Corrosion Science 51 (2009) 2316–2323

Contents lists available at ScienceDirect

Corrosion Science
journal homepage: www.elsevier.com/locate/corsci

Intergranular corrosion susceptibility in supermartensitic stainless steel weldments

J.M. Aquino *, C.A. Della Rovere, S.E. Kuri
São Carlos Federal University (UFSCar), Materials Engineering Department, Rodovia Washington Luís, km 235, CEP 13565-905, São Carlos, SP, Brazil

a r t i c l e i n f o a b s t r a c t

Article history: The intergranular corrosion susceptibility in supermartensitic stainless steel (SMSS) weldments was
Received 31 March 2009 investigated by the double loop – electrochemical potentiokinetic reactivation (DL-EPR) technique
Accepted 8 June 2009 through the degree of sensitization (DOS). The results showed that the DOS decreased from the base
Available online 14 June 2009
metal (BM) to the weld metal (WM). The heat affected zone (HAZ) presented lower levels of DOS, despite
of its complex precipitation mechanism along the HAZ length. Chromium carbide precipitate redissolu-
Keywords: tion is likely to occur due to the attained temperature at certain regions of the HAZ during the electron
C. Intergranular corrosion
beam welding (EBW). Scanning electron microscopy (SEM) images showed preferential oxidation sites in
C. Pitting corrosion
Stainless steel
the BM microstructure.
Welding Ó 2009 Elsevier Ltd. All rights reserved.

1. Introduction In welded joints, chromium and molybdenum carbonitride pre-

cipitations in the heat affected zone (HAZ), particularly in the
Intergranular corrosion is a selective process that occurs in sen- supermartensitic stainless steel (SMSS), are responsible for its sus-
sitized regions of stainless steels as a result of inadequate heat ceptibility to a corrosion attack. Those precipitations mainly occur
treatments, welding, or high-temperature service [1]. In the classi- at the prior austenite grain boundaries and also at the martensite/
cal sensitization process there is the formation of chromium rich d-ferrite phase boundaries during a multipass welding, with an
precipitates at grain boundaries, with the subsequent impoverish- additional intensification promoted by the post weld heat treat-
ment of that element in the adjacent matrix. Consequently, passive ments (PWHT) [10]. Carbide precipitates occur due to the combi-
films formed over those depleted regions are not stable, thus, it nation of carbon matrix saturation and tempering effect of the
leads to a more susceptible region to a corrosion attack. In that as- subsequent welding passes. That process is not expected to occur
pect, intergranular corrosion could resemble a galvanic cell in in the cap layer or in the single pass welded joints [10]. Intergran-
which the grains are the cathodic area, and the corresponding grain ular stress corrosion cracking (IGSCC), which was not expected in
boundaries are the anodic one; which results in a high cathodic the HAZ of the SMSS, can occur due to small amounts of Cr-carbide
area in relation to an anodic area [2]. Martensite induced sensitiza- precipitates at the prior austenite grain boundaries [11]. According
tion is another type of sensitized region susceptible to an intra- to the literature [12,13], the HAZ is the most susceptible region to a
granular corrosion attack [1,3]. This process occurs due to the crack initiation, due to its sensitization in weldments made by con-
chromium carbide precipitation in the martensitic lath boundaries ventional welding processes.
[3] which is more intensive in tempered than in quenched mar- Additionally, the welding technique exerts influence on the cor-
tensite [4]. Other metallurgical phases may contribute to influence rosion behavior of the weldment. The high power density pro-
the corrosion resistance. According to the literature [4] austenite cesses are interesting due to their high heat input; which is
promotes carbon and nitrogen dissolution, having a consequent confined in a small region of the workpiece, and also to their high
reduction in the chromium and molybdenum precipitates. Thus, cooling rates. These characteristics promote the dissolution of Cr-
pit potential demonstrated to be dependent on the austenite con- carbide precipitates as well as their suppression. That precipitate
tent; which gave noble potentials [5]. Otherwise, the d-ferrite suppression is caused by very short time periods in the tempera-
phase presented in low carbon 13%Cr steels deteriorates the corro- ture precipitation zone [14]. Corrosion resistance and mechanical
sion resistance [6] due to the higher carbide precipitation around properties, when comparing the high power density to the conven-
that phase, that is occasioned by its low carbon solubility [7]. Other tional processes [14,15], can lead to better welded joints.
corrosion and mechanical problems are associated with that phase Pitting corrosion is also likely to occur at chromium depleted
[8,9]. regions [5] as well as near the MnS inclusions [16–21]. However,
a direct relationship was not established. According to the litera-
* Corresponding author. Tel.: +55 16 33518506; fax: +55 16 33615404. ture [22], the HAZ of a high power density process weldment is
E-mail address: dsek@power.ufscar.br (J.M. Aquino). not the most susceptible region to the pitting corrosion due to

0010-938X/$ - see front matter Ó 2009 Elsevier Ltd. All rights reserved.
doi:10.1016/j.corsci.2009.06.009
 J.M. Aquino et al. / Corrosion Science 51 (2009) 2316–2323 2317

the redissolution process of the Cr-carbide precipitates. Further- Table 2

more, the HAZ microstructure does not seem to influence the pit- Electron beam welding parameters for the top pass.

ting corrosion [23]. Parameters Conditions

The double loop electrochemical potentiokinetic reactivation Vacuum 1 mbar
(DL-EPR) is a corrosion testing method characterized by a rapid, Welding voltage 60 kV
quantitative, and non-destructive test to measure the degree of Working distance 50 mm
sensitization [24]. This technique was primarily developed for Welding speed 7 mm s 1
Root opening 0.5 mm
austenitic stainless steel using standard practices. However, inter- Wire feed rate 3.78 m min 1
granular corrosion concerning the martensitic stainless steel (MSS) Welding current 130 mA
is a potential problem due to its use in the tempered condition for Heat input 1.1 kJ mm 1
hardness adjustment. As the conventional MSS contains high levels
of carbon, it is not regularly welded. On the other hand, with the
development of new weldable supermartensitic stainless steel The electrochemical testing samples were mounted in polyester
(SMSS), which has lower carbon content, there is the need to inves- resin to avoid the presence of crevices after the electric contact.
tigate the behavior of those welded joints regarding the sensitiza- The exposed area was 0.5 cm2. Before the polarization measure-
tion. In order to combine less susceptible welded joints, electron ments, the samples were wet-grinded on 600 silicon carbide
beam welding is inserted, considering intergranular and pitting (SiC) paper, washed in distilled water, and immersed in the elec-
corrosion. That process links high welding velocities to narrow trochemical cell.
heat affected zones of a few millimeters, with single passes, and The test solution was 0.5 mol L 1 H2SO4 with 0.01 mol L 1
no PWHT. These characteristics enable a more resistant weldment, KSCN, reagent-grade, in distilled water. The electrolyte was natu-
considering precipitation as well as corrosion. rally aerated and the temperature was held at 25 °C. A conven-
The aim of this work was to measure, comparatively, the degree tional electrochemical cell composed of a platinum counter
of sensitization (DOS) of the SMSS weldments, through the DL-EPR electrode and a satured calomel reference electrode (SCE), con-
technique. nected to a Solartron SI 1287A potentiostat, was used.
The DL-EPR polarization curves were obtained after two steps:
2. Experimental First, the working electrode was subjected to open circuit condi-
tions, until a steady state potential (Ecorr) was reached. This was
Electrochemical measurements were done in samples of the accomplished in 25 min. Then, an anodic potentiodynamic sweep-
base metal (BM), the heat affected zone (HAZ), and the weld metal ing rate of 1.67 mV s 1, from 100 mV/Ecorr to +600 mVSCE, was
(WM); in two classes of supermartensitic stainless steel that were imposed. At +600 mV, the potential scanning was reversed back
welded by an electron beam in a chamber of low vacuum. to 100 mV/Ecorr. At least four curves for each sample region were
The full penetration butt weld was done between two 20 mm done to obtain good reproducibility. The test results were ex-
thick plates, which were previously hot-rolled and tempered at pressed in the current densities’ ratio, iR/ia, which was used to eval-
600 °C for 10 min. The welding was done by using a matching filler uate the DOS. The iR term is the reactivation current density
wire metal, in two passes, without PWHT. The chemical composi- (maximum current density in the reverse scan), and the ia term
tion of the plates, the filler wire, and the welding operation condi- is the activation current density (maximum current density in
tions are showed in Tables 1 and 2, respectively. the anodic scan).
The 1 mm thick samples were extracted by electroerosion, from The DOS was compared to the pit potential results, which were
the inner part of the weldment top. Fig. 1(a–c) shows a schematic done in a separate work [22]. The used weldment samples were
representation of the samples’ extraction. Each extraction step was the same in both works. The pit potential determination was done
carried after etching the weldment sample with a villela’s reagent, in a 3.56 mass% (reagent-grade) sodium chloride solution.
through the identification of the characteristic microstructures
(BM, HAZ, and WM), as well as through an optical microscopic 3. Results and discussion
analysis. In order to attempt disregarding any galvanic influence,
the samples’ electrochemical investigation was carried separately Fig. 2(a–c) shows some of the initial microstructures of the BM,
in the BM and in the WM regions. A clear HAZ sample cutout the HAZ, and the WM. Due to their similar microstructural appear-
was not possible to be obtained due to its small size. Fig. 1(d) ance in the weldment regions of both steels, there were chosen
shows that the sample also presented BM and WM region residues. only three micrographs. Tempered martensite is the main metal-
X-ray diffraction measurements were done separately in the lurgical phase presented in the BM as a result of the tempering
BM, HAZ, and WM samples for identifying and quantifying the heat treatment. The HAZ microstructure also exhibited tempered
metallurgical phases. The austenite phase quantification (volumet- martensite as a consequence of the thermal heat gradient. That
ric fraction, %) was carried according to the standard practice process enabled distinct levels of tempering in the HAZ, as a conse-
ASTM-E975 [25]. The X-ray beam was focused in the middle of quence of the different temperatures reached. Additionally, the
all samples. That was particularly important when considering HAZ was characterized by a heterogeneous microstructure (rela-
the HAZ samples. The analysis was a 2h–h with a continuous tive to grain size). Quenched martensite is the WM characteristic
sweeping rate equal to 2 degrees a minute, ranging from 5 degrees phase that is formed due to the high cooling rates attained. Re-
to 120 degrees. During the X-ray measurements the samples were tained austenite was presented in the BM (A steel = 11.1% and B
rotationed. steel = 39.0%), in the HAZ (A steel = 5.5% and B steel = 4.3%), and

Table 1
Chemical composition (mass%) of steels and filler metal.

Material C Si Mn P S Cr Ni Mo Ti V Cu O N
A 0.02 0.3 0.9 0.03 0.004 12.6 5.1 1.8 0.01 0.05 0.3 0.01 0.01
B 0.007 0.07 1.6 0.007 0.003 11.4 6.1 2.6 0.02 0.05 0.5 0.01 0.01
Filler 0.012 0.45 0.65 0.005 0.005 12.3 6.4 2.6 – – – 0.008 0.01
2318 J.M. Aquino et al. / Corrosion Science 51 (2009) 2316–2323

Fig. 1. Sample extraction schematic representation of an electron beam weldment top pass: (a) welded joint, (b) weldment slice extracted by electroerosion, (c) testing
samples having dimensions of 10 mm  5 mm  1 mm, and (d) the HAZ sample was composed of the BM and the WM region. Weldment regions 1, 2, and 3 refer to the base
metal (BM), heat affected zone (HAZ), and weld metal (WM). The asterisk indicates where the electrochemical testings were done.

in the WM (A steel = 26.4% and B steel = 18.5%). According to the

literature [7], the austenitic phase in low carbon MSS can only be
visualized through transmission electron microscopy (TEM). d-Fer-
rite was detected in the BM, and also next to the fusion line in the
HAZ. Its morphological appearance resembles dark stringers as
indicated by the arrows in Fig. 2(a and b).
Fig. 3(a and b) shows representative curves of the DL-EPR test
for the BM region of the A and B steels. The activation (anodic
direction) and reactivation (cathodic direction) current densities
are similar, and very high. That indicates a high sensitized material
(iR/ia = 0.589 and 0.643 for the A and B steel, respectively). Scan-
ning electron microscopy (SEM) images of Fig. 4(a and b), reveal
intergranular and intragranular corrosion attacks, occasioned by
the chromium carbide (Cr-carbide) precipitates, and possibly other
precipitates like Fe2Mo [26]. That process occurred during the BM
tempering heat treatment.
Fig. 3 also shows the presence of two anodic peaks in the acti-
vation process of both steels. In order to separate and investigate
the origin of these peaks, a potentiokinetic sweeping rate at
0.67 mV s 1 was only conducted in the anodic direction. Fig. 5
shows that it was not possible to separate, completely, the two
anodic peaks of the B steel. According to the literature [3,27–30],
there are some interpretations for the so-called second anodic cur-
rent maximum (SACM), which are: (1) nickel enrichment on the
surface; (2) adsorbed hydrogen oxidation; (3) Fe2+ ions effect; (4)
chromium depleted zone effect, and (5) microstructural and com-
positional effects. Fig. 6(a and b) shows SEM images obtained after
the potential sweeping rate interruption at E1 = 200 mVSCE, and at
E2 = 50 mVSCE, respectively, which are indicated by the arrows in

Fig. 3. DL-EPR curves for the A and B steel BMs. The potential scanning direction is
Fig. 2. Optical micrograph of the weldment: (a) B steel BM, (b) A steel HAZ (next to indicated herein (passive vertex potential at 600 mV). The ia and iR current densities
the fusion line), and (c) A steel WM. The arrows indicate the d-ferrite phase. are also indicated.
 J.M. Aquino et al. / Corrosion Science 51 (2009) 2316–2323 2319

Fig. 4. SEM micrographs of the BM samples after a DL-EPR testing: (a) A steel, and Fig. 6. SEM micrographs of the B steel BM after a DL-EPR testing up to: (a)
(b) B steel. The samples were previously polished in alumina 1 lm. E1 = 200 mVSCE, and (b) E2 = 50 mVSCE. Microstructure revealed after a single
polarization on the anodic direction, for the same testing sample. The sample was
previously polished in alumina 1 lm.

Fig. 5. B steel BM anodic polarization at 0.67 mV/s from 100 mVOCP to

+600 mVSCE.
Fig. 7. B steel BM cathodic polarization at 0.67 mV/s from +600 mVSCE to
100 mVOCP.

Fig. 5. The micrographs revealed a generalized corrosion attack

over the microstructures, without a significant difference between investigation of these processes were conducted through a voltage
them. The expected chromium depleted zone effect for the SACM scan only in the cathodic direction, from +600 mVSCE to 250 mV/
was not confirmed in the SMSS, according to the used experimental Ecorr, after the open circuit conditions were reached. Fig. 7 shows
techniques. The same occurrence was observed on the A steel, but the resulting polarization curve. The arrows at E3 = 100 mVSCE,
without any conclusion. and at E4 = 190 mVSCE, indicate the potential sweeping rate inter-
During the reactivation process on Fig. 3, the B steel BM also ruption in order to obtain SEM images. Fig. 8(a) shows that the first
presented two reactivation maximums. The separation and the reactivation peak (E3 = 100 mVSCE) corresponds to the martensite
2320 J.M. Aquino et al. / Corrosion Science 51 (2009) 2316–2323

Fig. 10. SEM micrographs of the HAZ samples next to the BM (2 mm from the
fusion line) after a DL-EPR testing: (a) A steel, and (b) B steel. The samples were
Fig. 8. SEM micrographs of the B steel BM after a DL-EPR testing up to: (a) previously polished in alumina 1 lm.
E3 = 100 mVSCE, and (b) E4 = 190 mVSCE. Microstructure revealed after a single
polarization on the cathodic direction, for the same testing sample. The sample was
previously polished in alumina 1 lm.

Fig. 9. DL-EPR curves for the A and B steel HAZs. The potential scanning direction is
indicated herein (passive vertex potential at 600 mV).

phase contour laths corrosion attack. The second peak

(E4 = 190 mVSCE) is due to the entire martensitic matrix corrosion
attack (intragranular), as showed in Fig. 8(b). According to the lit-
erature [24], the second reactivation peak could correspond to a Fig. 11. SEM micrographs in the middle of the HAZ samples (1 mm from the fusion
nickel rich phase, which is probably related to the retained austen- line) after a DL-EPR testing: (a) A steel, and (b) B steel. The samples were previously
ite phase in the SMSS. In addition, the ascribed nickel phase corro- polished in alumina 1 lm.
 J.M. Aquino et al. / Corrosion Science 51 (2009) 2316–2323 2321

sion attack at E3 is associated with a chromium content decrease

that caused the potential shift toward negative values. Such reacti-
vation occurrence was not observed in the A steel BM sample due
to its lower austenite content. The reactivation splitting occurrence
of the current density peak occurs on the martensite–austenite
steels containing the following [31]: (i) at least 13% of chromium;
(ii) alloyed with molybdenum; (iii) and an austenite content higher
than 25% resulting from a suitable tempering heat treatment. That
occurrence, on the DL-EPR reactivation curve, was detected in
stainless steel type 304 [1]; however, having interpretations differ-
ent from the ones herein.
The DL-EPR curves of the HAZ samples are representatively
shown in Fig. 9. The resulted curve is a measurement of the entire
HAZ from its fusion line to its 2 mm-away point. The B steel degree
of sensitization (iR/ia = 0.005) is much smaller than the A steel one
(iR/ia = 0.08). That is because A steel has a higher carbon content
than B steel, which results in a higher Cr-carbide precipitation, as
it will be showed in the micrographs. However, when comparing
the HAZ samples to the BM samples, a much lower DOS is obtained
because of the Cr-carbide precipitates possible redissolution at the Fig. 13. DL-EPR curves for the A and B steel WMs. The potential scanning direction
HAZ specific zones. That process occurred due to the gradient’s is indicated herein (passive vertex potential at 600 mV).
high-temperature in the HAZ, which is proper to the precipitates
redissolution [14,32]. Furthermore, a reduced time for the Cr-car-
bide precipitation could also contribute to the DOS lower levels SEM images were obtained in three distinct regions within the
in that region due to the EBW high cooling rate [14]. Studies HAZ samples of both steels, which were: (1) next to the BM
regarding the Cr-carbide precipitate redissolution are in progress. (2 mm from the fusion line), (2) in the middle of it (1 mm from
The second maximum current density in the activation/reacti- the fusion line), and (3) next to the fusion line.
vation processes were not observed in the HAZ samples, even Fig. 10(a and b) shows representative images of the HAZ next to
when 0.67 mV s 1 sweeping rates were used. the BM. An outstanding corrosion attack, with a high number of
holes, can be observed on the grain boundaries, and also inside
the A steel grains. The B steel samples are mainly characterized

Fig. 12. SEM micrographs of the HAZ samples next to the fusion line after a DL-EPR
testing: (a) A steel, and (b) B steel. The samples were previously polished in alumina Fig. 14. SEM micrographs of the WM samples after a DL-EPR testing: (a) A steel, and
1 lm. (b) B steel. The samples were previously polished in alumina 1 lm.
2322 J.M. Aquino et al. / Corrosion Science 51 (2009) 2316–2323

by an intragranular corrosion attack over the martensitic matrix. However, new methodologies regarding sample extraction and
This difference is a consequence of the alloying element distinct analysis [34,35] are being considered.
levels, mainly carbon [33]. The WM region DL-EPR curves are shown in Fig. 13. Both steels
Fig. 11(a and b) shows representative images in the middle of presented very similar activation current densities. However, the A
the HAZ. Clearly, the intergranular and the intragranular corrosion steel samples were more sensitized than the B steel ones due to
attacks are decreased in both steels. This is a consequence of the their higher reactivation current density (iR/ia = 0.013 and 0.004
generated thermal heat gradient, which enabled the Cr-carbide for the A and B steel, respectively). Nevertheless, the WM region
precipitate dissolution [14,33]. This process is more prominent in was the least sensitized region within the weldment. No evidence
the martensitic matrix because of the higher carbon diffusion coef- of the second maximum current density was observed in the ano-
ficient in that phase [7]. dic nor in the cathodic scan directions, even after 0.67 mV s 1
Fig. 12(a and b) shows HAZ images next to the fusion line. With- sweeping rates were applied. Fig. 14(a and b) shows scanning elec-
in that region, there is a formation of the d-ferrite phase, which is tron WM images after the DL-EPR test. The A steel presented very
the lighter color phase in the micrographs. An intensive corrosion small corrosion attack cavities, which were almost absent in the B
attack can be observed adjacent to that phase, which was revealed steel samples. There is no evidence of preferential precipitation
by the DL-EPR tests. This process occurred due to the lower carbon sites, intragranular or intergranular corrosion attacks in the
solubility in the d-ferrite phase [7], which resulted in the matrix quenched martensite. The carbide precipitates lower levels in the
supersaturation and the consequent precipitation of the Cr-car- WM region were observed due to the high cooling rates that oc-
bides that were adjacent to the d-ferrite phase. As the A steel has curred during the welding procedure [36]. This process enabled a
a higher carbon content than the B steel, it exhibited a more inten- very short remaining time in the Cr-carbide precipitation region,
sive precipitation, which was confirmed by the numerous holes of and a lower alloying element microsegregation which led to a re-
the corrosion attack parallel to the d-ferrite phase. The precipita- duced intergranular corrosion susceptibility. The high cooling rate
tion also occurred in the B steel samples, which were observed effect also supported the sensitization lower levels in the HAZ due
through the small cavities in Fig. 11(b), in spite of their very small to the reduced time period within which the material stayed in the
carbon content. The HAZ micrographs elucidated the difference be- Cr-carbide precipitation region [14].
tween the reactivation current densities of the A and B steels The degree of sensitization was compared within different re-
(Fig. 9). gions of the weldment, and is shown in Fig. 15(a and b). The HAZ
It was not possible to do the HAZ samples’ TEM due to their re- samples presented levels of sensitization extremely lower which
duced number, and to the difficulty of obtaining these samples. were comparable to the WM ones. The HAZ performance concern-

Fig. 15. DOS and pitting potential values within the supermartensitic stainless steel weldments: (a) A steel, and (b) B steel.
 J.M. Aquino et al. / Corrosion Science 51 (2009) 2316–2323 2323

ing the intergranular corrosion was completely different from [7] E. Folkhard, Welding Metallurgy of Stainless Steel, first ed., Springer-Verlag,
New York, 1988.
what is reported in the literature [12,13,35]. However, few re-
[8] K. Kondo, K. Ogawa, H. Amaya, H. Hirata, M. Ueda, H. Takabe, Y. Miyazaki, Alloy
searches on SMSS, as well as on its weldments that are done design of super 13Cr martensitic stainless steel (Development of super 13Cr
through high power density processes, are available. The most martensitic stainless steel for line pipe – 1), in: Paper of the Supermartensitic
important point to consider is the welding process, which has dif- Stainless Steels, Brussels, Belgium, 1999.
[9] Y.C. Lin, S.C. Chen, Effect of residual stress on thermal fatigue in a type 420
ferent characteristics when comparing it to the conventional pro- martensitic stainless steel weldment, J. Mater. Process. Technol. 138 (2003)
cess [14,15]. Among those characteristics, the sharp temperature 22–27.
variation of the HAZ is the most important because it mainly [10] E. Ladanova, J.K. Solberg, T. Rogne, Carbide precipitation in haz of multipass
welds in titanium containing and titanium free supermartensitic stainless
causes the Cr-carbide precipitation/dissolution in it. steels part 1 – proposed precipitation mechanisms, Corros. Eng. Sci. Technol.
Fig. 15 also includes the weldment pitting potential values. 41 (2006) 143–151.
These results are found in paper [22]. According to the literature [11] H. Nakamichi, K. Sato, Y. Miyata, M. Kimura, K. Masamura, Quantitative
analysis of Cr-depleted zone morphology in low carbon martensitic stainless
[16–20], the presence of chromium depleted regions are related steel using FE-(S)TEM, Corros. Sci. 50 (2008) 309–315.
to a pit nucleation and growth. As the sensitization increases, the [12] L. Coudreuse, V. Ligier, C. Lojewski, P. Toussaint, Environmental induced
pitting potential decreases, which results in a lower corrosion cracking (SSC and SCC) in supermartensitic stainless steels (SMSS), in: Paper of
the Supermartensitic Stainless Steels, Brussels, Belgium, 2002.
resistant material. [13] T. Rogne, M. Svenning, Intergranular corrosion of supermartensitic stainless
steel – a high temperature mechanism? in: Paper of the Supermartensitic
4. Conclusion Stainless Steels, Brussels, Belgium, 2002.
[14] H.T. Lee, J.L. Wu, The effects of peak temperature and cooling rate on the
susceptibility to intergranular corrosion of alloy 690 by laser beam and gas
The BM was the most sensitized region within the SMSS weld- tungsten arc welding, Corros. Sci. 51 (2009) 439–445.
ment due to its tempering process, which resulted in its Cr-carbide [15] H.T. Lee, J.L. Wu, Correlation between corrosion resistance properties and
thermal cycles experienced by gas tungsten arc welding and laser beam
precipitation. The WM and the HAZ were the lower sensitized re- welding alloy 690 butt weldments, Corros. Sci. 51 (2009) 733–743.
gions due to the WM high cooling rates, as well as to the Cr-carbide [16] G.T. Burstein, S.P. Vines, Repetitive nucleation of corrosion pits on stainless
precipitate redissolution at certain regions of the HAZ. That was steel and the effects of surface roughness, J. Electrochem. Soc. 148 (2001)
B504–B516.
confirmed by the SEM images within the HAZ samples. [17] G.S. Frankel, Pitting corrosion of metals, J. Electrochem. Soc. 145 (1998) 2186–
The HAZ of an electron beam top pass weld is susceptible to 2198.
sensitization. Moreover, the precipitation within that region [18] C.O.A. Olsson, D. Landolt, Passive films on stainless steels: chemistry, structure
and growth, Electrochim. Acta 48 (2003) 1093–1104.
showed to vary according to the attained temperature during
[19] I. Reynaud-Laporte, M. Vayer, J.P. Kauffmann, R. Erre, An electrochemical-AFM
welding. The d-ferrite phase (microstructural effect) also influ- study of the initiation of the pitting corrosion of a martensitic stainless steel,
enced the HAZ sensitization, because of its low carbon solubility. Microsc. Microanal. Microstruct. 8 (1997) 175–185.
[20] M.P. Ryan, D.E. Williams, R.J. Chater, B.M. Hutton, D.S. McPhail, Why stainless
The austenite content did not seem to influence the sensitization
steel corrodes, Nature 415 (2002) 770–774.
or the pit potential. [21] T.L.S.L. Wijesinghe, D.J. Blackwood, Real time pit initiation studies on stainless
The splitted reactivation peak in the A steel BM samples corre- steels: the effect of sulphide inclusions, Corros. Sci. 49 (2007) 1755–1764.
sponded to two oxidation processes, which were possibly related [22] J.M. Aquino, C.A. Della Rovere, S.E. Kuri, Localized corrosion susceptibility
of supermartensitic stainless steel in welded joints, Corrosion 64 (2008)
to: (1) the martensitic phase contour laths, and (2) the martensitic 35–39.
matrix, which may be associated to the austenitic phase content in [23] J. Enerhaug, O. Grong, U.M. Steinsmo, Factors affecting initiation of pitting
the SMSS. The second maximum current density in the activation corrosion in super martensitic steels weldments, Sci. Technol. Weld. Join. 6
(2001) 330–338.
process was not splitted into two peaks. Furthermore, that occur- [24] V. Cíhal, R. Stefec, On the development of the electrochemical potentiokinetic
rence did not seem to be related to the chromium depleted effect. method, Electrochim. Acta 46 (2001) 3867–3877.
A correlation between sensitization and pitting potential was [25] ASTM standard E 975-95, Standard practice for X-ray determination of
retained austenite in steel with near random crystallographic orientation,
established in the SMSS weldments, indicating that the probability ASTM, PA, 1995, pp. 1–6.
of pitting corrosion enhanced as the sensitization increased. How- [26] V. Vodarek, M. Tvrdy, A. Korgak, Heat treatment of supermartensitic steels, Inz.
ever, it was not confirmed that chromium depleted regions were Mater. 5 (2001) 939–941.
[27] L. Felloni, S.S. Traverso, G.L. Zucchini, G.P. Cammarota, Investigation on the
the preferential sites for pit nucleation and growth.
second anodic current maximum on the polarization curves of commercial
stainless steels in sulphuric acid, Corros. Sci. 13 (1973) 773–779.
Acknowledgements [28] A.A. Hermas, M.S. Morad, K. Ogura, A correlation between phosphorous
impurity in stainless steel and a second anodic current maximum in H2SO4,
Corros. Sci. 41 (1999) 2251–2266.
The authors thank the financial support provided by CAPES and [29] O.L. Riggs Jr., The second anodic current maximum for type 430 stainless steel
Dr. Celso Roberto Ribeiro for supplying the steel samples. in 0.1N H2SO4, Corrosion 31 (1975) 413–415.
[30] M.B. Rockel, Interpretation of the second anodic current maximum on
polarization curves of sensitized chromium steels in 1N H2SO4, Corrosion 27
References (1971) 95–103.
[31] V. Cíhal, M. Blahetová, J.Hubácková, Z. Krhutová, S. Lasek, K. Mazanec,
[1] V. Kain, K. Chandra, K.N. Adhe, P.K. De, Detecting classical and martensite- Corrosion and structural testing of martensitic steels by electrochemical
induced sensitization using the electrochemical potentiokinetic reactivation polarization method, in: Paper of the Supermartensitic Stainless Steels,
test, Corrosion 62 (2005) 587–593. Brussels, Belgium, 2002.
[2] G.H. Aydogdu, M.K. Aydinol, Determination of susceptibility to intergranular [32] P.H.S. Cardoso, C. Kwietniewski, J.P. Porto, A. Reguly, T.R. Strohaecker, The
corrosion and electrochemical reactivation behaviour of AISI 316L type influence of delta ferrite in the AISI 416 stainless steel hot workability, Mater.
stainless steel, Corros. Sci. 48 (2006) 3565–3583. Sci. Eng. A 351 (2003) 1–8.
[3] N. Alonso-Falleiros, M. Magri, I.G.S. Falleiros, Intergranular corrosion in a [33] O.M. Akselsen, G. Rorvik, P.E. Kvaale, C. van der Eijk, Microstructure-property
martensitic stainless steel detected by electrochemical tests, Corrosion 55 relationships in HAZ of new 13% Cr martensitic stainless steels, Weld. J. 83
(1999) 769–778. (2004) 160–167.
[4] M. Kimura, Y. Miyata, T. Toyooka, Y. Kitahaba, Effect of retained austenite on [34] H. Yanliang, B. Kinsella, T. Becker, Sensitisation identification of stainless steel
corrosion performance for modified 13% Cr steel pipe, Corrosion 57 (2001) to intergranular stress corrosion cracking by atomic force microscopy, Mater.
433–439. Lett. 62 (2008) 1863–1866.
[5] P.D. Bilmes, C.L. Llorente, L. Saire Huamán, L.M. Gassa, C.A. Gervasi, [35] C. Garcia, M.P. de Tiedra, Y. Blanco, O. Martin, F. Martin, Intergranular
Microstructure and pitting corrosion of 13CrNiMo weld metals, Corros. Sci. corrosion of welded joints of austenitic stainless steels studied by using an
48 (2006) 3261–3270. electrochemical minicell, Corros. Sci. 50 (2008) 2390–2397.
[6] T. Hara, H. Asahi, Effect of d-ferrite on sulfide stress cracking in a low carbon 13 [36] S.A. David, S.S. Babu, J.M. Vitek, Welding: solidification and microstructure, J.
mass% chromium steel, ISIJ Intern. 40 (2000) 1134–1141. Miner. Met. Mater. Soc. 55 (2003) 14–20.