See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/226945864

Haq, Effect of shielding gases on mechanical and metallurgical properties of

duplex stainless-steel welds

Article  in  Journal of Materials Science · January 2008

DOI: 10.1007/s10853-008-3098-8

CITATIONS READS

87 367

4 authors:

Paulraj Sathiya S. Aravindan

National Institute of Technology Tiruchirappalli Indian Institute of Technology Delhi
118 PUBLICATIONS   2,158 CITATIONS    202 PUBLICATIONS   4,259 CITATIONS   

SEE PROFILE SEE PROFILE

Soundararajan .R A. Noorul Haq

Sri Krishna College of Engineering and Technology National Institute of Technology Tiruchirappalli
94 PUBLICATIONS   597 CITATIONS    120 PUBLICATIONS   4,961 CITATIONS   

SEE PROFILE SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Thermal Fatigue View project

CMT welding View project

All content following this page was uploaded by Soundararajan .R on 10 March 2020.

The user has requested enhancement of the downloaded file.

J Mater Sci (2009) 44:114–121
DOI 10.1007/s10853-008-3098-8

Effect of shielding gases on mechanical and metallurgical

properties of duplex stainless-steel welds
P. Sathiya Æ S. Aravindan Æ R. Soundararajan Æ
A. Noorul Haq

Received: 7 August 2008 / Accepted: 30 October 2008 / Published online: 5 December 2008
Ó Springer Science+Business Media, LLC 2008

Abstract Duplex stainless steels (DSS) are well known Introduction

for their higher mechanical strength and better corrosion
resistance. DSS is commonly used in the marine construc- Tungsten inert-gas (TIG) welding uses a nonconsumable
tion, petrochemical, and chemical industries. DSS (2205) tungsten electrode and an inert gas for arc shielding. It is
has equal amounts of a and c phases. However, unlike the commonly used for fabricating joints of difficult-to-weld
parent metal, the solidification microstructure of the fusion metals such as aluminum and stainless steels. Since the
zone in the weld does not have nearly equal amounts of the a weld pool geometry plays an important role in determining
and c phases. Thus the mechanical properties and corrosion the mechanical properties of the weld, this aspect strongly
resistance of DSS welds are degraded. The interpass tem- influences TIG weld quality [1]. A nonconsumable tung-
perature plays a vital role in achieving balanced a and c sten electrode, shielded by an inert gas, is used to strike an
phases, which in turn results in improved mechanical and electric arc with the base metal (BM), providing the heat
corrosion resistance. Gas tungsten arc welding (GTAW) is necessary to melt the BM. The quality of TIG weld is
employed for welding of thin sheets/plates. The GTAW higher than that of any of the arc welding processes due to
process ensures small amounts of slag formation during the reliability and strength of the welded joints. The gas
welding, which eliminates slag crevices and sites for cor- tungsten arc welding (GTAW) process ensures small
rosion attack. Standard 2205 (UNS S31803) DSS sheets of amount of slag formation during welding, which eliminates
5 mm thickness, with 22.37% Cr and 5.74% Ni, were used slag crevices and sites for corrosion attack [2]. Duplex
in this study. Weld beads were produced with Zeron-100 stainless steels (DSS) consist of nearly equal amounts of
super DSS filler wire with higher alloy content (26% Cr and ferrite and austenite phase at room temperature. DSS
8% Ni). Argon (Ar) and helium (He) were employed as exhibits greater toughness and better weldability than fer-
shielding gases. Heat input was maintained \1 kJ/mm. ritic stainless steel. DSS 2205 provides pitting and crevice
Interpass temperatures were maintained at 120°C. The corrosion resistance superior to that of 316L or 317L
ferrite number of the weld metal for the two different austenitic stainless steels in almost all corrosive media. The
shielding gases was investigated. Mechanical properties of yield strength of DSS is about twice that of austenitic
joints such as impact strength and hardness were evaluated. stainless steel. This allows a designer to save weight and
Microstructure evaluation was also carried out. makes the alloy more cost competitive when compared
with 316L or 317L austenitic stainless steels. DSS is
mainly used in the marine industry, in cargo tanks for ships
and trucks, for power generation applications, and in
P. Sathiya (&)
R. Soundararajan
A. Noorul Haq
Department of Production Engineering, National Institute pressure vessels, tanks piping, and heat exchangers in the
of Technology, Tiruchirappalli 620015, Tamil Nadu, India chemical processing industry. Though the cost of pro-
e-mail: psathiya@nitt.edu cessing of DSS is lower than that of austenitic stainless
steel, maintaining equal amounts of ferrite and austenite
S. Aravindan
Department of Mechanical Engineering, Indian Institute phases is crucial. The Schaeffler diagram and WRC-1992
of Technology Delhi, New Delhi 110016, India can be used for prediction of the ferrite content in the weld

123
J Mater Sci (2009) 44:114–121 115

metal. The Schaeffler diagram, now 50 years old, is out of Experimental details
date for ferrite prediction in stainless-steel welds. Since
1995, the WRC-1992 diagram has been recommended for The size of the base material chosen for this investigation was
ferrite prediction. This diagram is based on the nickel and 90 9 45 9 5 mm3 sheets corresponding to UNS S31803,
chromium equivalent to asses the effects of addition of with chemical composition as given in Table 1. Zeron-100
alloying elements on microstructural constituents of stain- filler wire with a diameter of 3 mm was used in this study;
less-steel weld metal [3]. its chemical composition is also presented in Table 1.
The main advantages of DSS are the equal amounts of a Welding was performed with two different shielding
and c phases present in the steel. Unlike the parent metal, gases (argon and helium) with two different welding
the weld metal microstructure dose not have equal amounts parameters. The welding parameters were chosen in such a
of a and c phases. For this reason, the mechanical prop- way as to keep the heat input below 1 kJ/mm. The details
erties and corrosion resistance of the DSS weld is of the welding parameters are presented in Table 2.
degraded. This can be attributed to the reduction of the c The interpass temperature was maintained as 120°C.
phase in the fusion zone. Interpass temperature is very The ferrite content of the weld metal was estimated by the
important to achieve a structure with balanced a and c point count method. The ferrite percentages were deter-
phases. The ferrite-to-austenite ratio also depends on the mined metallographically using the point count method by
heat input of the welding process. It has been reported that following ASTM E 562 standards. Metallographic exam-
the problem of low fraction of c phase in the weld metal inations were carried out by using optical microscopy. For
can be alleviated by addition of nitrogen gas with different metallographical examination, the samples were polished
combinations in the plasma and laser beam, or by alloy and electrolytically etched in 40% KOH solution. This
addition in filler materials and even by carrying out post- etching was performed at 2 V for 15 s. In all cases, the
weld heat treatment in the a/c coexistence temperature austenite phase is seen as white while the dark phase is
range of 800–1,150°C [4]. ferrite. The microhardness of weldment was measured
Faster cooling rate significantly affects the fusion zone from the BM, across the heat-affected zone (HAZ) to the
microstructure in pulsed GTAW welding for the same heat weld zone (WZ), in transverse direction using a Metco
input condition [5]. Nickel is very effective in controlling SMV 1000 micro Vickers hardness machine under 1-kg
the ferrite-to-austenite ratio compared with the cooling rate load, maintained for 15 s. Charpy impact tests were per-
for the range of compositions studied [6]. Ferrite contents formed to asses the notch toughness of samples extracted
greater than 90% have been reported in electron beam from the weldment. Notches were made such that fracture
welding (EB) in UNS 31803 [7]. It is thus desirable to occurred only with in the weld fusion zone. The impact
control welding conditions such that cooling should be tests were performed both at room temperature and at
slow enough for adequate austenite formation, but fast -40°C on subsized samples (55 9 10 9 4 mm3). The
enough to prevent deleterious precipitation. An acceptable fractured surfaces of impact tested surfaces were studied
phase balance may still be achieved, however, by the use of using scanning electron microscopy (SEM) fractography.
preplaced nickel foils [8] or through post-weld solution
treatment, as recommended for laser welded super DSS [9].
Table 2 Welding parameters
Interpass temperature should be maintained at a specific
value in order to alleviate the faster cooling rate over the Argon Helium
temperature range 300–500°C, which is the critical region Current intensity (A) 135 140
for phase transformation to occur in selective identification
Voltage (V) 11 12
feature (SIF) 2205 DSS [10, 11]. This paper describes the
Weld speed (mm/min) 100 110
effect of two different shielding gases (argon and helium)
Heat input (kJ/mm) 0.891 0.916
on mechanical and metallurgical characteristics of the
Gas flow rate (lpm) 10 10
welded joint.

Table 1 BM, filler wire (FW), and weld metal chemical composition (wt%)
Element (%) C S P Si Mn Cr Ni Mo N Fe

Base material 0.020 0.025 0.021 0.4 1.52 22.37 5.74 2.85 0.171 Balance
Filler wire 0.013 0.001 0.017 0.51 1.51 26.0 8.0 3.14 0.15 Balance
Weld metal—Ar 0.018 0.020 0.020 0.44 1.50 22.82 7.84 3.01 0.146 Balance
Weld metal—He 0.020 0.022 0.019 0.48 1.54 22.60 8.02 2.95 0.158 Balance

123
116 J Mater Sci (2009) 44:114–121

Results and discussion higher arc energy. Compared with argon, helium provides
better side-wall penetration and greater welding speeds by
The microstructure of as-received 2205 DSS as observed generating an energy-rich arc.
optically consists of elongated a and c grains, as shown in
Fig. 1; the morphology of the a–c grain boundary is relatively
Effects of chemical composition on DSS weld metal
smooth. The elongated a and c phases have a grain size of
about 45–50 lm in length and about 15 lm in width. The
The ferrite-to-austenite ratio in DSS weld metal is deter-
measured volume fractions of a and c phases are almost equal.
mined by the chemical composition and the cooling rate
A macrograph of the welded samples, with argon and
experienced during welding. Cooling rate is a function of
helium shielding, is presented in Fig. 2. The helium-shielded
weld geometry. Heat input and preheat temperature depend
weld bead has a wider width compared with that using argon.
on the type of welding and the process parameters
In order to improve understanding, weld bead profiles
employed. Filler wire chemical composition, the nature of
were measured by using a ‘‘cam scope’’ and the profiles are
the shielding gas, and the degree of dilution determine the
presented in the Fig. 3a and b. From Fig. 3a it can be
composition of the weld metal. Two different shielding
observed that the bead width of the argon-shielded weld is
gases were used and the corresponding weld metal chem-
12.43 mm and the depth of penetration is 4.499 mm. From
ical compositions are presented in Table 1.
Fig. 3b it can be observed that the bead width of the
From Table 1 it can be seen that the fraction of ferrite-
helium-shielded weld is 18.82 mm and the depth of pen-
stabilizing elements such as Cr and Mo is greater in the
etration is 4.556 mm. The greater penetration and wider
argon-shielded weld whereas the fraction of the austenitic-
bead width achieved by using helium shielding is due to the
stabilizing elements Ni and N is greater in the helium-
shielded weld metal. All DSS weld metals solidify primarily
as ferrite and, in the subsequent solid-state transformation,
part of the ferrite transforms to austenite. This ferrite-
to-austenite transformation takes place over a range of
temperatures. However, the transformation cannot attain
equilibrium due to rapid cooling, so that more ferrite
remains untransformed than under equilibrium conditions.
The transformation of ferrite to austenite during the welding
of DSS is similar to that in low-carbon-steel weld metals in
which austenite transforms to ferrite on cooling. However, a
major difference exists between the transformations in
low-carbon steels and DSS. In low-carbon steels all of the
austenite is transformed to a combination of ferrite, bainite,
and martensite, except for a small amount of retained aus-
tenite when present. However, in DSS, a considerable
Fig. 1 BM microstructure volume of ferrite remains untransformed. The solidification
models of the two investigated shielding gas weld metals
were calculated from their corresponding chromium and
nickel equivalents. The solidification mode for both of the
shielding gases was ferrite.
The Cr and Ni equivalents were calculated by using the
following formulae (with values presented):

Creq ¼ %Cr þ %Mo þ 0:7% Nb

ðArgon - 25:83 & Helium - 25:55;
Creq =Nieq ¼ 2:26Þ;
Nieq ¼ %Ni þ 35% C þ 20% N þ 0:25% Cu
ðArgon - 11:39 & Helium - 11:88;
Creq =Nieq ¼ 2:15Þ:

From these values, it can be clearly seen that the

Fig. 2 Macrograph of the weld joint chromium and nickel equivalents are higher in the

123
J Mater Sci (2009) 44:114–121 117

Fig. 3 Weld bead profile:

a argon shielding gas and
b helium shielding gas

argon-shielded weld metal. Due to the high Creq/ ferrite content calculated from the Creq and Nieq values is
Nieq ratios, a large amount of ferrite is present in the also presented.
argon-shielded weld. This is further confirmed by
superimposing the Creq/Nieq ratios of the weld metals Charpy V-notch impact test
on a pseudobinary diagram, which confirms that the mode
of solidification is ferrite. The WRC-1992 (Fig. 4) The Charpy impact toughness of the DSS parent material is
diagram is used for estimating weld-metal ferrite on the 85 J. The TIG welded specimens were tested at room and
basis of chemical composition. The ferrite content based subzero temperature (-40°C). The toughness of the test
on the ferritoscopic analysis for both of the shielding results welded with different shielding gases is tabulated in
gases is indicated in Fig. 4. In the same diagram the Table 3.

Fig. 4 WRC-1992 diagram

[12]

123
118 J Mater Sci (2009) 44:114–121

Table 3 Impact test results DSS weldment deteriorates with increasing volume of
Temperature (°C) Impact values (J)
a-ferrite within the HAZ [14]. Careful control of heat input
and interpass temperatures during welding are required to
Argon gas Helium gas correct the a-to-c ratio in the DSS weld. Less heat input
25 87 92 results in a faster cooling rate, preventing the transforma-
-40 65 78 tion of primary a into c phase in the weld [15]. The low c
content in the fusion zone is responsible for its poor impact
toughness [16]. The results of impact SEM fractography
Three samples were tested at room temperature and at are presented in Fig. 5a–d. For all of the impact fractured
-40°C, and the average values are tabulated in Table 3. specimens, the grains are elongated in the grain-boundary
The helium-shielded weld sample exhibited higher tough- direction and fine dimples are present on the fractured
ness than did the argon-shielded weld. The reasons for this surfaces, revealing that the ductile fracture mode with fine
can be attributed to the cooling rate, the larger amount of dimples occurred in the fracture of welds with both
arc energy, and the presence of larger amounts of Mn in the shielding gases.
WZ. In general, increase of the ferrite content of the weld
metal decreases its toughness. Argon-shielded welds have a Hardness testing
larger amount of ferrite phase whereas helium-shielded
welds have a greater amount of austenite-stabilizing ele- The samples were prepared and electrolytically etched with
ments such as Ni and N. Hence, the helium-shielded weld 40% KOH solution. Three distinguished zones were iden-
has a greater amount of austenite phase and exhibits higher tified and a set of ten microhardness readings was taken in
toughness than the argon-shielded weld. Manganese is one each. The average values are presented in Table 4.
of the powerful alloying elements for improving the From Table 4 it can be clearly seen that the hardness of
toughness of the weld metal. The observed differences in the weld metal is slightly higher than that of the HAZ and
toughness were not only due to changes in the ferrite the base material for both the argon- and helium-shielded
content, but also due to the reduction in the ferrite-to- cases. The hardness of the argon-shielded weld is less than
austenite ratio. The decrease in toughness values at the that of the helium-shielded weld, since the Ar-shielded
subzero temperature for both of the weld metals can be welds have less austenite phase compared with the helium-
attributed to the presence of ferrite, which suffers from a shielded weld metal. It has been reported that the ferrite
ductile brittle transition [13]. The impact toughness of the and austenite phases do not differ greatly in composition

Fig. 5 SEM fracture surface of

impact tested specimens.
a -40°C with argon shielding
gas. b Room temperature with
argon shielding gas. c -40°C
with helium shielding gas. d
Room temperature with helium
shielding gas

123
J Mater Sci (2009) 44:114–121 119

Table 4 Average microhardness values method using a Fisher ferrite scope. The advantage of the
Serial no. Shielding gas Microhardness value (VHN)
point counting method is that it can be applied to all
microstructures, including the narrow HAZ. However, it is
BM HAZ Weld metal difficult to apply in situ and is relatively expensive. In this
1 Argon 173 259 308 work Fisher ferrite scope was used. In the case of the
2 Helium 174 283 325 argon-shielded GTA weld, the ferrite content is estimated
to be 55% while the remaining 45% in the weld metal is
austenite. Due to nitrogen loss as well as the lower amount
because substitutional elements do not have time to parti- of austenite-stabilizing elements, as seen in Table 3, the
tion significantly during DSS welding. amount of austenite phase is lower in the argon-shielded
The authors have conducted investigations on several weld metal. In the helium-shielded weld metal the amount
commercial and experimental alloys to understand the of austenite-stabilizing elements was higher than in the
effect of nitrogen on strengthening of DSS and concluded argon-shielded weld metal. A similar effect is also
that, if the fraction of nitrogen in DSS is \0.13%, ferrite is observed in the microstructure; as shown in Fig. 7c and d
harder than austenite, whereas for a nitrogen content of the austenite phases comprised 65% of the helium-shielded
0.13% in DSS, the hardness levels of austenite and ferrite weld metal and while the remaining 35% was ferrite phase.
are equal and austenite becomes harder than ferrite when
the nitrogen content exceeds 0.13% [17]. In the present Microstructural studies
study the nitrogen content of the weld metals for both the
argon- and helium-shielded weld metals exceeded 0.13%. The microstructure of the welds with different shielding
Therefore, in these weld metals austenite is harder than gas is presented in Fig. 6a–d.
ferrite. When using pure argon as the shielding gas in GTAW
joints, the result is normally nitrogen loss and a high
Ferrite measurement amount of ferrite-stabilizing element present. Due to this
the balance between the two phases is not even. As shown
There are two methods for measuring the ferrite content of in Fig. 6a, a large amount of ferrite phase is present in the
weld metals: the point count procedure and a magnetic microstructure. As an effective stabilizer of austenite,

Fig. 6 Weld metal and HAZ

microstructures. a Argon shield,
weld metal (2009). b Argon
shield, HAZ (2009). c Helium
shield, weld metal (2009).
d Helium shield, HAZ (2009)

123
120 J Mater Sci (2009) 44:114–121

nitrogen has played an important role in the development been covered by allotrimorphic austenite perturbation.
of DSS. Many studies on high-nitrogen grades have been Widmanstätten austenite has a much higher aspect ratio
reported, especially in terms of improvement of mechani- (length/thickness). The large amount of austenite phase
cal and corrosion properties [18–21]. The weldability of present is due to the high Mn content and the large amount
DSS has been improved by the addition of nitrogen to of austenite-stabilizing elements present in the weld metal
stabilize austenite at elevated temperature [22]. In general, (Fig. 6c, d). Due to the energy-rich arc, a greater amount of
the austenite phase in DSS weld metal is formed from Widmanstätten austenite structure was presented in the
ferrite in three modes: allotrimorphs at the prior-ferrite helium-shielded WZ compared with in the Ar-shielded WZ.
grain boundaries, Widmanstätten side-plates growing into
the grains from these allotrimorphs, and intragranular
Conclusions
precipitates. The possibility of chromium nitride formation
during welding of DSS has been reported [23]. In the
DSS weld metals with different shielding gases were
described study pure argon and helium were used as the
welded and examined, the results are as follows:
shielding gas. From the previous literature said pure argon
as shielding gas, which is occurred in the nitrogen loss. (1) When using the helium shielding gas the weld bead
Hence, it can be avoid for the nitrogen loss we can properly aspect ratio (width/penetration) is higher than in the
maintained by the heat input and interpass temperature, can Ar-shielded weld.
control the nitrogen loss in the weld metal. By appropriate (2) Helium-shielded welds exhibited higher toughness
selection of the heat input and interpass temperature, one due to the large Mn content and the smaller amount of
can achieve a less significant difference in the amount of a ferrite phase and larger amount of austenite phase
and c phases (55% ferrite and 45% austenite) present in the present in the weld metal.
Ar-shielded weld metal compared with the helium-shielded (3) Due to the high value of the Creq/Nieq ratio, a large
weld. Chromium nitride precipitations can also be observed amount of ferrite is present in the argon-shielded weld
in Fig. 7. metal.
More ferrite phase was present in argon-shielded (4) Chromium nitride precipitation is observed in the
weld metal. So, the solubility of nitrogen in ferrite decreases argon-shielded weld metal.
rapidly with decreasing temperature. A significant amount (5) For the argon-shielded weld metal an even phase
of the ferrite phase is observed, and also it can be clearly balance (55% ferrite and 45% austenite) can be
seen that the grains in the argon-shielded weld metal are achieved by appropriate selection of the heat input
finer (Fig. 6a). The austenitic phases seen within the grains and interpass temperature.
could be either intragranular precipitates or Widmanstätten (6) The hardness of the weld metal is much higher than
austenite transverse to the long-axis direction, as shown that of the BM and HAZ for both studied shielding
in Fig. 6d. Widmanstätten austenite also nucleates gases.
directly from prior a-ferrite boundaries which have not (7) Due to the high arc energy the Widmanstätten
austenite structure was present in the microstructure
of the helium-shielded weld.

Acknowledgement The authors would like to express their heartfelt

thanks to Mr. P. Veerappan, Proprietor, Keerthana Engineering
Works, Tiruchirappalli-15, Tamil Nadu, India, for conducting the
welding trials in his factory.

References

1. Folkhard E (1988) Welding metallurgy of stainless steel.

Springer-Verlag, Vienna, p 186
2. Karlsson L, Ryen L, Pak S (1995) Weld J 74(11):28s
3. Kotecki DJ (2000) Weld J 79(12):346s
4. Hilkes J, Bekkers K (1995) Weld J 74(11):51
5. Badij R, Belkessa B, Maza H, Bouabdallah M, Bacroix B,
Kahloun C (2004) Mater Sci Forum 217:467
6. Wang S-H, Chiu P-K, Yang J-R, Fang J (2006) Mater Sci Eng A
420:26
Fig. 7 Argon-shielded weld with chromium nitride precipitations 7. Muthupandi V, Bala Srinivasan P, Seshadri SK, Sundaresan S
(2009) (2003) Mater Sci Eng A 358:9

123
J Mater Sci (2009) 44:114–121 121

8. Bonnefois B, Charles J, Dupoiron F, Soulignac A (1991) Pro- 16. Honeycombe J, Gooch TG (1978) Proceedings of trends in steels
ceedings of the conference on duplex stainless steels ‘91, vol 1, and consumables, The Welding Institute, London, p 28
Beaune, France, p 347 17. Wahlberg G, Dunlop GL (1987) Proceedings of the conference on
9. Bonollo F, Tiziani A, Zambon A, Penasa M (1994) Proceedings stainless steels, The Institute of Metals, University of New York,
of fourth international conference on duplex stainless steels, p 291
paper 108, vol 2, Glasgow, Scotland, p 13 18. Hertzman S, Jargelius Pettersson R, Blom R, Kivineve E,
10. Brown EL, Burnett ME, Purtscher PT, Krauss G (1983) Metall Eriksson J (1996) ISIJ Int 7(36):968
Mater Trans A 14:791 19. Siegmund T, Werner E, Fisher FD (1993) Comput Mater Sci
11. Zambon A, Bonollo F (1994) Mater Sci Eng A 178:203 1:234
12. Kou S (2003) Welding metallurgy, 2nd edn. Wiley, Hoboken, 20. Kokawa H, Tomita M, Kuwana T (1995) Weld Int 9(3):8
NJ, p 225 21. Ogawa T, Koseki T (1989) Weld J 68:181s
13. Ramirez AJ, Lippold JC, Brandi SD (2003) Metall Mater Trans A 22. Charles J (1991) Proceedings of the conference on duplex
34:1575 stainless steels ‘91, Les editions de physique, Beaune, p 3
14. Gooch TG (1982) ASM conference on duplex ferritic-austenitic 23. Ogawa K, Komizo Y, Azuma S, Kudo T (1992) Trans Jpn Weld
stainless steels, St. Louis, p 573 Soc 23(1):40
15. Ku JS, Ho NJ, Tjong SC (1997) J Mater Process Technol 63:770

123

View publication stats