding on (da/dN)cf = (da/dN)e – (da/dN)r
(surface, weld morphology, angular distortion, Where (da/dN)e is the measured crack growth
ferrite and austenite content) with flux and rate and (da/dN)r is the reference crack growth
without flux has be experimental examine was rate, corresponding to pure mechanical fatigue
done and finally mechanical properties found are in air.
tensile strength σ =765 Mpa, and 35 %
elongation. By using SiO2, MoO2 and Cr2O3 flux 5. Optimization techniques.
penetration capacity increased and improved
mechanical strength. The greater weld depth-to- Different methods, approach and techniques
width ratio take place with activated TIG welding have been used to achieve excellent weld bead
process as compare with the normal TIG welding. with best tensile strength on DSS, Low carbon

7
steel, high carbon steel. The follow is a review the effect of Nd-YAG laser welding parameters
of some researches that utilized these (shielding gas type, laser energy, conveying
techniques for the purpose of optimizing the speed, laser focus, pulse frequency and pulse
welding process. shape) on the ultimate tensile stress. Their result
indicated that the pulse shape and energy of the
5.1 Taguchi technique. laser contributed most to thin plate butt-welding.
It was found that the optimal combination of
Juang and Tarng [34] have adopted a modified welding parameters for laser welding were argon
Taguchi method to analyze the effect of each as a shielding gas, a 360 W laser energy, a
welding process parameter (arc gap, flow rate, workpiece speed of 25 mm/s, a focus distance of
welding current and speed) on the weld pool 0 mm, a pulse frequency of 160 Hz and type III
geometry (front and back height, front and back pulse shape. It was also found that the superior
width) and then to determine the TIG welding ultimate tension stress was 169 MPa at an
process parameters combination associated overlap of the welding zone of approximately
with the optimal weld pool geometry. It was 75%.
experimentally reported that, the four smaller-
the-better quality characteristics,‘ four 5.2 Response surface methodology (RSM).
responses’ of the weld pool in the TIG welding
of S304 stainless steel of 1.5 mm in thickness Wang and Rasmussen [53] have investigated
are greatly improved by using this approach. the inertia welding process of low-carbon steels
using RSM, with the purpose of establishing an
empirical functional relationship between the
Lee HK and Han HS. [35] have used the process parameters (the axial pressure, the initial
Taguchi method and regression analysis in rubbing velocity and the total moment of inertia)
order to optimize Nd-YAG laser welding and the breaking strength of the joint. It was
parameters (nozzle type, rotating speed, title concluded that a relatively wide range of
angle, focal position, pumping voltage, pulse operating conditions would produce successful
frequency and pulse width) to seal an iodine- welds. Also, they observed that the average
125 radioisotope seed into a titanium capsule. microhardness at the weld was about 27%
The accurate control of the melted length of the higher than the base material and the ideal weld
tube end was the most important to obtain a should be made with the least possible amount
sound sealed state. It was demonstrated that the of kinetic energy as long as full penetration at
laser pulse width and focal position were the the interface is achieved.
laser welding parameters that had the greatest
effects on the S/N ratios of the melted length. Benyounis and Bettamer [56] have proposed
The optimal welding conditions were obtained models using RSM to investigate the effect of
at a pulse width of 0.86 ms and a focal position welding parameters in SAW (welding current,
of 3.18–3.35 mm. Furthermore, confirmation arc voltage and welding speed) on the impact
experiments were conducted at the optimal strength at two testing temperatures of 50 _C
welding conditions, it can be said that the and 27 _C. The aim was to predict and optimize
titanium tube ends were sealed perfectly the impact strength of the spiral-welded joints
with respect to the process parameters. It was
Laser butt-welding of a thin plate of observed that the welding current was the most
magnesium alloy using the Taguchi method has significant factor associated with the impact
been optimized by Pan et al. [71]. They studied

8
strength, then the welding speed, whereas the (a plasma welding torch followed by a gas
welding voltage has no significant effect within tungsten arc (GTA welding torch)
the factors domain investigated. They listed the simultaneously during welding.SAF 2205 DSS
optimal welding conditions that would lead to with 5.5 mm thickness material were performed
acceptable impact strength with improving the dual torch technique, potentiodynamic
process productivity measurements and electro chemical impedance
spectroscopy were performed. Polarization
5.3 Artificial neural networks (ANN). curves of the base metal and weld metal are
examined at different torch pitch ( 25,30 and 35
Control of distortion and overall quality of mm).Corrosion rates for the DSS welds also
welds were investigated by Casalino et al. [65] examined and demonstrated the potential of
in order to select the GMAW process parameters using dual torch plasma welds to improve
that minimize thermal deformation and evaluate weldability of duplex stain less steels.
weld quality. They integrated the artificial
intelligence techniques and FEM with the aid of 6. Comparison between the optimization
experimental trials of bead-on-plate welds. The techniques
base metal was 1.6 mm thick low-carbon steel, a
Derived from the above literature review
0.9 mm diameter copper- coated wire was used
some insight has been gained into the use of
as an electrode with a shielding gas consisting of
DoE, ANN, Dual torch technique, Taguchi
a 75% Ar–25% CO2 mixture with flow rate of
method and other techniques for modelling and
10–15 ft3/h.
optimizing different welding processes on DSS.
ANN was used at first to link the process
It was noted that Taguchi performs better than
parameters to the geometry of the molten zone,
other techniques, especially ANN and RSM,
which allowed the geometries throughout a
when a large number of experiments are not
range of process parameters to be calculated.
affordable. The trend in the modelling using
Then FEM was applied to predict the residual
Taguchi method has a low order non-linear
stress value and distortion in the welded joint.
behaviour with a regular experimental domain
Finally, fuzzy C-means clustering algorithm was
and relatively small factors region, due to its
applied to evaluate the quality joints.
limitation in building a model to fit the data over
Mathematical models for GMAW were
an irregular experimental region. This ability is
constructed. Experimentally butt welded joint
powerful in identifying the insignificant factors
were validated. It was concluded that the
main effect, insignificant interactions or
experimental result are in good agreement with
insignificant quadratic terms in the model and
the mathematical model.
thereby can reduce the complexity of the
problem. On the other hand, this technique
5.4. Dual torch technique on DSS welds.
required good definition of ranges for each
factor to ensure that the response(s) under
Z.Sun and M.Kuo have investigated the
consideration is changing in a regular manner
effect of the dual-torque technique on the
within this range.
microstructure changes and corrosion properties
The Taguchi method is one of the
of duplex stainless steel were examined. Ferrite
powerful optimization techniques comparing
percentage is very important to reduce stress- with RSM, ANN and Dual torch technique
corrosion cracking resistance, therefore to which characterize with improving the product
control this technique uses two welding torches’ quality and reliability at low cost. However,

9
RMS analysis approach of SN may lead to non-
optimal solutions, less flexibility and the
conduction of needless experiments.

6.1 Comparison between RSM and ANNs

Techniques.
Table 1
Comparison Technique
ANNs RSM
Time taken long Medium
Experimental irregular Irregular
domain
Understandin Difficult Moderate
g
Optimization Through Straight
model
Availability in Available Available
software
Application Rarely Rarely

6.2 Comparison between Taguchi and Dual

torch Techniques.
Table 2
Comparison Technique
Taguchi Dual torch
Time taken Short Long
Experimental Regular Regular
domain
Understandin Normal Easy
g
Optimization Through Straight
model
Availability in Available Available
software
Application Frequently Not frequently

Table 1 and Table 2 presents a comparison

between the above mentioned common
modelling /optimizing methods based on this
literature review.

7. Conclusion

10
11