Engineering, 2010, 2, 391-396
doi:10.4236/eng.2010.25051 Published Online May 2010 (http://www.SciRP.org/journal/eng)
391
Influence of Clad Metal Chemistry on Stress Corrosion
Cracking Behaviour of Stainless Steels Claddings in
Chloride Solution
Edmilson O. Correa1, Reginaldo P. Barbosa2, Augusto J. A. Buschinelli3, Eduardo M. Silva1
1Universidade Federal de Itajuba, Instituto de Engenharia Mecânica, Itajubá, Brazil
2Arcelor Mittal Inox Brazil, Praça 1º de maio, S/N, Centro, Timóteo, Brazil
3Universidade Federal de Santa Catarina, Departamento de Engenharia Mecânica, Florianópolis, Brazil
Received December 1, 2009; revised February 7, 2010; accepted February 12, 2010
Abstract
The effect of clad metal composition on stress corrosion cracking (SCC) behavior of three types of SMAW
filler metals (E308L-16, E309-16 and E316L-16), used for cladding components subjected to highly corro-
sive conditions, was investigated in boiling 43% MgCl2 solution. In order to evaluate the stress corrosion
cracking susceptibility of the top layer, constant load tests and metallographic examinations in tested SCC
specimens were conducted. The susceptibility to stress corrosion cracking was evaluated in terms of the
time-to-fracture. Results showed that the E309-16 clad metal presented the best SCC resistance. This may be
attributed to the presence of a discontinuous delta-ferrite network in the austenitic matrix, which acted as a
barrier to cracks propagation. Concerning to E308-16 and E316L-16 clad metals, results showed that these
presented a similar SCC test performance. Their higher SCC susceptibility may be attributed to the presence
of continuous vermicular delta-ferrite in their microstructure.
Keywords: Stainless Steels, Cladding, Stress Corrosion Cracking
1. Introduction
It is well established that sensitization-induced intergranu-
lar stress corrosion cracking of similar austenitic stainless
steels weldments occurs predominantly in the heat af-
fected zone (HAZ). Because that, traditionally, these
steels are joined each other with weld metals containing
5% to ~10% residual delta ferrite in the interdendritic
boundaries with the unique objective of reducing the
occurrence of hot cracking and microfissuring in the
weld metal. In general, little importance is given to the
evaluation of the SCC resistance of the weld metal once
the HAZ is the region less resistant to SCC and it needs
much more attention [1-4].
However, in applications where the austenitic stainless
steels are used for cladding components subjected to
highly corrosive conditions, the evaluation of the suscep-
tibility to SCC of the weld metal turns more relevant [5].
It is worthwhile mentioning that, in cladding applications,
reduction of corrosion is achieved by applying a corro-
sion resistant surface onto a cheaper and a tougher core
material by welding, generally mild carbon steels.
According to the literature [4,5], during the early
stages of solidification of the major of the austenitic
stainless filler metals, the liquid phase initially trans-
forms to delta ferrite in a dendritic manner. Simultane-
ously, elements with low solubility in the ferrite phase (C,
Ni, N, Mn) are rejected to the remaining liquid. As cool-
ing continues, the formed delta ferrite undergoes a solid-
state transformation to austenite. This solid-state trans-
formation occurs by diffusion-controlled migration of the
delta ferrite interface and involves the transport of ferrite
stabilizers to the ferrite and redistribution of the austenite
stabilizers. Previous work has shown that the ferritic
phase is more active electrochemically than the austenite
phase, resulting in the preferential corrosion of the ferrite
on exposure to aggressive environments [6].
The weld metal composition together with the welding
thermal cycle experimented by the austenitic stainless
steel during the cladding operation have influence on the
content of delta ferrite and on the elemental partitioning
in the weld metal. Consequently, the SCC resistance of
weld metal can differ to that observed in this same mate-
rial heat treated by annealing.

E. O. CORREA ET AL. 392
Therefore, the aim of the paper is to evaluate the sus-
ceptibility to stress corrosion cracking of three different
types of austenitic stainless steel deposits in an enviro-
nment containing a hot chloride solution. The correct
selection of the filler metal for cladding can be an effi-
cient procedure to eliminate or minimize the occurrence
of SCC in clad petrochemical industry equipments.
2. Materials and Experimental Procedure
2.1. Welding
Cladding was done on a mild steel plate of 10 mm thick
under the conditions shown in the Table 1.
The weld deposit consisted of 2 layers overlapped and
was processed by depositing several beads by SMA
welding. It was possible with this procedure to eliminate
the effect of dilution upon the microstructure of the clad
metal. Three different types of filler metals (AWS
E308L-16, E309-16 and E316L-16) and heat input of
9.0 kJ/cm were used in this study. The chemical compo-
sition of the undiluted top layer of the clad metal, ob-
tained by optical emission spectrometry, after the weld-
ing process, is shown in Table 2.
Considering that all of the filler metals used were for-
mulated such that solidification occurs in the FA mode,
that is, with delta ferrite as primary phase, it can be noted
from Table 2 that the most dominant difference in com-
position of them is the presence of higher amounts of
very strong austenite stabilizers elements (nickel and
carbon) in the E309-16 clad metal and higher amount of
Mo (ferrite stabilizer) in the E316L-16 clad metal.
Table 1. Welding parameters.
Heat input
(KJ/cm)
Voltage
(V)
Current
(A)
Travel speed
(cm/min)
layers
number
9.0 30 90 17 2
Table 2. Chemical composition (wt%) of the top layer of
stainless steels in as-welded condition.
C Mn Si Cr Ni Mo N
E308
L-16 0.0340.840.5819.63 9.44 0.0340.0387
E309-
16 0.1140.800.7923.79 12.43 0.0520.0786
E316
L-16 0.0220.690.9619.51 11.79 2.700.0580
2.2. Microstructural Characterization and
Hardness test
After welding, 30 mm × 10 mm rectangular specimens
were sectioned transverse to the surface, then polished
and etched electrolytically in 10% oxalic acid to reveal
the microstructure of the clad metals. This etching re-
veals the ferrite as a dark phase in the bright austenite
matrix. Metallographic specimens were also etched using
sodium hyposulfite to obtain a better contrast between
the phases. Microstructural examinations of the speci-
mens were carried out using standard optical microscopy
and SEM. The bulk hardness of the clad layer was meas-
ured using a Vickers hardness tester less than 10 Kg load.
2.3. SCC testing
A constant-load lever arm apparatus was used to evaluate
the susceptibility to SCC of the three types of stainless
steel clad metals in aerated boiling 43% MgCl2 test solu-
tion at 145 ± 3ºC. 3 mm thick smooth rectangular tensile
specimens, as can be seen in Figure 1, obtained from the
top layer of clad metals, were used for the SCC tests. The
preparation of the specimens was in accordance with
ASTM G58 and ASTM E8 standards.
Specimens were put in a glass cylinder in which the
test solution was placed. The glass vessel was coupled to
a reflow condenser so that all measurements were moni-
Figure 1. Design of the smooth tensile specimen used in SCC tests (unit: mm).

E. O. CORREA ET AL.393
tored continuously under closed-circuit conditions. Sea-
ling between glass vessel and specimen was done by
silicon rubbers.
The tests were conducted at a tensile load correspo-
nding to the nominal stress level of 240 MPa. Time-to-
failure was the SCC parameter evaluated
3. Results and Discussion
3.1 Cladding Hardness
According to NACE MR0175 standard [7], a maximum
hardness of 250 HV is required to austenitic stainless
steels to be used in petroleum process components in
order to avoid stress corrosion in environments contain-
ing hydrogen sulfide. The NACE standard does not men-
tion the maximum hardness required to the as-welded
austenitic stainless steels to avoid stress corrosion crack-
ing when exposed to environments containing chlorides,
however, it is believed that the maximum hardness value
of 250 HV may also be adopted in this condition.
Figures 2, 3 and 4 show the Vickers hardness profile
of the cladding metal using the filler metals studied. The
maximum hardness value found was 190 HV10, which
classify all claddings as acceptable according to the cri-
terion adopted by NACE standard.
3.2. Microstructure
The microstructure of the top layer of the all filler metals
studied presented variable amount of delta-ferrite. In
general, the ferrite morphology was predominantly fine
and vermicular (see Figure 5). The delta-ferrite present
was typically intradentritic and it is a result of the incom-
plete transformation of primary ferrite into austenite. Ac-
cording to the literature, this vermicular morphology is
present when weld cooling rate is not so high or moder-
ate (characteristic of the SMA welding process) and/or
Figure 2. Vickers hardness vs. depth from surface of the
E308L-16 clad metal.
Figure 3. Vickers hardness vs. depth from surface of 316L-
16 clad metal.
Figure 4. Vickers hardness vs. depth from surface of the
E309-16 clad metal.
when the Creq/Nieq is low but still within the FA
mode (Creq/Nieq calculated was between 1.48 and 1.89) [4].
3.3. SCC Results
As mentioned in the experimental procedure, the neces-
sary time for occurrence of the complete fracture was the
parameter adopted to evaluate the susceptibility to the
stress corrosion cracking of the clad specimens. Table 3
shows the results obtained in the SCC tests of the speci-
mens.
The SCC results showed that the welds carried out us-
ing filler metal AWS E309-16 were significantly more
resistant to the cracking than those carried out using the
filler metals AWS E308L-16 and E316L-16. The SCC
behavior of these last ones was quite similar. From the
Figure 6, it can be observed that the minimum time-to-
fracture of the specimens using the electrode E309-16 is
approximately 1.5 and 2 times superior, respectively, to
the maximum time-to-fracture verified for the electrodes
E308L-16 and E316L-16.
The microstructure of the E309-16 clad metals (ferrite
content of 10FN) consisted of a discontinuous delta-ferrite
network distributed in the austenitic matrix as well as a

E. O. CORREA ET AL. 394
Figure 5. Microstructure of the top layer of the cladding
with filler metal E308L-16. Note the presence of two
delta-ferrite (dark phases) morphologies-Vermicular mor-
phology (right) and network morphology (left). Etching:
sodium Hyposulfite.
Table 3. SCC Results for the specimens (heat input of 9.0
kJ/cm).
Filler Metal E308L-16 E309-16 E316L-16
Testing time
(min) 248 298 194 639 1026 325243
Temperature
(ºC) 147 145 147 148 148 148144
tendency of globularization of certain amount of ferrite
(see Figure 7). This result is in accordance with Krish-
nan et al. [3] who verified that this tendency for globu-
larization of ferrite increases when the network is dis-
continuous and the ferrite morphology is finer. In fact,
the ferrite thickness of the E309-16 clad metal was
slightly lower than that observed in the E308L and E316
clad metals. A possible explanation for this fact is the
lower ferrite content verified in the E309-16 clad metal.
On the other hand, the top layer of the E308L-16 clad
(ferrite content of 12FN) and E316L-16 (ferrite content
of 18FN) presented a microstructure constituted of con-
tinuous delta-ferrite network in the austenite matrix (see
Figure 8).
Therefore, these microstructures indicate that the main
reason for the superior resistance to SCC of the E309-16
clad metal is the presence of discontinuous network of
delta-ferrite in the grain boundaries of austenite dendrites.
This discontinuous network of delta-ferrite in the micro-
structure of E309-16 weld metal: 1) acted as a barrier
that restricted the growth of austenitic grains during the
welded joint cooling. It is worthwhile mentioning that
large austenitic grains increase the susceptibility to SCC
of austenitic stainless steels [8], 2) reduced the stress
corrosion crack propagation rate once a continuous crack-
ing path through the delta-ferrite could not develop [6].
The presence of the discontinuous network of delta-
Figure 6. Stress corrosion cracking tests results (minimum
and maximum times-to-fracture) for the clad metals studied.
Figure 7. Microstructure of the top layer of the E309-16
stainless steel showing discontinuous delta-ferrite networks
(dark phase) distributed in the austenitic matrix (gray
phase). Etching: oxalic acid.
Figure 8. Microstructure observed in the weld metal of
E308L-16 and E316L-16 stainless steels showing a continu-
ous delta-ferrite network (dark phase) distributed in the
austenitic matrix (gray phase). Etching: Oxalic acid.
ferrite in the 309-16 clad metal may be mainly attributed
to the two factors: 1) the presence of higher amounts of
Ni and C in its composition and 2) the relatively fast
weld metal cooling rate. As mentioned before, during the
FA solidification of the E309-16 clad metal, Ni and C,
which have low solubility in the ferrite phase, are rejected
to the remaining liquid during the transformation of liquid
to delta-ferrite. As cooling continues, the formed delta-

E. O. CORREA ET AL.395
ferrite becomes unstable and undergoes a solid-state
transformation to austenite. Thus, austenite consumes the
ferrite via a diffusion-controlled reaction across the aus-
tenite-ferrite interface until the ferrite is sufficiently en-
riched in ferrite-promoting elements (Cr and Mo) and
depleted in austenite-promoting elements (Ni and C) [4].
Having in mind these observations and considering that
the welding cooling rate was relatively fast; at the end of
solidification where diffusion is limited, probably, very
small regions along to the austenite-ferrite boundaries
remained enriched with Ni and C. As a consequence, in
these small regions along of the austenite-ferrite bounda-
ries there was no formation of delta ferrite, making it
discontinuous along to the interface.
On the other hand, considering also the fast weld
metal cooling rate, the continuous network of delta fer-
rite observed in the E316L-16 clad metal may be attrib-
uted to the higher content of Mo (ferrite stabilizer) as
well as the lower content of carbon and in its composi-
tion in comparison with the E309-16 filler metal. As a
result, at the end of the FA solidification, the austen-
ite-ferrite boundaries remains enriched of Mo, which acts
to facilitate the formation of higher amount of delta fer-
rite in them. Also, considering that the E316L-16 filler
metal presents a nickel content slightly lower than that of
E309-16 and content of C and Cr similar to the E308L
filler metal, it is possible to affirm that Mo influenced
significantly to the formation of the continuous network
of delta-ferrite in the microstructure of this clad metal. In
fact, the 316L clad metal presented the highest FN num-
ber (18FN) among the filler metals studied.
With relation to the E308-16 clad metal, the presence
of the continuous network of delta-ferrite in its micro-
structure may be attributed to the lesser contents of Ni
and C in its composition in comparison with the E309-16
filler metal. This facilitates the formation and stabiliza-
tion of delta-ferrite in the austenite-ferrite boundaries at
the end of the FA solidification.
It is also worthwhile mentioning that, based on the
observations above, it may be suggested that the influ-
ence of the delta-ferrite on the susceptibility to SCC of
austenitic welds is more related with its morphology and
distribution in the austenitic matrix than with its content.
In fact, the higher resistance to SCC of the weld metal
E309-16 can not be attributed arbitrarily to the lower
content of delta-ferrite in its microstructure, considering
that the weld metal E308L-16 presented an amount of
delta-ferrite very close to that found in weld metal E309-
16. However, its performance in the stress corrosion tests
was quite inferior to the performance of this last one.
Figures 9 and 10 show stress corrosion cracks devel-
oped in the welds using the filler metals E308L-16 and
E309-16, respectively. As it can be seen from Figure 9
the crack propagation occurred under a relatively straight
and smooth cracking path through the ferrite boundaries.
This continuous cracking path resulted in more rapid
SCC failure and it was developed due to the relatively
continuous vermicular network of ferrite present in the
microstructure. On the other hand, it can be seen from
Figure 10 that, due to the discontinuous morphology of
the ferrite, the crack propagation in the weld metal
E309-16 occurred under a nonplanar cracking path. Thus,
once the crack was initiated, it became more difficult for
it to propagate along the tortuous boundaries of ferrite [9].
It can also be noted from the both figures that the
stress corrosion cracks were developed along to the
boundaries of ferrite. This is attributed to the fact that the
ferrite phase (anode) is attacked more easily by the envi-
ronment than the austenitic phase (cathode), particularly
due to the segregation of impurities in it [6].
Further research should be carried out to verify the in-
fluence of different temperatures, different loads and
environments on the susceptibility to stress corrosion
cracking of these stainless steels.
Figure 9. SCC crack developed along to the dendrites in the
E308L-16 weld metal. Note the relatively straight and
smooth cracking path through the ferrite boundaries.
Etching: oxalic acid.
Figure 10. SCC crack developed along to the dendrites in
the E309-16 clad metal. Note a relatively tortuous cracking
path through the ferrite boundaries. Etching: oxalic acid.

E. O. CORREA ET AL. 396
3. Conclusions
1) The filler metal AWS E309-16 produced a clad layer
with higher resistance to SCC than filler metals AWS
E308L-16 and E 316L-16 and it is more recommended
for cladding equipments subject to SCC. Its microstruc-
ture formed by a discontinuous network of delta-ferrite
reduced both the austenite grain size during the weld
metal solidification and the propagation rate of stress
corrosion crack during the SCC tests.
2) The results indicated that the contribution of the
delta ferrite in the SCC resistance is much more related
with its morphology and distribution than with its con-
tent in the austenitic welds.
4. Acknowledgements
The authors are grateful to PETROBRAS for supplying
the welding consumables and the Brazilians government
organizations CNPq and FAPEMIG for the financial
support.
5
. References
[1] O. M. Alyousif and R. Nishimura, “The Stress Corrosion
Cracking Behavior of Austenitic Stainless Steels in Boil-
ing Magnesium Chloride Solutions,” Corrosion Science,
Vol. 49, No. 7, 2007, pp. 3040-3051.
[2] ASM Handbook, “Corrosion,” Metals Park, USA, Vol.
13, 1999.
[3] K. N. Krishnan and K. P. Rao, “Effect of Microstructure
on Stress Corrosion Cracking Behaviour of Austenitic
Stainless Steel Weld Metals,” Materials Science and En-
gineering A, Vol. 142, No. 1, 1991, pp. 79-85
[4] J. C. Lippold and D. J. Kotecki, “Welding Metallurgy and
Weldability of Stainless Steel,” 5th Edition, John Willey
& Sons, Hoboken, 2005.
[5] C. D. Lundin, “Dissimilar Metals Welds: Transition
Joints Literature Review,” Welding Journal, Vol. 61, No.
2, 1982, pp. 58-63.
[6] G. Sui, E. A. Charles and J. Congleton, “The Effect of
Delta-Ferrite Content on the Stress Corrosion Cracking of
Austenitic Stainless Steels in a Sulphate Solution,” Cor-
rosion Science, Vol. 38, No. 5, 1996, pp. 687-703
[7] H. L. Logan, “Stress Corrosion,” 11th Edition, In.: NACE
Basic Corrosion Course. Anton deS. Brasunas, Houston,
1990.
[8] W. A. Baeslack, W. F. Savage and D. J. Duquette, “Ef-
fect of Nitrogen on the Microstructure and Stress Corro-
sion Cracking of Stainless Weld Metals,” Welding Jour-
nal, Vol. 58, No. 3, 1979, pp. 83-90.
[9] V. Y. Gertsman and S. M. Bruemmer, “Study of Grain
Boundary Character along Intergranular Stress Corrosion
Crack Paths in Austenitic Alloys,” Acta Mater, Vol. 49,
No. 9, 2001, pp. 1589-1598.