Case Studies in Engineering Failure Analysis 2 (2014) 6975

Contents lists available at ScienceDirect

Case Studies in Engineering Failure Analysis

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

Case study

Weld failure analysis of 2205 duplex stainless steel nozzle

Jingqiang Yang, Qiongqi Wang, Zhongkun Wei, Kaishu Guan *
Key Lab of Safety Science of Pressurized System, Ministry of Education, School of Mechanical Engineering, East China University of Science
and Technology, Shanghai 200237, China

A R T I C L E I N F O A B S T R A C T

Article history: Failure analyses of weld joint between the nozzle and the head of the reactor made of 2205
Received 20 May 2013 duplex stainless steel was performed by optical microscopy (OM) and scanning electron
Received in revised form 25 April 2014 microscopy (SEM). Cracks were found in HAZ of the weld. The depth of the cracks is equal
Accepted 7 May 2014 to the thickness of the inner weld. Localized uneven distribution of ferrite/austenite with
Available online 13 May 2014 8090% ferrite in weld is found. Results show that the cracks occurred along columnar
granular with cleavage fracture. Poor weld process probably results in these cracks.
Keywords: 2014 Kaishu Guan. Published by Elsevier Ltd. This is an open access article under the CC
Duplex stainless steel
BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/3.0/).
Weld
Cracking

1. Introduction

The duplex stainless steels (DSSs) with both ferrite and austenite phases have better resistance to localized corrosion than
single-phase austenitic stainless steels (SSs) in chloride-containing solutions. Therefore, the DSSs have gained increasing
interest in recent years for a number of applications as structural materials in various industrial sectors including chemical,
petrochemical, power generation, and oil renery [13].
The good mechanical properties and corrosion resistance of DSSs depend upon the proper austeniteferrite balance,
which is approximately 1:1, and high Cr, Mo and N contents [4]. However, the DSSs weld joints have a drawback that the
mechanical properties deteriorate signicantly due to the very uneven distribution of ferrite and austenite in the weld joints.
Since welding is widely used in the fabrication of pressure vessels, understanding the causes that result in the cracking is
critical to the application of the duplex stainless steel vessels.
In this work, the failure mechanism of a reactor vessel after 1 year of operation was studied.
The working temperature and working pressure of the reactor are around 100 8C and 0.2 MPa, respectively. The medium
is processed gas. The chemical composition of the material of the reactor is listed in Table 1.

2. Visual examination and weld process

Fig. 1 shows the schematic drawing of the head and nozzles of the reactor. The diameter of the reactor is 4000 mm with
34 mm thickness in the head. Five outer nozzles are placed onto the head to enter and hold the coils. The central nozzle holds
the agitator. The connections between the nozzles and the head are welded by SMAW (shielded metal arc welding) with a

* Corresponding author. Tel.: +86 21 64253055.

E-mail address: guankaishu@ecust.edu.cn (K. Guan).

http://dx.doi.org/10.1016/j.csefa.2014.05.001
2213-2902/ 2014 Kaishu Guan. Published by Elsevier Ltd. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/
licenses/by-nc-nd/3.0/).
70 J. Yang et al. / Case Studies in Engineering Failure Analysis 2 (2014) 6975

Table 1
Chemical composition of the material of the reactor (wt%).

C Mn Si S P Cr Ni Mo N

0.020 0.87 0.38 0.001 0.021 22.6 5.94 3.16 0.16

Fig. 1. Schematic structure and crack between nozzle and the heat for the reactor with (a) outside and (b) inside.

K-grove preparation. The welds are completed in four to ve passes. The cracks can be found in all of the ve outer nozzles
and all are located at the outer ring of the nozzles where it is poor accessible to the welder.
Fig. 2 shows the schematic crack position. The crack originated from the HAZ and the depth of the cracks is exactly equal
to the inner weld thickness of 17 mm of the K-grove depth. Fig. 3 shows an image of the cracks. The cracks are in the nozzle
side. No wall thinning or plastic deformation is observed near the cracks, which reveals that the failure is brittle fracture in
nature.
A small piece of boat sample of crack was cut to examine the fractography and metallography. The microstructure was
characterized by optical microscopy (OM), and the fractography was observed by scanning electron microscopy (SEM).
Ferrite measurements were carried out using image analyses.

Fig. 2. Detail weld structure between nozzle and head.

 J. Yang et al. / Case Studies in Engineering Failure Analysis 2 (2014) 6975 71

Fig. 3. An actual photograph of the crack parallel to the weld.

3. Fractography

A boat samples was taken from the nozzle. Fig. 4 shows the fractography of the sample. The crack shows a rather coarse
cleavage like fracture with general view as shown in Fig. 4(a). Fig. 4(b) shows the cleavage like fracture surface along with
dendrite weld.

4. Metallography

Fig. 5 shows the metallography of the cross-section of the weld. It can be seen that the microstructure in the upper part of
the weld seems to predominantly ferritic (dark phase) with only some thin and strong orientated (columnar) austenite
grains (white phase).
Fig. 6 shows the cross-section metallography of the sample. It can be seen that the weld is not homogenous. It can be
observed that the top layer of the weld is ne grained and contains numerous small austenite grains. Below this zone the
microstructure is mainly ferritic with the orientated austenite grains. In the columnar zone, ferrite contents varying between
80% and 90% can be observed (magnication about 200). Fig. 7 shows a small crack beside the large crack surface, which is
intercolumnar cracking of the weld metal at the beginning, however it becomes transgranular form at the tip of cracking.
Some transgranular form of cracking was also found in Fig. 8.

5. Discussion

It can be summarized that the cracks are located at the outer circle of the nozzles and have a depth equal to the thickness
of the inner weld. The cracks can be described as the result of intercolumnar cracking of the weld metal and some little
transgranular form of cracking. The weld presents very uneven distribution of ferrite/austenite with 8090% ferrite. From
above analysis, the cracks are probably caused by poor weld process, revealed by the poor accessibility of the welder. The
poor weld process caused it could not input enough welding heat and then the welding line energy was low. Furthermore,
the fast cooling rate after welding caused inevitably high ferrite content.
The intercolumnar cracks suggest that they are formed during the solidication process of the weld. These cracks are
primarily formed in heavy sections and with much restraint, which low austenite content.

5.1. Ferrite/austenite ratio

The duplex stainless steel base metal and weld are subjected to a series of thermal cycles. As a result, complex
microstructural transformations occur, affecting the d/g phase balance in the steel. The difference in cooling rate between
the central region of the weld and the regions near the fusion boundaries affects the d/g equilibrium in the weld metal.
During heating, near the high temperature peak, the structure transformed to ferrite almost entirely. Then the ferrite
partially transformed to austenite.
72 J. Yang et al. / Case Studies in Engineering Failure Analysis 2 (2014) 6975

Fig. 4. SEM morphology of the sample.

 J. Yang et al. / Case Studies in Engineering Failure Analysis 2 (2014) 6975 73

Fig. 5. Metallography of the cross-section of the sample.

These welds were dressed after welding, which means that after welding the surface of the weld was smoothed by using a
TIG torch without ller metal. So the very high quenching rate results in high ferrite content and caused local uneven
distribution of ferrite/austenite. Since the local duplex structure is severely ferritized by the high peak temperature and by
the fast cooling rate. In the outer ring of the nozzles, the temperature may be uneven and the cooling rate was also different
which cause homogenous in the weld shown in Fig. 6.
It is well known that the impact toughness of the DSSs welds decreases with the increasing of ferrite in the HAZ and
weldment [57], since the local duplex structure is severely ferritized by the high peak temperature and by the fast cooling
rate of the thermal cycle. Another problem associated with fusion welding of these materials is their susceptibility to
solidication cracking, which is greater than that of the 304 L austenitic stainless steels [8,9]. When duplex stainless steels
solidify to ferrite, the austenite formation occurs solely in the solid state. This may reduce the benecial effects of austenite
formation on solidication cracking hindrance. And it has been shown that compositions solidifying in a ferritic mode can be
susceptible to solidication cracking. The alloys with higher ferrite content have a coarser columnar structure and are more
likely to exhibit solidication cracks.

Fig. 6. Cross-section of sample electrolytically etched in 10 N KOH.

74 J. Yang et al. / Case Studies in Engineering Failure Analysis 2 (2014) 6975

Fig. 7. Intercolumnar crack in cross-section of sample at the beginning.

5.2. Stress concentration in the weld

The welding sequence is: weld the outer weld rstly, and then weld the inner weld. It indicates that the highest residual
weld stresses are inside, because the outer weld is heat treated during welding of the inner weld.

5.3. The process of cracking

From above analysis, the intercolumnar cracks may be formed in solidication process which can be veried in Figs. 6 and
7. The cracks further developed in the transgranular form during hydrostatic pressure test due to the embrittlement of the

Fig. 8. Transgranular crack in the cross-section of the sample.

 J. Yang et al. / Case Studies in Engineering Failure Analysis 2 (2014) 6975 75

weld joint which can be veried in Figs. 7 and 8. However, the cracks were failed to detected after the hydrostatic pressure
test since the cracks did not penetrate the nozzle, so the vessel was put into operation.

5.4. Repairing

It is necessary to ensure the continuity of duplex structure properties across the weld by controlling the phase balance
both in the FZ and in the HAZ during repairing these welds. For practical application of this kind of welded joints, an adequate
proportion of ferrite in the welded joints would be in the range of 3070%. This ferrite content depends on the chemical
composition of the welded joints and the cooling rates of the weld, which are determined by the input energy applied during
welding [10]. The welds are going to be repaired by GTAW (gas tungsten arc welding) method, which is better method
compared with SMAW (shielded metal arc welding) method used before, since it allows cleaner welds, better weld
penetration and better controlled application of heat. ER2209 with 2.0 mm diameter will be used as ller material.
Consequently, better performance of welds could be achieved by controlling weld process.

6. Conclusion

Cracks were found between nozzles and head of reactors. The cracks were located at the outer ring of the nozzles and have
a depth equal to the thickness of the inner weld. The cracks are resulted from intercolumnar cracking of the weld metal
during solidication process of the weld. Then the cracks further develop with transgranular form during hydrostatic
pressure test. The weld presents very uneven distribution of ferrite/austenite with 8090% ferrite. The cracks are probably
caused by poor weld process, revealed by the poor accessibility of the welder. So in the future work, better performance of
welds could be achieved by controlling weld process.

References

[1] Liou H-Y, Hsieh R-I, Tsa W-T. Microstructure and stress corrosion cracking in simulated heat-affected zones of duplex stainless steels. Corr Sci
2002;44:284156.
[2] Nilsson JO. Super duplex stainless steels. Mater Sci Technol 1992;8:685700.
[3] Ravindranath K, Mailhotra SN. The inuence of aging on the intergranular corrosion of 22 chromium5 nickel duplex stainless steel. Corr Sci
1995;37:12132.
[4] Gunn RN. Duplex stainless steels: microstructure, properties and applications. Cambridge, England: Abington Publishing; 1997.
[5] Gooch TG. Weldability of duplex ferriticaustenitic stainless steel. In: Conf proc duplex stainless steels82. St. Louis: ASM; 1983. p. 573.
[6] Sun Z, Kuo M, Annergren IY, Pan D. Effect of dual torch technique on duplex stainless steel welds. Mater Sci Eng A 2003;356:27482.
[7] Muthupandi V, Bala Srinivasan P, Seshadri SK, Sundaresan S. Effect of weld metal chemistry and heat input on the structure and properties of duplex
stainless steel weld. Mater Sci Eng A 2003;358:916.
[8] Varol I, Baeslack III . Characterization of weld solidication cracking in a duplex stainless steel. Metallography 1989;23:119.
[9] Nelson DE, Baeslack WA, Lippold JC. Investigation of weld hot cracking in duplex stainless steels. Weld J 1987;66:24150.
[10] Urena A, Otero E, Utrilla MV, Munez CJ. Weldability of a 2205 duplex stainless steel using plasma arc welding. J Mater Process Technol 2007;182:624
31.